Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Differential Responses of Dinitrogen Fixation, Diazotrophic Cyanobacteria and Ammonia Oxidation Reveal a Potential Warming-Induced Imbalance of the N-Cycle in Biological Soil Crusts

  • Xiaobing Zhou,

    Affiliations Arizona State University, School of Life Sciences,Tempe, AZ 85287, United States of America, Xinjiang Institute of Ecology and Geography, Key Laboratory of Biogeography and Bioresource in Arid Land, Chinese Academy of Sciences, Urumqi, Xinjiang 830011, China

  • Hilda Smith,

    Affiliation U. S. Geological Survey, Southwest Biological Science Center, Moab, UT 84532, United States of America

  • Ana Giraldo Silva,

    Affiliations Arizona State University, School of Life Sciences,Tempe, AZ 85287, United States of America, Center for Fundamental and Applied Microbiomics, Biodesign Institute, Arizona State University, Tempe, AZ 85287, United States of America

  • Jayne Belnap,

    Affiliation U. S. Geological Survey, Southwest Biological Science Center, Moab, UT 84532, United States of America

  • Ferran Garcia-Pichel

    ferran@asu.edu

    Affiliations Arizona State University, School of Life Sciences,Tempe, AZ 85287, United States of America, Center for Fundamental and Applied Microbiomics, Biodesign Institute, Arizona State University, Tempe, AZ 85287, United States of America

Differential Responses of Dinitrogen Fixation, Diazotrophic Cyanobacteria and Ammonia Oxidation Reveal a Potential Warming-Induced Imbalance of the N-Cycle in Biological Soil Crusts

  • Xiaobing Zhou, 
  • Hilda Smith, 
  • Ana Giraldo Silva, 
  • Jayne Belnap, 
  • Ferran Garcia-Pichel
PLOS
x

Abstract

N2 fixation and ammonia oxidation (AO) are the two most important processes in the nitrogen (N) cycle of biological soil crusts (BSCs). We studied the short-term response of acetylene reduction assay (ARA) rates, an indicator of potential N2 fixation, and AO rates to temperature (T, -5°C to 35°C) in BSC of different successional stages along the BSC ecological succession and geographic origin (hot Chihuahuan and cooler Great Basin deserts). ARA in all BSCs increased with T until saturation occurred between 15 and 20°C, and declined at 30–35°C. Culture studies using cyanobacteria isolated from these crusts indicated that the saturating effect was traceable to their inability to grow well diazotrophically within the high temperature range. Below saturation, temperature response was exponential, with Q10 significantly different in the two areas (~ 5 for Great Basin BSCs; 2–3 for Chihuahuan BSCs), but similar between the two successional stages. However, in contrast to ARA, AO showed a steady increase to 30–35°C in Great Basin, and Chihuhuan BSCs showed no inhibition at any tested temperature. The T response of AO also differed significantly between Great Basin (Q10 of 4.5–4.8) and Chihuahuan (Q10 of 2.4–2.6) BSCs, but not between successional stages. Response of ARA rates to T did not differ from that of AO in either desert. Thus, while both processes scaled to T in unison until 20°C, they separated to an increasing degree at higher temperature. As future warming is likely to occur in the regions where BSCs are often the dominant living cover, this predicted decoupling is expected to result in higher proportion of nitrates in soil relative to ammonium. As nitrate is more easily lost as leachate or to be reduced to gaseous forms, this could mean a depletion of soil N over large landscapes globally.

Introduction

Biological soil crusts (BSCs) are complex soil surface communities with cyanobacteria, green algae, lichens or mosses as primary producers [1]. They also contain significant populations of heterotrophic bacteria, archaea [2] and fungi [3], as well as chemolithotrophic microbes [4]. They are common in dryland regions, where they can reach up to 70% of the living cover [5], and thus are an important element of these ecosystems both locally and globally [6]. Such communities have colonized soil surfaces at least since the Proterozoic and represent the only known terrestrial ecosystem at that time [7]. Cryptogamic covers around the world, much of which are BSCs, can fix as much as 49 Tg of nitrogen (N) per year, which accounts for nearly half of the global biological N2 fixation on land [8]. Biological soil crusts affect most dryland ecosystem processes. They stabilize the soil surface [9, 10], fix atmospheric carbon (C) and N [8, 1115], alter the soil content of many metals and metalloids [16], may help maintain vascular plant diversity [17, 18], and affect hydrological processes [1921].

Dinitrogen fixation is also the entry point of an active N cycle in BSCs. Most N2 fixation in BSCs is done by heterocystous cyanobacteria (either free-living or in association with lichens). Surveys of nifH genes have revealed that three generic entities (Nostoc spp., Scytonema spp. and Tolypothrix/Spirirestis spp.) are dominant in BSCs [22], but a few heterotrophic bacteria are also important in early successional BSCs that are devoid of heterocystous cyanobacteria [23]. Variable rates in time and space and with BSC successional stage have been documented in a wealth of studies [12, 14, 15, 24, 25]. Within a given BSC type, moisture and temperature (T) are the two most important abiotic factors that determine N2 fixation rates [12, 26, 27].

Ammonia oxidation (AO) behaves as the crucial step to transform the fixed, reduced N to nitrate (NO3-). It is the first and rate-limiting step of nitrification, with chemolithoautotrophic Proteobacteria and archaea in the phylum Thaumarchaeota conducting this process in soils and BSCs [2, 28]. The AO bacteria and archaea seem to differ in their relative importance with changes in factors such as moisture, pH, and fertilization treatments [2932]. Ammonia oxidation is a prominent transformation in BSCs across biogeographic regions in the Southwestern USA [14] and seasons [33], usually showing rates similar in magnitude to those of N2 fixation [14, 33].

Denitrification indicates the reduction of nitrates to gaseous N2 by microorganisms. In North American soil BSCs, denitrification and anaerobic ammonia oxidation (anammox), whenever measured concurrently, show orders-of -magnitude rates lower than N2 fixation and AO [14, 33]. It is not known why rates of denitrification and anammox are so low, although it is possibly due to low denitrifier population or insufficient NH4+ and NO3-to act as substrates [14, 34]. Thus, the balance between AO and N2 fixation rates largely determines the net N budget in BSCs.

Nitrogen cycling in microbial communities of drylands are structured by abiotic factors to a stronger degree than what is generally observed in more mesic systems [35]. Temperature is one of chief factors that affect all aspects of the N cycle, including N2 fixation and AO. In drylands, T varies greatly within a day and among seasons and years. Indeed, T changes (between 5°C and 30°C) can be more important than variation in soil water content (between 30% and 80% of field capacity) in determining the relative dominance of N transformation rates in BSCs from a semi-arid grassland [36]. In BSCs, T can also shape microbial community composition, including those of N2 fixing cyanobacteria [37] and AO microbes, selecting for AO archaea in warmer deserts and for AO bacteria in cooler deserts [35]. And yet, many studies that have simultaneously compared multiple transformations in the N cycle have not directly addressed the potential effects of T on interactions among different processes. Because these effects do not need to scale equally for different N transformations, T variations may decouple these processes, producing divergent net outcomes.

In this study, we describe the T sensitivity of both the acetylene reduction assay (ARA), as an indicator of N2 fixation, and AO, on BSCs using short-term acclimation experiments. In order to gauge potential effects of long-term acclimation to different T regimes, we included BSCs from a cool (Great Basin) and hot (Chihuahuan) desert. The BSCs were incubated under different T treatments with sufficient moisture supply to explore the sensitivities of both ARA and AO to site-relevant T variations, and also to ascertain potentially differential responses to BSCs from different locations under current and future Ts. We hypothesized that rates of ARA and AO would display optimal Ts that would be relatively higher in hot deserts compared with cool deserts, due to long-term adaption of their respective microbial populations, and sought to establish if, by virtue of having differential effects on different processes, future changes in T may change the balance in dryland N cycles.

Materials and Methods

Site descriptions and sample collection

Our samples were obtained from Hill Air Force Base in the Great Basin Desert (USA) (41.10° N, 113.00° W) and from the Jornada Experimental Range in the Chihuahuan Desert (USA) (32.55° N, 106.72° W), which are typical of cool and hot desert ecosystems, respectively. At Hill, mean annual precipitation is 200 mm and mean air T is 3°C in January and 34°C in July. At the Jornada, the mean annual precipitation is 280 mm and the mean air T is 14°C in January and 36°C in July.

Intact BSCs were collected to 1.5 cm depth with a putty knife in October of 2014 in the Great Basin and in March 2015 in the Chihuhuan Desert, respectively. The patches of intact BSC were randomly collected at each site and transported to Arizona State University in wooden frames that preserved their integrity. Samples were kept dry (inactive) under room temperature (around 25°C) and dark until the start of the laboratory experiment. The BSCs in the field was extremely dry so that the microbial activity was very low, and humidity in the storage unit was also kept low [38]. The experiment was started within days of each sampling. Visual inspection (such as color and surface roughness determination) is an indicator of rough BSCs species composition and biomass, and was used in our study to describe our BSC types. We used early successional (light; dominated by motile Microcoleus-like cyanobacteria and late successional crusts from each site (dark or lichen crusts), according to the naturally occurring climax community [39].

In the Great Basin soils were silty clays. In the Chihuahuan Desert, light and lichen BSCs were used. The two BSCs in the Chihuahuan desert were on sandy soil. Light BSCs refers to the light coloration of the soil surface, and are dominated by early successional, motile cyanobacterial species [39]. Dark BSCs refer to the darker color imparted to the soil surface by heavily pigmented non-motile cyanobacteria and develop secondarily from light crust in a later successional stage [1]. Lichen BSCs are typically a late successional crust type, where the surface is dominated by lichens [40].

Soil parameters

Pigment, geochemical parameters, and pH were determined and three replicates were included for each measurement. For pigment concentrations, 2 cm2 of collected BSC sample was ground, put into a 15 ml falcon tube, and mixed with 10 ml 90% acetone. The mixture was kept in dark at 4°C for 24 h, and then centrifuged at 3000 × g for 10 min. Supernatant absorbance was read at 384nm, 490 nm, and 663nm. Chlorophyll a and scytonemin contents were calculated based on Garcia-Pichel and Castenholz [41]. Soil geochemical parameters (total carbon (C), total N (N), NH4+-N and NO3--N) for each sample were determined commercially at the Core Research Facilities Administration, Arizona State University. Ten g fresh soil samples were exacted with 50 ml 2M KCl solution, and NH4+-N and NO3--N were analyzed by a AQ2 Discrete Analyzer system (SEAL Analytical, Inc., WI, U.S.A.). For total C and N, air-dried soil samples were ground in a ball mill (SPEX SamplePrep, NJ, U.S.A.) and analyzed on a PE2400 Elemental Analyzer (PerkinElmer, Inc, MA, U.S.A.). Soil solution pH was measured with a 1: 5(W/V) of BSCs sample: double-dionized water that was shaken for 3 min, allowed for equilibrate for 24 h, and determined with a pH meter.

DNA exaction and real-time PCR

Sample DNA was extracted from 0.4 g BSCs using the PowerSoil DNA isolation Kit (MO BIO Laboratories, Inc. USA), and quantified using the SYBR Green assay as described by Brankatschk et al. [42]. SYBR green based real-time PCR assays were run on an ABI7900HT thermocycler (Applied Biosystems, Foster City, CA). Reaction volume was 20 μL and iTaq SYBR GREEN Fast PCRmaster mix. PCR primer for N2 fixation gene were nifHF (AAAGGYGGWATCGGYAARTCCACCAC) and nifHR (TTGTTSGCSGCRTACATSGCCATCAT), for amoA were amoA1f mod (GGGGHTTYTACTGGTGGT) and AmoA-2R’(CCTCKGSAAAGC- CTTCTTC). Each PCR run included triplicate sample templates, calibration standards and no-template control.

Temperature pre-treatments

One cm deep, intact BSCs pieces were cut and modified carefully with cutter blade from the patch of the field sampled BSCs and then placed into the wells of a six well (diameter = 3.5 cm) tissue culture plate. One well was reserved for autoclaved samples (controls) and five for treatment samples. Each well contained around 10 g of BSC material. Double deionized water was filled until the soil reached its water holding capacity (a thin film of water remained on the sample surface). The plates were covered and incubated at different Ts in separate incubators, all illuminated at a light intensity of around 80 μmol (photons) m-2 s-1. Analyzed Ts were -5, 5, 10, 15, 20, 25, 30 and 35°C. The levels of illumination roughly correspond to that of a heavily overcast, rainy morning. The temperature ranges were selected to encompass the variability in air temperature changes around the two sites. During the whole incubation period, the samples were kept intact. Water loss in the well was slow during the incubation and soil water content was stable. The pre-incubation time was 24 h, and the incubated samples were used for the measurements of potential N2 fixation rates and AO rates, each at their respective T. Eight independent replicates were used in potential N2 fixation and five for AO rates measurements, because ARA values are often more variable.

Potential N2 fixation rates

Potential N2 fixation rates were estimated using ARA according to the methods of Belnap [12]. The samples for incubation were cut out of the collected intact material using clear, gas–tight tubes (diameter of 2.54 cm, length of 9.5 cm), put in the tubes, and closed by rubber stoppers with the top end having septum ports for sampling (eight replicates for each treatment). Tubes were injected with acetylene (C2H2) to create a 10% C2H2 atmosphere and incubated for 5 h at the various Ts. Gas (4 ml) of the headspace within the tubes was collected and analyzed for C2H2 and ethylene (C2H4) content. The samples were analyzed on a Shimadzu GC-14 A gas chromatograph, using helium as the carrier gas (30 ml min-1). Calibrations with ethylene standards were done at the time of observations. Results of the observed nitrogenase activity as nmol C2H4 m-2 h-1 were transformed to N input (μmol N m-2 h-1) using the theoretical conversion ratio 3 [26].

Potential AO rates

Potential (ammonium amended) aerobic AO rates were determined according to the methods described by Strauss et al. [14] and Maruseko [35]. A mixture of 20 mM sodium chlorate (NaClO3, an inhibitor of nitrite reduction) and 1mM ammonium sulfate ((NH4)2SO4) was prepared and adjusted to pH 7.2 using KH2PO4 and K2HPO4. The samples, which had been pre-incubated under different Ts, were slurried with this mixture (1:4 w/v) (five replicates for each treatment). These samples were incubated for 6 h in the dark in an environmental shaker, set at the corresponding pre-incubation T. Aliquots of 5 ml were removed every 3 h and mixed with 5 ml 4M KCl. Samples were centrifuged at 5,000 × g for 5 min to pellet the debris. The supernatant was filtered with Whatman #42 filters and then nitrite (NO2-) was determined spectrophotometrically at 520 nm, after color reagent reactions. NO2- production was calculated as the linear increase in NO2- concentrations over time (μmol N m-2 h-1).

Cyanobacterial cultures isolation, growth and thermophysiology

We isolated the N2-fixing cyanobacteria, Nostoc spp., Tolypothrix spp. and Scytonema spp., from each of the field locations by using enrichment cultures in liquid media where N2 was the only N source (B110.; see Yeager et al. [22]). Resulting colonies were then streaked on 1.5% agar plates and observed with optical microscopy to corroborate the presence of only one of the desired morphotypes per culture (unialgal). For each culture the 16S rRNA gene was sequenced using cyanobacterial specific primers [43], to corroborate identity by blasting against GenBank using BLASTN [44]. Cultures are maintained in the culture collection of the Garcia Pichel’s laboratory at ASU and are publicly available upon request. To evaluate responses to T under N2-fixing conditions, strains were inoculated in 20 mL of N free media (B110) at 5% v/v in 50 mL culture bottles. The cultures were incubated at different T (15, 25, 30, 35 and 40°C) in a 12 h photoperiod, and illuminated for 30 days at 20–27 μmol m-2 s-1. Growth was estimated visually in three categories: optimal, suboptimal and no-growth, which included dying cultures. The experiment was replicated in full, and growth in either one of the trials reported as a positive.

Statistics

For each type of BSC, one-way ANOVA was used to determine the differences in ARA and AO rates among the different T treatments, and LSD method was applied for multi-comparisons. Normality and homoscedasticity were tested before ANOVA analysis. Within each desert, t-test statistics was performed on the two indices between light and dark/lichen BSCs at each T. The statistical analysis were conducted using SAS software (Version 8.0, SAS Institute Inc., Cary, NC, U.S.A.) at the α = 0.05 level. The Arrhenius equation, which displays the logarithm of kinetic constants versus inverse T, was used to gauge the effect of T on the rates of N transformation. Only the portion of each dataset where the rates increased with T was used so as to explicitly avoid T ranges where process rates were saturated. The form of the equation can be written as ln (K) = ln (β)+ α(1/T), where K is rate, T is absolute T in degrees Kelvin, and ln (β) is the value of the true y-intercept (1/T = 0) and α is the slope of the regression line. The relationship between ln (k) and 1/T was determined with reduced major axis (RMA) regression. The slope α and ln (β) were obtained using standardized major axis (SMA) regression SMATR software package protocols. The software package was also used for post-hoc multiple comparison of slopes among the eight different groups. The Q10, which indicates the average fold increase in rate for an increase of 10°C, was also calculated from exponential regressions of the same datasets. Significance was defined as P < 0.05.

Results

Biological soil crust characterization

The main characteristics of the BSC used are presented in Table 1. As intended, differences in BSC successional stage (maturity) were reflected in the proxy of phototrophic biomass (chlorophyll a), which for a given locality were lower in light than in dark/lichen BSCs. Samples from the Great Basin had significantly more chlorophyll a than those of Chihuahuan Desert (P < 0.05), regardless of successional stage. More mature BSCs had a higher concentration of scytonemin, reflective of larger numbers of heterocystous cyanobacteria [45]. Soil pH was moderately alkaline in all BSCs, and slightly, but significantly, lower in samples from the Chihuahuan than those from the Great Basin deserts. The abundance of the functional genes amoA and nifH and NH4-N concentrations was higher in later than earlier successional BSCs in both deserts. The abundance of amoA and nifH in early successional BSCs was similar in both deserts; however, the abundance of amoA and nifH in late successional BSCs from the Great Basin were 5.1 and 2.0 times higher, respectively, than those in late successional BSCs from the Chihuahuan desert. The BSCs from the Great Basin had relatively higher concentrations of NH4-N and NO3-N than those in the Chihuahuan desert.

thumbnail
Table 1. Pigment concentrations and soil chemistry and gene abundance of nifH and amoA in different BSCs types of the Great Basin and Chihuahuan Desert (mean ± s.e., n = 3).

https://doi.org/10.1371/journal.pone.0164932.t001

Acetylene reduction rates

We found similar trends with T of potential N2 fixation rates for light, dark and lichen BSCs from either desert (Fig 1). In all BSCs, rates increased with increasing T until reaching an optimal range where they leveled off, which was reached between 15 and 20°C. The plateau in optimal T ranged between 15 and 30°C, with ARA declining markedly at 35°C, in the Great Basin samples. In the Chihuahuan, optima plateaus were around 15–25°C and above that there was a slight but significant decrease. Rates were significantly higher in dark/lichen compared to light successional BSCs in both deserts at most Ts. Rates in the Great Basin BSCs were much higher than those in the samples from the Chihuahuan. Below the optimum range, the sensitivity of ARA (as judged by Q10 or the slope of Arrhenius plots on Table 2 and S1 Fig) was not significantly different between the two BSC types from the same region, but BSCs from the northern Great Basin had a much stronger response to T than Chihuahuan Desert BSCs.

thumbnail
Fig 1. Nitrogen fixation (Acetylene reduction) rates of dark (solid black symbols), light (grey symbols) and lichen (solid black symbols) BSCs in the Great Basin and Chihuahuan Desert.

Symbols show the mean of n = 8 determinations and error bars depict standard errors. Different lowercase letters indicated significant differences (p < 0.05) for comparisons among temperatures within single BSC type. An asterisk indicates significant differences between dark and light BSCs for a single temperature.

https://doi.org/10.1371/journal.pone.0164932.g001

thumbnail
Table 2. Responsivity of N transformations to T among the BSCs studied, gauged by estimates of the slope of a regression between the natural logarithm of the process rates and the inverse of temperature (Arrhenius plots; see full data in S1 Fig).

Only the portion of any dataset where rates increased with T was used. Slopes that were not significantly different at the 95% were assigned same letter in the multi-comparison column. Q10s, giving the average fold increase in rate for an increase of 10°C, are also included, calculated from regression curves of the same datasets.

https://doi.org/10.1371/journal.pone.0164932.t002

Potential AO rates

The potential AO rates of the BSCs in the two deserts also increased with T, but did not show a wide plateau of close-to-optimal Ts such as those found for potential N2 fixation (Fig 2). Chihuahuan Desert samples increased steadily within the ranges tested; Great Basin Desert samples suffered losses at the highest T. Similar to ARA, the sensitivity of AO to T, as gauged by Q10 or the slopes of Arrhenius plots (S1 Fig, Table 2), was similar among BSC types in a given desert, but much higher in BSCs from the Great Basin than in those from the Chihuahuan Desert.

thumbnail
Fig 2. Potential AO fixation rates of dark (solid black symbols) and light (grey symbols) and lichen (solid black symbols) BSCs in the Great Basin and Chihuahuan deserts.

Symbols show the mean of n = 5 determinations, anderror bars depict standard errors. Different lowercase letters indicated significant differences (p < 0.05) for comparisons among temperatures within a single BSC type. An asterisk indicates significant differences between dark/lichen and light BSCs for a single temperature.

https://doi.org/10.1371/journal.pone.0164932.g002

Growth of N-fixing cyanobacteria as a function of T

All 30 strains of diazotrophic cyanobacteria showed robust diazotrophic growth at 15 and 25°C, regardless of their origin (Table 3). However, clear differences were seen at high Ts. No Nostoc spp. Cultures grew well at 30°C or above, and only 4 out of 8 strains of Tolypothrix spp. did, but all strains of Scytonema spp. grew well. At 35 and 40°C, from all strains of all taxa, only 11 out of 12 Scytonema spp. strains grew.

thumbnail
Table 3. Diazotrophic growth capability as a function of temperature in cultured cyanobacteria isolated from the Great Basin and the Chihuahuan deserts.

https://doi.org/10.1371/journal.pone.0164932.t003

Discussion

Temperature influence on the N transformations of biological soil crusts

We found a similar pattern of N2 fixation in the four BSC types/origins: ARA increased with T until reaching a wide optimal plateau, after which it decreased. The plateau varied somewhat with locale and maturity, around the range 15–30°C. Most studies have found a similar pattern and optimal ARA rates at this T range, regardless of BSC type or location. In the vicinity of Moab, UT, Barger et al. [26] found that both light and dark BSCs from cool deserts (Canyonlands National Park, USA) had peak rates in the range of 20 to 30°C. In the Tenneger Desert, China, Su et al. [15] found an optimal range of ARA at 15–30°C. Global literature reviews that include studies from the Arctic, Antarctica, Scotland, Canada, South Africa, and subalpine regions also found the optimal T range of 20–30°C [26, 46]. Because similar optimal ranges have been observed across so many soil types and environmental conditions, it could represent basic 1) thermal, 2) nutrient, or 3) energy constraints on the few dominant organisms generally associated with N2 fixation in BSCs (largely the cyanobacteria Nostoc spp., Scytonema spp. and Tolypothrix/Spirirestis spp. [22]; although some heterotrophic diazotrophs do help when those cyanobacteria are absent [23]). To compare this optimal range for N2 fixation with the T dependence of their diazotrophic metabolism, we carried out growth experiments with representative isolates of each of the major cyanobacterial diazothrophs (Table 3). It is clear from this data that a large proportion of the strains cannot continue growing at T above 30°C, and that the genera that dominate in Great Basin crusts (Tolypothrix spp. and Nostoc spp. [37]) are much more sensitive than those that are more common in the Chihuahuan, warmer crusts (Scytonema spp. [37]). Thus the phenomenology of N2-fixation (Fig 1) can likely be traced to the community composition and physiology of their inhabiting diazotrophic flora.

The patters in the crusts are also consistent with nitrogenase degradation above 39°C [47, 48]. Zhu and Brill [49] and Brooks et al. [47, 49] also both found that regulatory proteins coded by nifA were suppressed at 37–41°C. In general terms, the T dynamics of N2 fixation are similar to those found for net photosynthesis in BSCs [50, 51]. It is possible that the effects seen in N2 fixation are a consequence of a limited C or ATP supply, both in the field and in our cultures. Regardless of the mechanisms behind the observed decline in ARA at higher T, it has implications for scaling of T with other processes involved in the BSC N cycle, as is discussed below.

Interestingly, while there were no major differences in the optimal ranges for the BSCs between the cool and hot desert tested, the samples in the cool Great Basin did show a much more precipitous decrease at high T, probably due to the low incidence of Scytonema spp. cyanobacteria there. In this dual sense our results are only partially consistent with our initial hypothesis that ARA rates would display relatively higher optimal Ts in hot compared with cool deserts. This implies that the N fixing cyanobacteria present in these deserts are not just being selected for their ability to fix N under normal thermal regimes. Much clear effects can be seen from the literature: optimal T for ARA is 35–40°C for cyanobacterial BSCs s in the cool Loess Plateau region of China [27], but only 25°C in BSCs from the High Arctic [52]. Our ARA rate did, however, show significant (Table 2) and consistent differences at the low end of the T range: late successional BSCs from the cooler Great Basin showed a much more marked response to T than those from the hot desert, with Q10 values almost twice as large within below 20°C, indicating that the extant populations in those samples can make better use of “warm spells” or seasons. A Q10 value measures the T sensitivity of an enzymatic reaction rate or physiological process with a T increase of 10°C. Most biological systems show Q10 values about 2–3 [53], but they can vary significantly between processes, even in a single system. For instance, between 25 and 15°C, Q10 values in the H2-evolving symbioses of Rhizobium japonicum ranged from 2.0 to 2.7 for CO2 evolution, 1.3 to 2.4 for C2H2 reduction, and 3.2 to 3.7 for H2 evolution [54].

In the case of potential AO, we did not observe a wide plateau of optimal rates. The relatively higher optimal T for BSCs from the Chihuahuan than from Great Basin was consistent with our initial hypothesis that AO rates would be higher in hot desert compared to cool desert soils. The relationship of AO rates was clearly exponential with T for most of the T range we tested (Fig 2; Table 2), thus behaving much like a purely chemical reaction and fitting well a log-linear relationship. In the present study, only samples from the Great Basin showed AO inhibition at higher Ts (a symptom of the organismal or biochemical machinery failing), which is consistent with an adaptation of the ammonia oxidizers to the local T regime, as would be anticipated based on past molecular studies [35].

Balance between N2 fixation and AO as influenced by temperature

We found consistent and significant trends in T response of ARA and AO. The Arrhenius slopes and Q10‘s of the Chihuahuan Desert samples (Q10 = 2.50–3.06) were roughly half of the Great Basin Desert samples (Q10 = 4.80–5.15), an indication that ammonia oxidizers, like N fixers in the cool deserts are much less T-dependent that those found in the hot desert samples and unusually responsive to T. The reasons behind this are unclear to us.

The major differential effect of T on ARA and AO was the presence and absence, respectively, of a ‘saturation’ or plateau at the mid to warmer end of the T range, rather than a differential response to T at the lower end (in fact the Q10 values for both processes at lower Ts were not significantly different in either geographic area; Table 2), regardless of geography or BSC type. This resulted in an imbalance between ARA and AO at warmer Ts that was not documented at lower Ts (Fig 3). However, it is important to note that our experiments could only assess potential N2 fixation and AO. Therefore, the absolute values and sign of the imbalance may not be accurate. In fact, it is likely that actual AO rates are somewhat overestimated and N2fixation rates somewhat underestimated in our assays [33] because of artificially increased diffusion of oxygen in the AO assay, and artificially increased diffusion of acetylene/ethane in the ARA assay. Thus, our results should be used for assessment of relative changes. Despite this limitation, we see a strong and consistent pattern of AO becoming relatively more important at T above 15–20°C, as ARA (indicating N inputs) stay constant or decline. Because the denitrification rates in the two deserts are at least an order of magnitude smaller than N2 fixation [14], we therefore predict that as drylands warm, the relative proportion of soil N will shift away from ammonium and towards nitrate, which is more easily lost via leaching, resulting in less overall N.

thumbnail
Fig 3. Trend of the balance between Nitrogen fixation (acetylene reduction) and AO rates, as the net difference for each BSC type an origin, according to the legend inserted.

Columns show the mean of n = 40 determinations (differences between 8 ARA value and 5 AO value), and error bars depict standard errors. Different lowercase letters indicated significant differences (p < 0.05) for comparisons among temperatures within a single BSC type.

https://doi.org/10.1371/journal.pone.0164932.g003

While experiments examining warming effects on soils crusts do exist, none have focused directly on gauging effects on the N cycle. A small increase of 2°C did not seem to affect soil chemistry in the short term [55, 56]. In another study where warming treatments were applied in the field, the warming treatment decreased levels of available N [36], although it is not known whether this was a result of the concomitant decline in BSC cover or because AO increased during times of declining N2 fixation, as we observe.

Conclusions and Outlook

As global temperatures rise in the future, many ecosystem processes will likely be altered. Our findings suggest that N cycles may be one process that is significantly affected. Many studies have documented that as higher temperatures become more common, N inputs via fixation are likely to decrease. Our results show that in addition to N inputs declining with rising temperatures, AO rates will likely continue to rise, thus decoupling the input and transformation processes. If our laboratory experiments can be extrapolated to the field, we predict this decoupling will result in a higher nitrate to ammonium ratio in dryland soils. Because nitrate is more easily lost both through leaching and as a gas than ammonium, we also expect an overall reduction in soil available N. As most desert soils are already N-limited, and cover ~35% of the Earth’s surface, the decoupling observed in this study could have large implications for soil fertility and thus plant productivity at the global scale. For example, a recent study by Poulter et al.[57] in Nature found that ~60% of global C sink anomalies were driven by semi-arid vegetation growth in Australia alone. Because of the relevance of BSCs to N budgets across all scales [8, 12], and the subsequent effect of N limitation on C cycling [58], we need long-term field experimentation to test the validity of our prediction. This is especially important in the face of the multiplicity of factors that may affect the N balance under natural conditions (e.g., soil moisture, CO2), many of which will likely be affected by future warming.

Supporting Information

S1 Fig. Linear regression between ln(N2 fixation rate) (A) and ln (AO rate) (B) with 1/T based on the Arrenius equation.

GB-Dark: Dark BSCs of the Great Basin; GB-Light: Light BSCs of the Great Basin; CH-Lichen: Lichen BSCs of the Chihuahuan Desert; CH-Light: Light BSCs of the Chihuahuan Desert. Only the portion of dataset where the rates increase with T was used.

https://doi.org/10.1371/journal.pone.0164932.s001

(DOCX)

Acknowledgments

We thank Sergio Velasco Ayuso at ASU and Anita Antoninka at Northern Arizona University for their help with soil sample collection. Thanks to Yevgeniy Marusenko (University of California, Irvine) for experimental suggestions. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

Author Contributions

  1. Conceptualization: XBZ FGP.
  2. Data curation: XBZ AGS.
  3. Formal analysis: XBZ HS.
  4. Funding acquisition: FGP.
  5. Investigation: XBZ HS AGS.
  6. Methodology: XBZ JB FGP.
  7. Project administration: XBZ FGP.
  8. Resources: FGP JB.
  9. Software: XBZ.
  10. Supervision: FGP.
  11. Validation: XBZ AGS.
  12. Writing – original draft: XBZ FGP.
  13. Writing – review & editing: AGS JB FGP.

References

  1. 1. Garcia-Pichel F. Desert Environments: Biological Soil Crusts. In: Bitton G, editor. Encyclopedia of Environmental Microbiology. New York: John Wiley; 2002. p. 1019–23.
  2. 2. Soule T, Anderson IJ, Johnson SL, Bates ST, Garcia-Pichel F. Archaeal populations in biological soil crusts from arid lands in North America. Soil Biol Biochem. 2009;41:2069–74.
  3. 3. Bates ST, Nash TH III, Garcia-Pichel F. Patterns of diversity for fungal assemblages of biological soil crusts from the southwestern United States. Mycologia. 2012;104:353–61. pmid:22123652
  4. 4. Marusenko Y, Bates ST, Anderson I, Johnson S, Soule T, Garcia-Pichel F. Ammonia-oxidizing archaea and bacteria are structured by geography in biological soil crusts across North American arid lands. Ecological Processes. 2013;2:9.
  5. 5. Belnap J, Prasse R, Harper K. Influence of biological soil crusts on soil environments and vascular plants. Biological Soil Crusts: Structure, Function, and Management: Springer; 2003. p. 281–300.
  6. 6. Garcia-Pichel F, Johnson SL, Youngkin D, Belnap J. Small-scale vertical distribution of bacterial biomass and diversity in biological soil crusts from arid lands in the Colorado Plateau. Microb Ecol. 2003;46:312–21. pmid:14502414
  7. 7. Beraldi-Campesi H. Early life on Earth and the first terrestrial ecosystems. Boletin De La Sociedad Geologica Mexicana. 2014;66:65–83.
  8. 8. Elbert W, Weber B, Burrows S, Steinkamp J, Budel B, Andreae MO, et al. Contribution of cryptogamic covers to the global cycles of carbon and nitrogen. Nat Geosci. 2012;5:459–62.
  9. 9. Belnap J, Gillette DA. Disturbance of biological soil crusts: Impacts on potential wind erodibility of sandy desert soils in southeastern Utah. Land Degradation & Development. 1997;8:355–62.
  10. 10. Zhang YM, Wang HL, Wang XQ, Yang WK, Zhang DY. The microstructure of microbiotic crust and its influence on wind erosion for a sandy soil surface in the Gurbantunggut Desert of Northwestern China. Geoderma. 2006;132:441–9.
  11. 11. Barger NN, Herrick JE, Van Zee J, Belnap J. Impacts of biological soil crust disturbance and composition on C and N loss from water erosion. Biogeochemistry. 2006;77:247–63.
  12. 12. Belnap J. Nitrogen fixation in biological soil crusts from southeast Utah, USA. Biol Fertil Soils. 2002;35:128–35.
  13. 13. Garcia-Pichel F, Belnap J. Microenvironments and microscale productivity of cyanobacterial desert crusts. J Phycol. 1996;32:774–82.
  14. 14. Strauss SL, Day TA, Garcia-Pichel F. Nitrogen cycling in desert biological soil crusts across biogeographic regions in the Southwestern United States. Biogeochemistry. 2012;108:171–82.
  15. 15. Su YG, Zhao X, Li AX, Li XR, Huang G. Nitrogen fixation in biological soil crusts from the Tengger desert, northern China. Eur J Soil Biol. 2011;47:182–7.
  16. 16. Beraldi-Campesi H, Hartnett HE, Anbar A, Gordon GW, Garcia-Pichel F. Effect of biological soil crusts on soil elemental concentrations: implications for biogeochemistry and as traceable biosignatures of ancient life on land. Geobiology. 2009;7:348–59. pmid:19573165
  17. 17. Langhans TM, Storm C, Schwabe A. Biological soil crusts and their microenvironment: Impact on emergence, survival and establishment of seedlings. Flora. 2009;204:157–68.
  18. 18. Li XR, Wang XP, Li T, Zhang JG. Microbiotic soil crust and its effect on vegetation and habitat on artificially stabilized desert dunes in Tengger Desert, North China. Biol Fertil Soils. 2002;35:147–54.
  19. 19. Belnap J. The potential roles of biological soil crusts in dryland hydrologic cycles. Hydrol Process. 2006;20:3159–78.
  20. 20. Li XR, Tian F, Jia RL, Zhang ZS, Liu LC. Do biological soil crusts determine vegetation changes in sandy deserts? Implications for managing artificial vegetation. Hydrol Process. 2010;24:3621–30.
  21. 21. Zhang J, Zhang YM, Downing A, Cheng JH, Zhou XB, Zhang BC. The influence of biological soil crusts on dew deposition in Gurbantunggut Desert, Northwestern China. J Hydrol. 2009;379:220–8.
  22. 22. Yeager CM, Kornosky JL, Morgan RE, Cain EC, Garcia-Pichel F, Housman DC, et al. Three distinct clades of cultured heterocystous cyanobacteria constitute the dominant N(2)-fixing members of biological soil crusts of the Colorado Plateau, USA. FEMS Microbiol Ecol. 2007;60:85–97. pmid:17381526
  23. 23. Pepe-Ranney C, Koechli C, Potrafka R, Andam C, Eggleston E, Garcia-Pichel F, et al. Non-cyanobacterial diazotrophs dominate dinitrogen fixation in biological soil crusts during early crust formation. The ISME journal. 2016;10:287–98. pmid:26114889
  24. 24. Evans RD, Belnap J. Long-term consequences of disturbance on nitrogen dynamics in an arid ecosystem. Ecology. 1999;80:150–60.
  25. 25. Wu N, Zhang YM, Downing A. Comparative study of nitrogenase activity in different types of biological soil crusts in the Gurbantunggut Desert, Northwestern China. J Arid Environ. 2009;73:828–33.
  26. 26. Barger NN, Castle SC, Dean GN. Denitrification from nitrogen-fixing biologically crusted soils in a cool desert environment, southeast Utah, USA. Ecological Processes. 2013;2:1–9.
  27. 27. Zhao Y, Xu M, Belnap J. Potential nitrogen fixation activity of different aged biological soil crusts from rehabilitated grasslands of the hilly Loess Plateau, China. J Arid Environ. 2010;74:1186–91.
  28. 28. Marusenko Y, Garcia-Pichel F, Hall SJ. Ammonia-oxidizing archaea respond positively to inorganic nitrogen addition in desert soils. FEMS Microbiol Ecol. 2015;91:1–11.
  29. 29. Bowen JL, Kearns PJ, Holcomb M, Ward BB. Acidification alters the composition of ammonia-oxidizing microbial assemblages in marine mesocosms. Marine Ecology Progress Series. 2013;492:1–8.
  30. 30. He J, Shen J, Zhang L, Zhu Y, Zheng Y, Xu M, et al. Quantitative analyses of the abundance and composition of ammonia-oxidizing bacteria and ammonia-oxidizing archaea of a Chinese upland red soil under long-term fertilization practices. Environ Microbiol. 2007;9:2364–74. pmid:17686032
  31. 31. Sher Y, Zaady E, Nejidat A. Spatial and temporal diversity and abundance of ammonia oxidizers in semi-arid and arid soils: indications for a differential seasonal effect on archaeal and bacterial ammonia oxidizers. FEMS Microbiol Ecol. 2013;86:544–56. pmid:23855990
  32. 32. Zhang X, Liu W, Schloter M, Zhang G, Chen Q, Huang J, et al. Response of the abundance of key soil microbial nitrogen-cycling genes to multi-factorial global changes. Plos One. 2013;8. pmid:24124568
  33. 33. Johnson SL, Budinoff CR, Belnap J, Garcia-Pichel F. Relevance of ammonium oxidation within biological soil crust communities. Environ Microbiol. 2005;7:1–12. pmid:15643930
  34. 34. Johnson SL, Neuer S, Garcia-Pichel F. Export of nitrogenous compounds due to incomplete cycling within biological soil crusts of arid lands. Environ Microbiol. 2007;9:680–9. pmid:17298368
  35. 35. Marusenko Y. Niche differentiation of ammonia-oxidizing microbial communities in arid land soils: Arizona State University; 2013.
  36. 36. Delgado-Baquerizo M, Maestre FT, Gallardo A. Biological soil crusts increase the resistance of soil nitrogen dynamics to changes in temperatures in a semi-arid ecosystem. Plant Soil. 2013;366:35–47.
  37. 37. Garcia-Pichel F, Loza V, Marusenko Y, Mateo P, Potrafka RM. Temperature drives the continental-scale distribution of key microbes in topsoil communities. Science. 2013;340:1574–7. pmid:23812714
  38. 38. Rajeev L, da Rocha UN, Klitgord N, Luning EG, Fortney J, Axen SD, et al. Dynamic cyanobacterial response to hydration and dehydration in a desert biological soil crust. ISME J. 2013;7:2178–91. pmid:23739051
  39. 39. Garcia-Pichel F, Wojciechowski MF. The evolution of a capacity to build supra-cellular ropes enabled filamentous cyanobacteria to colonize highly erodible substrates. Plos One. 2009;4. pmid:19924246
  40. 40. Bates ST, Nash TH Iii, Sweat KG, Garcia-Pichel F. Fungal communities of lichen-dominated biological soil crusts: Diversity, relative microbial biomass, and their relationship to disturbance and crust cover. J Arid Environ. 2010;74:1192–9. http://dx.doi.org/10.1016/j.jaridenv.2010.05.033
  41. 41. Garcia-Pichel F, Castenholz RW. Characterization and biological implications of scytonemin, a cyanobacterial sheath pigment. J Phycol. 1991;27:395–409.
  42. 42. Brankatschk R, Fischer T, Veste M, Zeyer J. Succession of N cycling processes in biological soil crusts on a Central European inland dune. FEMS Microbiol Ecol. 2013;83:149–60. pmid:22816620
  43. 43. Nübel U, Garcia-Pichel F, Muyzer G. PCR primers to amplify 16S rRNA genes from cyanobacteria. Applied and environmental microbiology. 1997;63:3327–32. pmid:9251225
  44. 44. Zhang Z, Schwartz S, Wagner L, Miller W. A greedy algorithm for aligning DNA sequences. Journal of Computational biology. 2000;7:203–14. pmid:10890397
  45. 45. Couradeau E, Karaoz U, Lim HC, da Rocha UN, Northen T, Brodie E, et al. Bacteria increase arid-land soil surface temperature through the production of sunscreens. Nature communications. 2016;7.
  46. 46. Belnap J, Lang OL. Biological soil crusts: structure, function, and management. Germany, Berlin: Springer-Verlag; 2003.
  47. 47. Brooks SJ, Collins JJ, Brill WJ. Repression of nitrogen fixation in Klebsiella pneumoniae at high temperature. Journal of bacteriology. 1984;157:460–4. pmid:6363385
  48. 48. Hennecke H, Shanmugam K. Temperature control of nitrogen fixation in Klebsiella pneumoniae. Archives of microbiology. 1979;123:259–65. pmid:393199
  49. 49. Zhu JB, Brill WJ. Temperature sensitivity of the regulation of nitrogenase synthesis by Klebsiella pneumoniae. Journal of bacteriology. 1981;145:1116–8. pmid:7007342
  50. 50. Lange OL, Belnap J, Reichenberger H. Photosynthesis of the cyanobacterial soil-crust lichen Collema tenax from arid lands in southern Utah, USA: Role of water content on light and temperature responses of CO2 exchange. Functional Ecology. 1998;12:195–202.
  51. 51. Grote EE, Belnap J, Housman DC, Sparks JP. Carbon exchange in biological soil crust communities under differential temperatures and soil water contents: implications for global change. Glob Change Biol. 2010;16:2763–74.
  52. 52. Zielke M, Ekker AS, Olsen RA, Spjelkavik S, Solheim B. The influence of abiotic factors on biological nitrogen fixation in different types of vegetation in the High Arctic, Svalbard. Arct Antarct Alp Res. 2002;34:293–9.
  53. 53. Reyes BA, Pendergast JS, Yamazaki S. Mammalian peripheral circadian oscillators are temperature compensated. Journal of biological rhythms. 2008;23:95. pmid:18258762
  54. 54. Layzell D, Rochman P, Canvin D. Low root temperatures and nitrogenase activity in soybean. Canadian Journal of Botany. 1984;62:965–71.
  55. 55. Johnson SL, Kuske CR, Carney TD, Housman DC, Gallegos-Graves L, Belnap J. Increased temperature and altered summer precipitation have differential effects on biological soil crusts in a dryland ecosystem. Glob Change Biol. 2012;18:2583–93.
  56. 56. Zelikova TJ, Housman DC, Grote EE, Neher DA, Belnap J. Warming and increased precipitation frequency on the Colorado Plateau: implications for biological soil crusts and soil processes. Plant Soil. 2012;355:265–82.
  57. 57. Poulter B, Frank D, Ciais P, Myneni RB, Andela N, Bi J, et al. Contribution of semi-arid ecosystems to interannual variability of the global carbon cycle. Nature. 2014;509:600–3. pmid:24847888
  58. 58. Gallardo A, Schlesinger WH. Carbon and nitrogen limitations of soil microbial biomass in desert ecosystems. Biogeochemistry. 1992;18:1–17.