Magnitude and Pathways of Increased Nitrous Oxide Emissions from

Jul 9, 2018 - ... fluxes following upland thermokarst formation along a 20-year thaw sequence within a thermo-erosion gully in a Tibetan swamp meadow...
0 downloads 0 Views 2MB Size
Download PDF
Article Cite This: Environ. Sci. Technol. 2018, 52, 9162−9169

pubs.acs.org/est

Magnitude and Pathways of Increased Nitrous Oxide Emissions from Uplands Following Permafrost Thaw Guibiao Yang,†,‡ Yunfeng Peng,† Maija E. Marushchak,§ Yongliang Chen,† Guanqin Wang,†,‡ Fei Li,†,‡ Dianye Zhang,†,‡ Jun Wang,†,‡ Jianchun Yu,†,‡ Li Liu,†,‡ Shuqi Qin,†,‡ Dan Kou,†,‡ and Yuanhe Yang*,†,‡

Downloaded via UNIV OF LOUISIANA AT LAFAYETTE on January 17, 2019 at 11:34:11 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

State Key Laboratory of Vegetation and Environmental Change, Institute of Botany, Chinese Academy of Sciences, Beijing 100093, China ‡ University of Chinese Academy of Sciences, Beijing 100049, China § Department of Environmental and Biological Sciences, University of Eastern Finland, Kuopio 70211, Finland S Supporting Information *

ABSTRACT: Permafrost thawing may release nitrous oxide (N2O) due to large N storage in cold environments. However, N2O emissions from permafrost regions have received little attention to date, particularly with respect to the underlying microbial mechanisms. We examined the magnitude of N2O fluxes following upland thermokarst formation along a 20-year thaw sequence within a thermo-erosion gully in a Tibetan swamp meadow. We also determined the importance of environmental factors and the related microbial functional gene abundance. Our results showed that permafrost thawing led to a mass release of N2O in recently collapsed sites (3 years ago), particularly in exposed soil patches, which presented post-thaw emission rates equivalent to those from agricultural and tropical soils. In addition to abiotic factors, soil microorganisms exerted significant effects on the variability in the N2O emissions along the thaw sequence and between vegetated and exposed patches. Overall, our results demonstrate that upland thermokarst formation can lead to enhanced N2O emissions, and that the global warming potential (GWP) of N2O at the thermokarst sites can reach 60% of the GWP of CH4 (vs ∼6% in control sites), highlighting the potentially strong noncarbon (C) feedback to climate warming in permafrost regions.

■

landforms,24 may increase N2O emissions due to the unlocking of vast N stocks stored in frozen soils. However, current experimental evidences are mainly derived from lowland permafrost ecosystems,14,17 and limited observations are available for upland ecosystems. Upland permafrost landscapes prone to thermokarst formation cover as much as 0.91 × 106 km2, or ∼5% of the permafrost regions24 and belong to an important category of permafrost thawing around the world. More importantly, compared with lowlands which tend to be waterlogged following thermokarst formation,25 thermokarst subsidence in upland permafrost usually results in intermediate or even lowered soil moisture26,27 and the formation of adjacent aerobic and anaerobic soil environments.8 These intermediate soil moisture conditions may promote coupled nitrification (oxidation of ammonia to NO3−) and denitrification (anaerobic respiration using NO3− as an electron

INTRODUCTION Permafrost regions account for a large proportion of the global carbon (C) reservoirs, due to the slow microbial decomposition of soil organic matter under cold conditions.1−3 The large amount of organic C stored in permafrost soils may represent increasingly relevant sources of methane (CH4) and carbon dioxide (CO2) to the atmosphere under global warming, which is more pronounced at high latitudes than the global average.4−7 The effects of permafrost thaw on gaseous C release have been extensively studied to evaluate the magnitude of permafrost C climate feedback.4,7−12 However, the large amount of nitrogen (N) stored in permafrost soils, which is a potential source for another greenhouse gas, nitrous oxide (N2O),8,13−21 is often ignored. Understanding N2O emissions from thawing permafrost is particularly critical because it has a high global warming potential (GWP), which is ∼300-fold greater per unit mass than that of CO2 over a 100year horizon,22 and N2O is also the most important substance leading to ozone depletion in the stratosphere.23 Permafrost thaw, leading to either gradual active layer deepening or often abrupt development of thermokarst © 2018 American Chemical Society

Received: Revised: Accepted: Published: 9162

April 28, 2018 July 5, 2018 July 9, 2018 July 9, 2018 DOI: 10.1021/acs.est.8b02271 Environ. Sci. Technol. 2018, 52, 9162−9169

Article

Environmental Science & Technology

We measured the N2O flux along a thermo-erosion gully during two summers in 2015 and 2016. The gully occurred in the early 1990s according to local residents and was ∼240 m long and approximately 2 m at the maximum depth at the time of measurements. In May 2015, four sites with an area of 15 × 20 m2 were established outside and inside the gully. The control site was an undisturbed area outside the thermoerosion gully. The other three sites were dispersed at intervals of ∼80 m from the head to the end along the gully, representing different thawing stages (early stage: collapse occurred 3 years ago; midstage: collapse occurred 12 years ago; and late stage: collapse occurred 20 years ago). N2O Flux Measurements. Fluxes of N2O were determined using static chamber and gas chromatography techniques.34 In 2015, ten vegetated patches were randomly chosen at the control site and at each of the three sites within the gully. A PVC ring (26-cm diameter, 12-cm high) with a collar was inserted into the soil to a depth of 10 cm. Because thermokarst formation often creates a mosaic of vegetated and exposed patches8 with potentially different N2O emission rate,13,17 we also measured the N2O flux in exposed patches inside the gully in 2016. The N2O flux was measured from 5 positions in the exposed patches in each site due to their smaller proportion of the feature area. During gas collection, an opaque cylindrical chamber (25 cm tall) was placed on the collar. The chamber was covered with insulation materials to avoid a rapid increase in chamber temperature and equipped with an electric fan for proper mixing of the headspace air. Gas samples (50 mL) were collected from the headspace at 10-min intervals (0, 10, 20, 30, and 40 min) and immediately injected into pre-evacuated air bags (Delin Inc., Dalian, Liaoning, China). Gas samples were then transported to the laboratory and analyzed for N2O concentration using a gas chromatograph (Agilent 7890A, Agilent Technologies Inc., Santa Clara, CA; for analytical methodologies, see Zheng et al.35). The rate of N2O flux was calculated from the slope of the N2O concentration−time function. Gas samples were collected between 9:00 am and 12:00 pm; in 2015, ca. every 10 days in July and August, and in 2016 once in early July, thrice during consecutive days at the end of July, once at the end of August. In total, 2275 gas samples were analyzed to obtain the N2O fluxes. At the time of gas collection, the soil temperature and volumetric water content (VWC) at a 0−10 cm soil depth were measured in each patch adjacent to the PVC ring using a digital thermometer and a portable TDR-100 soil moisture probe (Spectrum Technologies Inc., Plainfield, IL). Soil Sampling and Biochemical Analyses. To discern the mechanism underlying the thermokarst effects on N2O emissions, we collected soil samples from the four sites and measured the soil physicochemical characteristics and functional gene abundance of denitrifying microbes. Six soil cores (2.5 cm diameter and 15 cm deep) were collected near each PVC ring and homogenized to gain one composite sample during the intensive flux measurement campaign at the end of July 2016. In total, 240 soil cores were collected (40 soil samples). The soil samples were passed through a 2-mm-mesh sieve after removal of roots and divided into three parts. The first part was stored at 4 °C to determine the soil nitrate (NO3−) concentration. The second part was air-dried to measure the soil pH, and the third part was stored at −80 °C to quantify the key microbial functional gene abundances. The soil NO3− concentration was determined using a flow injection analyzer (Autoanalyzer 3 SEAL; Bran and Luebbe, Norder-

acceptor), which may lead to high N2O production and emissions.15,28 Moreover, disturbances of the vegetation cover may also promote N2O emissions via improved nutrient availability for soil microbes in the absence of plant uptake.15,17 In support of these deductions, elevated N2O concentrations in the soil have been observed in upland thermokarst features.8 However, direct field evidence of post-thaw N2O emissions in upland thermokarst remains unavailable. The escape of N2O from soils to the atmosphere after permafrost thaw is the result of the balance of multistep processes related to nitrification and denitrification,29 and these processes are primarily driven by microbial pathways.30 Multiple specialized microbial groups are responsible for nitrification-mediated (ammonia oxidizing archaea and bacteria) and denitrification-mediated pathways (denitrifying microorganisms). Further, the denitrification pathway is dependent on nitrification for the supply of nitrate, the initial electron acceptor for denitrification. Soil abiotic properties indirectly influence N2O flux via regulating microbial abundance and their activities.31 Although the importance of plant and abiotic factors in the regulation of N2O fluxes has been widely examined in permafrost areas (e.g., absence of plant N uptake, favorable moisture content and pH, low soil C:N ratios, and relatively high inorganic N concentrations),13−15,17 investigations of the microbial communities directly associated with N2O production under permafrost thaw remain lacking.19 The Tibetan permafrost region, covering 67%, or ∼1 500 000 km2, of the plateau area,32 has experienced significant climate warming and consequential permafrost thawing over the past decades. The alpine swamp meadows, which contain a high soil N density similar to that in the northern circumpolar permafrost region (Supporting Information Figure S1), offer an opportunity to examine the changes in N2O emissions following permafrost collapse together with the underlying mechanisms. Here, we measured the in situ N2O flux along a thermo-erosion gully in an alpine swamp meadow underlain by permafrost. The abiotic factors and microbial functional gene abundance were determined to gain a mechanistic understanding of how thermokarst formation affects N2O emissions. We hypothesize that (1) upland thermokarst formation enhances N2O emissions because of the mobilization of permafrost N and reduced plant N uptake, co-occurring with favorable moisture content, and that (2) thaw-induced changes in N2O flux are controlled by interactions between abiotic factors and soil microorganisms.

■

MATERIALS AND METHODS Experimental Site. The study site is an alpine swamp meadow underlain by discontinuous permafrost, located on the northeastern Tibetan Plateau, China (N 37°28′, E 100°17′, ∼3900 m a.s.l; Figure S2).33 The region is characterized by a wet alpine climate with a mean annual air temperature of −3.3 °C and a mean annual precipitation of 460 mm. The active layer, which is affected by seasonal thawing, is ca. 90 cm thick as measured by a GSSI georadar (SIR-20, Laurel, Santa Clara, CA) and validated by manual measurements by a thaw probe. Vegetation at the site is mainly dominated by graminoids, such as Kobresia tibetica, K. royleana, and Carex atrofuscoides. The soil in this area has a silty loam texture with 2.0% clay, 53.7% silt, and 44.3% sand. The chemical properties of the 0−15-cm soil layer are as follows: soil organic C, 19.4%; total N, 1.5%; bulk density, 0.27 g/cm3; and pH, 5.6. 9163

DOI: 10.1021/acs.est.8b02271 Environ. Sci. Technol. 2018, 52, 9162−9169

Article

Environmental Science & Technology stedt, Germany) after extracting fresh soil with 2 M KCl at a soil to solution ratio of 1:5 (v/v). Soil pH was measured in a soil water mixture (dry soil to deionized water ratio of 1:5 (v/ v)) with a pH electrode (PB-10, Sartorius, Germany). For the soil microbial functional gene quantification, the present study only focused on denitrifying microorganisms because the soil water pore space (WFPS) was between 62.6% and 84.1% across treatments and growing seasons. Under this condition, microbial denitrification is considered the predominant pathway of N2O production.36 Here, we quantified the abundance of genes that encoded nitrite reductase (nirK and nirS) and nitrous oxide reductase (nosZ). For the analysis, DNA was extracted from wet soil (0.4 g) using the Power Soil DNA Isolation Kit (MoBio Laboratories, Carlsbad, CA) according to the manufacturers’ instructions. The DNA quality was checked with a NanoDrop-2000 (Thermo Fisher Scientific, Madison, WI), which was assessed based on the absorbance values at 260/280 nm and 260/230 nm. A quantitative real-time polymerase chain reaction (qPCR) analysis was performed in triplicate on a StepOne Plus RealTime PCR system (Applied Biosystems, Inc., CA) in 20 μL of reaction mixture containing 10 μL of SYBR premix Ex Taq (Tli RNaseH Plus), 0.4 μL of forward and reverse primers, 0.4 μL of ROX Reference Dye (50×), 6.8 μL of DNA-free water, and 2 μL of 5-fold diluted template DNA. The amplification conditions, reading temperature, primer pairs, number of cycles, and references used to quantify the abundance of the nirK, nirS, and nosZ genes are summarized in Table S1. For quantification, standard curves were constructed with a known concentration of purified plasmid DNA carrying the cloned portion of the corresponding gene with r2 > 0.95. The abundances of the nirK, nirS, and nosZ genes are presented as a copy number per gram of dry weight (copies g−1 dw). Statistical Analysis. All data were checked for normality before the analysis. The N2O flux data were log-transformed to meet the assumption of a normal distribution. To assess the effects of thermokarst formation on N2O fluxes, a repeatedmeasures ANOVA was used to compare the differences in the N2O flux among the thawing stages and between vegetated and exposed patches, with the thawing stage or patch type as the between-subject factor, and the sampling date as the withinsubject factor. One-way ANOVA was employed to evaluate the differences in seasonal means of the N2O flux, soil temperature, and VWC as well as the soil pH, NO3− concentration, and the microbial functional gene abundance associated with denitrification (nirK, nirS, nosZ) among the thawing stages and between vegetated and exposed patches. We used post hoc tests (Tukey’s HSD) to test whether the above-mentioned parameters exhibited significance at the level of P < 0.05. Because of nonsignificant differences in N2O fluxes between patch types at the mid- and late stages of permafrost collapse (Figure 1b, Figure S3), we only examined the differences of biotic and abiotic parameters between the vegetated and exposed patches at the early stage. A linear regression was conducted to explore the relationships among N2O flux and the soil temperature, VWC, pH, abundance of the nirK, nirS, and nosZ genes, and the ratio of (nirK + nirS)/nosZ, which is usually used to represent the capacity for N2O production.37 In the regression analysis, the N2O flux, soil temperature, and VWC values were the means of the three consecutive measurements at the end of July in 2016. All of the above analyses were conducted using SPSS 20.0 (IBM SPSS, Chicago, IL).

Figure 1. Changes in N2O flux along the thaw sequence in 2015 (a) and 2016 (b) and between vegetated and exposed patches in 2016 (insert panel). Error bars represent the standard error (n = 10 for vegetated patches and n = 5 for exposed patches). Different letters on the bars represent a significant difference among treatments (P < 0.05). In the inset panel, * represents significant difference; ns, insignificant difference.

To determine how environmental factors (soil temperature, VWC, and pH), substrate availability (NO3− concentration), and denitrifying microorganisms individually and interactively affect changes in N2O flux under thermokarst formation, a variation partitioning analysis was performed to identify the N2O flux variation purely and jointly explained by the above three groups among the thawing stages and between patch types. A variation partitioning analysis is a quantitative statistical approach that is widely used in ecological studies to quantify the relative contributions of biotic and abiotic factors as well as their joint effect on many ecological processes.38−40 By using this method, the variation in dependent variables can be decomposed into independent components reflecting the relative importance of various groups of explanatory variables and their joint effects. For the microbial group, we only included the ratio of (nirK + nirS)/nosZ as an integrated index of denitrifiers to avoid collinearity in the analysis. Variation partitioning analysis was conducted using the “vegan” package for the R platform.41

■

RESULTS Changes in N2O Flux Following Thermokarst Formation. Permafrost collapse caused significant changes in N2O fluxes (Figure 1, Table S2). At the site where thermokarst formation occurred most recently (3 years ago), N2O flux increased significantly compared with that of the control site. 9164

DOI: 10.1021/acs.est.8b02271 Environ. Sci. Technol. 2018, 52, 9162−9169

Article

Environmental Science & Technology

whereas the NO3− concentration showed an opposite trend from the control to the fully thawed site, peaking at early thawing stage (Figure 2a−d). In the recently thawed site, the biogeochemical parameters in the exposed patches either significantly increased (NO3− concentration), decreased (soil temperature and pH), or remained unchanged (VWC) compared with those in the vegetated patches (Figure 3a−d). Linking N2O Fluxes to Biotic and Abiotic Factors under Thermokarst Formation. Both biotic and abiotic properties influenced the N2O fluxes under thermokarst (Table 1). Along the thaw sequence, the N2O fluxes exhibited positive correlations with NO3− concentration (r2 = 0.44, P < 0.001), copy number of the nirS gene (r2 = 0.21, P < 0.01), and ratio of (nirK + nirS)/nosZ (r2 = 0.49, P < 0.001), but negative correlations with the VWC (r2 = 0.57, P < 0.001), pH (r2 = 0.22, P < 0.01), and abundance of the nosZ gene (r2 = 0.38, P < 0.001). The relationship between N2O fluxes and soil temperature was nonsignificant (P = 0.89). Between patches in the recently formed thermokarst site, the N2O fluxes increased with the NO3− concentration (r2 = 0.46, P < 0.001) and the ratio of (nirK + nirS/nosZ (r2 = 0.64, P < 0.001), while it decreased with the soil temperature (r2 = 0.56, P < 0.001), pH (r2 = 0.57, P < 0.001), and the nosZ gene (r2 = 0.67, P < 0.001). A significant relationship was not observed between N2O fluxes and the abundance of the nirK gene (P = 0.095). The variation partitioning analysis showed that the total variation in N2O flux patterns was well explained by environmental conditions, substrate availability, and microbial gene abundance both along the thaw sequence (76.5%) or between vegetated and exposed patches (83.3%) (Figure 4). Among the different thawing stages, the environment was the primary contributor (65.5%, B+AB+BC+ABC), followed by the substrate (44.6%, C+AC+BC+ABC) and microorganisms (49.5%, A+AB+AC+ABC) (Figure 4a). Between the patch types, the total variance in the N2O flux was primarily related to microorganisms (67.5%, A+AB+AC+ABC), followed by the environment (70.1%, B+AB+BC+ABC) and substrate (45.9%, C+AC+BC+ABC) (Figure 4b). Strong overlaps occurred between microbial and abiotic factors, especially for the interaction of environment and microbe, which captured 44.7% (AB+ABC) and 55.5% (AB+ABC) of the total variances among the thawing stages (Figure 4a) and patch types stages (Figure 4b), respectively.

Specifically, in vegetated patches, the N2O emissions increased from 0.026 to 0.19 mg N2O m−2 d−1 following thermokarst formation. In exposed patches, permafrost collapse led to substantially elevated N2O emissions relative to the control (3.1 vs 0.026 mg N2O m−2 d−1). However, at the mid- and late stages of permafrost collapse (∼12 to ∼20 years after permafrost collapse), the N2O emissions dropped to the rate of the noncollapsed site. Effects of Thermokarst Formation on Biogeochemical Characteristics. The soil biogeochemical properties exhibited significant differences both among different thawing stages and between patch types (Figures 2-3). In the vegetated patches, the soil temperature at 0−10 cm gradually increased along the thaw gradient from 8.6 to 9.2 °C, while the changes in VWC were best depicted by the Covington curve, which is characterized by an initial decrease and a subsequent increase. Similar to the VWC, the soil pH also slightly declined at the early thawing stage and increased progressively afterward,

■

DISCUSSION Our field N2O measurements revealed that upland permafrost collapse substantially increased N2O emissions in the recently (3 years ago) collapsed site (Figure 1, Figure S3). Especially high N2O emissions were measured in the patches where the soil was exposed and lacking any vegetation: the value was 2 orders of magnitude higher than that at the noncollapse site (3.1 mg N2O m−2 d−1 as opposed to 0.026 mg N2O m−2 d−1). The emission rate at the exposed patches was comparable to that observed for the dry bare mesocosms in Voigt et al.17 (3.1 mg N2O m−2 d−1 as opposed to 2.8 mg N2O m−2 d−1), matching the in situ N2O flux from tropical forest soils (0.1− 2.5 mg N2O m−2 d−1), which is the largest terrestrial N2O source.42 Fluxes from vegetated patches were smaller, but also the N2O emissions from these areas also increased by more than 7-fold following thermokarst formation. These findings are consistent with our first hypothesis that upland permafrost collapse would cause increased N2O emissions. Nevertheless, our results conflicted with a recent study which reported

Figure 2. Changes in the soil temperature (a), VWC (b), pH (c), NO3− concentration (d), nirK (e), nirS (f), nosZ (g), and ratio of (nirK + nirS)/nosZ (h) in vegetated soil patches along the thaw sequence. Error bars represent the standard error (n = 10). Different letters on the bars represent significant differences among treatments (P < 0.05). ST, soil temperature at 10 cm depth; and VWC, volumetric water content at 10 cm depth. 9165

DOI: 10.1021/acs.est.8b02271 Environ. Sci. Technol. 2018, 52, 9162−9169

Article

Environmental Science & Technology

Figure 3. Comparison of the soil temperature (a), VWC (b), pH (c), NO3− concentration (d), nirK (e), nirS (f), nosZ (g), and ratio of (nirK + nirS)/nosZ (h) between the vegetated and exposed patches at the early collapsed site (thermokarst formation occurred 3 years ago). Error bars represent the standard error (n = 10 for vegetated patches and n = 5 for exposed patches). Different letters on the bars represent a significant difference among treatments (P < 0.05). ST, soil temperature at 10 cm depth; and VWC, volumetric water content at 10 cm depth.

Table 1. Relationships between the N2O Flux and Biotic and Abiotic Factors along the Thaw Sequence and between Vegetated and Exposed Patchesa N2O flux along the thaw sequence ST (°C) VWC (%) pH NO3− (mg kg−1) nirK (copies g−1 dw) nirS (copies g−1 dw) nosZ (copies g−1 dw) (nirK + nirS)/nosZ

N2O flux between vegetated and exposed patches

regression coefficient

2

r

P value

regression coefficient

r2

P value

−0.0018 −0.0029 −0.11 0.0037

0.00 0.57 0.22 0.44

0.89