Quantification of year-round methane and nitrous oxide fluxes in a typical alpine shrub meadow on the Qinghai-Tibetan Plateau

Agriculture, Ecosystems and Environment 255 (2018) 27–36

Contents lists available at ScienceDirect

Agriculture, Ecosystems and Environment journal homepage: www.elsevier.com/locate/agee

Quantification of year-round methane and nitrous oxide fluxes in a typical alpine shrub meadow on the Qinghai-Tibetan Plateau

T

⁎

Yongfeng Fua,b, Chunyan Liua, , Fei Lina,b, Xiaoxia Hua,b, Xunhua Zhenga,b, Wei Zhanga, Guangmin Caoc a

State Key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry (LAPC), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029, PR China b College of Earth Science, University of Chinese Academy of Sciences, Beijing 100049, PR China c Northwest Institute of Plateau Biology, Chinese Academy of Sciences, Xining 810008, PR China

A R T I C L E I N F O

A B S T R A C T

Keywords: Gas chromatography Carrier gas Nitrous oxide Spring thaw The Qinghai-Tibetan plateau

Alpine meadows are the largest grasslands in China. Greenhouse gas exchanges between alpine meadows and the atmosphere are highly uncertain due to the lack of year-round flux measurements. In this study, the methane (CH4) and nitrous oxide (N2O) fluxes in an alpine Potentilla fruticosa shrub meadow in the Qinghai-Tibetan Plateau were investigated using the static, opaque chamber-gas chromatography method between April 2012 and April 2015. The annual CH4 uptake (1.33–1.35 kg C ha−1 yr−1) and Q10 value (1.79) were at the low end of the range for natural grasslands in China, which indicated that global warming at the same extent would result in less of an increase in the CH4 sink in the Qinghai-Tibetan Plateau, compared to other grassland areas. The N2O emissions during the spring thaw period showed a tremendous inter-annual variation (0.03 to 0.14 kg N ha−1), which was closely linked to the variation in annual precipitation, especially the precipitation of the previous growing season. The high substrate concentrations and soil moisture during the spring thaw periods together provided the conditions for pulse N2O emissions. When the pulse N2O emissions occurred, emissions from the non-growing seasons dominated (67–74%) the annual total emissions. Thus, a proper sampling frequency (daily to weekly measurements) in non-growing seasons was needed for the quantification of annual fluxes. Four gas chromatographic set-ups (the Ar-CH4, N2-ascarite, N2-CO2, and pure N2 methods) were adopted for N2O flux measurements over natural grasslands in China. Using the N2-CO2 method, the annual N2O emissions from the alpine shrub meadow were quantified to be only 0.18–0.27 kg N ha−1 yr−1 in the present study. Based on measurements using the first three of the four gas chromatographic set-ups, the conclusion that can be drawn is that unfertilized natural grasslands in China function as marginally weak N2O sources, whereas the pure N2 method may remarkably overestimate the emissions.

1. Introduction Grasslands are the dominant ecosystems in China, accounting for approximately 40% of the national land area. As the largest grasslands in China, alpine meadows, with an area of approximately 64 million ha, are widespread on the Qinghai-Tibetan Plateau (Zhang et al., 2010). The Qinghai-Tibetan Plateau is the highest natural geographical unit in the world, with an average altitude of over 4000 m, and a total area of 2.5 million km2. Alpine meadows are one of the dominant ecosystem types on the Qinghai-Tibetan Plateau, covering 27% of the plateau area (Hu et al., 2010; Wei et al., 2015). The Kobresia (Kobresia humilis, Kobresia pygmaea and Kobresia capillifolia) and Potentilla fruticose are the typical zonal vegetation of alpine meadows. Alpine meadows support the development of animal husbandry, and are also extremely ⁎

important for carbon sequestration, water resources and biodiversity. Because of harsh climatic conditions, the alpine meadows are fragile and sensitive to climate change and human activities. Ongoing climate change, intensified activities of subterranean rodents (plateau zokor and pika), and overgrazing result in the degradation of alpine meadows. Degradation of the alpine meadows may have altered ecosystem communities and vegetation productivity, leading to the release of soil carbon and nitrogen (Zhang and Liu, 2003; Li et al., 2014), and intensifying the exchanges of greenhouse gases, such as methane (CH4) and nitrous oxide (N2O) (Zhang et al., 2014; Li et al., 2015; Zhao et al., 2017). The accurate quantification of greenhouse gas fluxes in alpine meadows is important to evaluate the effects of global change on ecosystems in the Qinghai-Tibetan Plateau, and their feedbacks to climate change.

Corresponding author. E-mail address: [email protected] (C. Liu).

https://doi.org/10.1016/j.agee.2017.12.003 Received 4 July 2017; Received in revised form 20 October 2017; Accepted 4 December 2017 0167-8809/ © 2017 Elsevier B.V. All rights reserved.

Agriculture, Ecosystems and Environment 255 (2018) 27–36

Y. Fu et al.

fluxes in the non-growing seasons to the annual totals; (c) accurately quantify the annual fluxes of both gases in the extensively distributed alpine shrub meadow of the Qinghai-Tibetan Plateau; and (d) assess the effects of improper measuring methods on the quantification of annual fluxes in natural grasslands.

Due to the cold climate in the alpine meadows of the QinghaiTibetan Plateau, measurements of the biosphere–atmosphere exchange of CH4 and N2O in the non-growing seasons (which last up to seven months per year) were either missing or carried out with very low measurement frequency. Previous intensive flux measurements have exclusively concentrated on the growing seasons (Jiang et al., 2010; Wei et al., 2012, 2015; Zhang et al., 2014; Zhu et al., 2015). Wang et al. (2014) and Wei et al. (2015) simply assumed a ∼30% contribution from the non-growing seasons to extrapolate the CH4 uptake in growing seasons to the annual cumulative uptake. Although Du et al. (2008) and Hu et al. (2010) carried out flux measurements in the non-growing seasons, the extremely low measurement frequencies (once or twice per month) did not allow capturing the short term flux dynamics during periods with freeze-thaw cycles (Holst et al., 2008). Due to the lack of observations in high temporal resolution during the non-growing seasons, large uncertainties may exist in the estimates of annual fluxes in the alpine meadows of the Qinghai-Tibetan Plateau. Amongst previous measurements of N2O fluxes with the static chamber method, different gas chromatographic set-ups have been used, yielding conflicting results. Previous measurements of N2O emissions from natural grasslands in China involved four different gas chromatographic set-ups: (I) the N2-CO2 method (Wei et al., 2012, 2014; Chen et al., 2013; Zhang et al., 2014), in which a pure nitrogen gas (N2) was used as a carrier gas, and a mixed gas of carbon dioxide (CO2) and N2 (10% CO2 in N2) was introduced directly into the electron capture detector (ECD) cell as a make-up gas; (II) the N2-ascarite method, in which N2 was used as the carrier gas while ascarite (coated sodium hydroxide particles) was added to the injection port to remove CO2 and water from the air samples (Holst et al., 2007, 2008; (III) the Ar-CH4 method (Pei et al., 2003; Jiang et al., 2010), in which a mixture gas of CH4 and argon gas (Ar) (with 5–10% CH4 in pure Ar) was used as a carrier gas; and (IV) the pure N2 method (Wang et al., 2005; Du et al., 2006, 2008; Lin et al., 2009; Li et al., 2012a, b), in which pure N2 was used as a carrier gas without any additional make-up gases or treatments for gas samples. For air samples with a stable N2O concentration but variable CO2 contents, a comparison study in the laboratory showed reliable and comparable N2O fluxes among the N2-ascarite and Ar-CH4 methods, whilst the pure N2 method significantly overestimated the N2O fluxes, with the bias magnitudes increasing with CO2 concentrations (Zheng et al., 2008). In a field comparison study conducted in croplands, the pure N2 method was found to significantly overestimate the low (< 200 μg N m−2 h−1) N2O fluxes compared to the N2-ascarite method (Zheng et al., 2008). Unfortunately, many of the N2O flux measurements in alpine grasslands have been conducted using the pure N2 method (Du et al., 2008; Lin et al., 2009; Li et al., 2012a, b), which might have resulted in a significant overestimate of the emissions. Both of these problems seriously hamper the accurate quantification of CH4 and N2O fluxes in alpine meadows of the Qinghai-Tibetan Plateau. Thus, there is a necessity to conduct year-round flux measurements in the alpine meadows using any of the N2-CO2, N2-ascarite, and Ar-CH4 methods, with frequencies high enough to capture the temporal dynamics of fluxes during the spring thaw periods. In this study, the CH4 and N2O fluxes in a typical alpine Potentilla fruticosa shrub meadow in the eastern Qinghai-Tibetan Plateau were investigated using the static chamber-gas chromatography method (the N2-CO2 method), with intensive field observations over 2.5 years. The Kobresia meadow and Potentilla fruticosa shrub meadow are the dominant types of alpine meadows in the Qinghai-Tibetan Plateau (Li et al., 2006; Yashiro et al., 2010). The former distributes on the sunny slope of the valley and the valley floor, whereas the latter locates in the shady slope of the valley and the floodplain. In spite of a relatively simple plant community, the Potentilla fruticosa shrub meadow has very high productivity, and thus becomes a fine pasture traditionally utilized for grazing (Wang et al., 1991). The aims of this study were to (a) characterize the year-round dynamics of the CH4 and N2O fluxes and the major drivers for the variations; (b) evaluate the importance of

2. Materials and methods 2.1. Experimental site The experimental site (37°38′27.720″N, 101°19′7.572″E; altitude: 3260 m) is located approximately 3 km north of the Haibei Alpine Meadow Ecosystem Research Station in Qinghai province, China. The experimental area (100 ha) extends along a river valley, and the landscape is therefore completely flat. The vegetation is composed of a shrub and herbaceous layer. The shrub coverage varies within 90–98%, with heights ranging from 24 to 64 cm. The shrub layer is dominated by Potentilla fruticosa. The plant species of the herbaceous layer mainly include Kobresia humilis, Elymus nutans, Ligularia sagitta (Maxim.) Maettf., Polygonum viviparum Linn. and Anaphalis lactea Maxim. The rotational grazing system (summer and winter-grazing pastures) has been applied since the 1960s. The Potentilla fruticosa shrub meadow has been utilized as a winter-grazing pasture. Therefore, grazing has been forbidden since then between June and August, but Tibetan sheep graze during the daytime from September to May of the following year. The stocking rate is approximately 5.4 sheep unit ha−1 during the grazing period, being equal to 4.0 sheep unit ha−1 yr−1. The soil is a Mol-Cryic Cambisol corresponding to a Gelic Cambisol with a pH of 8.3 ± 0.1 (0–10 cm, mean ± standard error) and a bulk density of 0.95 ± 0.07 (0–4 cm). The soil organic carbon, total nitrogen, and texture are listed in Table 1. The research area is subject to a continental monsoon climate, with a long cold winter and a short cool summer (Zhou et al., 2006). The growing season typically begins in early May and ends in late September (Li et al., 2004). Between 1980 and 2012, the annual precipitation amounted to 527.9 mm on average, of which about 80% fell in the growing seasons, while the annual average air temperature was −1.3 °C (Zhang et al., 2014). 2.2. Gas flux measurements The gas flux measurements were carried out during the period from 21 April 2012 to 21 April 2015, using the static opaque chamber-gas chromatography method (Zhang et al., 2014). Six spatial replicates (at least 30 m away from each other) were randomly selected for flux measurements. A chamber base frame made of stainless steel (length × width × height = 0.5 × 0.5 × 0.15 m) was permanently inserted into the soil at each spatial replicate. Prior to air sampling, a stainless steel chamber (length × width × height = 0.5 × 0.5 × 0.4 m), covered with thermal isolating styrofoam and radiation reflecting tinfoil to prevent dramatic temperature changes in the headspace during chamber closure, was mounted onto the fixed base frame. Rubber strips were used for gas-tight sealing between the base frame and the chamber. Table 1 The properties of the soil profiles in the experimental alpine shrub meadow. Soil depth (cm)

SOC (g kg−1)

TN

0–20 20–40 40–60 60–80 80–100

26.7 20.8 18.8 17.9 18.1

2.9 2.3 2.2 2.1 2.0

(0.7) (2.7) (1.8) (1.6) (1.1)

Sand (%) (0.04) (0.1) (0.2) (0.2) (0.1)

39.2 38.7 37.0 30.4 30.2

Silt

(1.1) (0.8) (1.7) (2.0) (0.7)

43.8 45.5 45.4 49.0 47.2

Clay

(1.0) (0.9) (1.4) (1.0) (0.5)

17.0 15.8 17.6 20.6 22.6

(0.5) (0.4) (0.5) (1.1) (1.0)

SOC: soil organic carbon content; TN: total nitrogen content; Sand, silt and clay: sand (0.02–2 mm), silt (0.002–0.02 mm) and clay (< 0.002 mm) contents. The values in parentheses indicate the standard error of five spatial replicates.

28

Agriculture, Ecosystems and Environment 255 (2018) 27–36

Y. Fu et al.

3. Results

A taller chamber (up to 0.8 m tall) was used in cases where the plant height was greater than 0.4 m. Each chamber was equipped with a digital platinum resistance thermometer (JM624, Jinming Instrument Co., Ltd., Tianjin, China) to monitor the headspace air temperature during chamber closure. A vent tube designed according to Hutchinson and Mosier (1981) was installed on the chamber wall to balance air pressure between the headspace and the ambient atmosphere. For this purpose, the vent tube was open during chamber installation and gas sampling, and it was closed during the remaining closure time. At each position, the gas fluxes were manually measured at frequencies of once every 1–7 days. The frequencies were set depending upon the variations and levels of N2O fluxes, with daily observations being applied during the spring thaw periods and following significant rainfall events, and weekly observations for the remaining observational period. On the day of observation, flux measurements at all spatial replicates were simultaneously implemented between 8:00 and 10:00 a.m., when the gas flux and air temperature were representative of the daily averages (Liu et al., 2010a, 2013). To determine a flux, five gas samples were collected at 20min intervals using plastic syringes. All gas samples were analyzed within eight hours using a gas chromatograph (Agilent 7890A, Santa Clara, CA, USA), which was equipped with a flame ionization detector and an ECD. A calibration gas (with 1.93 μmol mol−1 CH4 and 0.426 μmol mol−1 N2O in pure N2) was used to determine the concentrations of gas samples. The fluxes were determined by applying nonlinear and linear fits of gas concentrations as a function of sampling time, according to the method of Liu et al. (2010a, 2013). Nitrous oxide concentrations were detected by the N2-CO2 method as described by Wang et al. (2010). The detection limits of the CH4 and N2O fluxes were ± 2.5 μg C m−2 h−1 and ± 1.0 μg N m−2 h−1, respectively, for a chamber height of 0.4 m and a chamber closure time of 80 min. This corresponded to analysis precisions of 11.5 nmol mol−1 CH4 and 2.0 nmol mol−1 N2O at 95% confidence interval for ambient atmospheric concentrations. The N2-CO2 method has been extensively used in China, due to the low detection limits (suitable for more ecosystems with low gas exchange rates) and the low cost of using N2 and N2-CO2 as the carrier and make-up gases rather than the expensive Ar-CH4.

3.1. Environment The air and soil (5 cm) temperatures showed a similar seasonal pattern, but the daily averages of soil temperature had a narrower seasonal variation (−10.1 and 17.0 °C) compared with air temperatures (−20.2 and 15.5 °C, Fig. 1a). The annual averaged soil (5 cm) temperatures were 3.5–3.9 °C higher than the annual averaged air temperatures. The growing seasons normally started from the middle of April, when the daily averaged air temperatures were consistently higher than 0 °C, and ended in the beginning of October when the daily averaged air temperatures remained consistently below 5 °C. Therefore, the length of growing seasons varied between 161 and 171 days, and the non-growing seasons lasted as long as 194–202 days. The surface soil began to thaw from early March when the daily averages of air and soil temperatures were consistently above −10 and −3 °C, respectively. The spring thaw period continued 48–50 days (Table 2). The precipitation and soil moisture showed large seasonal and interannual variations (Fig. 1b). The total annual precipitation was 370.6, 485.9, and 594.9 mm during the 1st (from 21 April 2012 to 13 April 2013), 2nd (from 14 April 2013 to 21 April 2014), and 3rd (from 22 April 2014 to 21 April 2015; Table 2) year-round periods, respectively. The precipitation in growing seasons (337.2 to 507.5 mm) amounted to 80–90% of the total annual precipitation. The precipitation in nongrowing seasons fluctuated between 33.4 and 99.3 mm. The mean WFPS in the three growing seasons lightly increased from 62.3% to 66.7%. The gradual increase in precipitation through the growing seasons also steadily enhanced the soil moisture during the following non-growing seasons and spring thaw periods through the experimental period. The averages of soil WFPS in the 3rd non-growing season and spring thaw period were up to 77.6 and 86.8%. The remarkable increase in total annual precipitation from the 2nd to 3rd year-round period promoted the annual mean soil WFPS from 57.7 to 72.1% (Table 3). The soil NO3− concentrations were constantly low, and fluctuated between 0.01 and 5.3 mg N kg−1 dry soil. The soil NH4+ concentrations were always low in the growing seasons, but relatively high values (> 5 mg N kg−1 dry soil) were found during the spring thaw periods (Fig. 1c). In general, the annual averages of soil inorganic nitrogen concentration (NO3− + NH4+) were below 3 mg N kg−1 dry soil (Table 3).

2.3. Auxiliary measurements Air temperature (HMP155) and daily rainfall (RG13H with rain collector heater, VAISALA, Finland) were automatically recorded by the meteorological station at the Haibei Alpine Meadow Ecosystem Research Station, Chinese Academy of Sciences. The daily snowfall was additionally recorded by a manual rain gauge. The soil temperature at 5 cm depth was recorded at 30-min intervals by temperature loggers (StowAway TidbiT V2, Onset Computer Co., Irvine, CA, USA). The topsoil (0–6 cm) moisture was manually measured by the gravimetric method during freezing, and otherwise using a portable moisture probe (ML2x, ThetaKit, Delta-T Devices, Cambridge, UK). The soil moisture was normalized to water-filled pore space (WFPS, %). At all gas sampling dates, topsoil samples (0–10 cm depth) were taken and extracted for analysis of soil ammonium (NH4+) and nitrate (NO3−) concentrations (Zhang et al., 2014).

3.2. Characteristics of CH4 and N2O fluxes The CH4 fluxes were all negative, which indicated that the alpine meadow always functioned as a sink during the entire observation period (Fig. 2a). The CH4 uptake ranged from 1.7 to 38.6 μg C m−2 h−1 and showed a pronounced seasonal variation, being more intensive in the growing seasons compared to the non-growing seasons. The CH4 uptake in the non-growing seasons was normally below 20 μg C m−2 h−1. The cumulative uptake amounted to 0.84–0.96 and 0.39–0.46 kg C ha−1 in the growing and non-growing seasons, respectively. Although the CH4 uptake was relatively weak, the cumulative uptake during the long non-growing seasons (194–202 days) contributed to approximately one-third (29–35%) of the annual total uptake. The annual uptake ranged from 1.33 to 1.35 kg C ha−1 yr−1 and did not show a strong inter-annual variation during the observation period (Fig. 2b and Table 4). The N2O fluxes varied between −0.2 and 21.1 μg N m−2 h−1 during the whole experimental period (Fig. 2c). They were low, and close to the detection limit of the measurement system for most of the observational period. The highest emissions were detected during the 2nd and 3rd spring thaw periods. These emissions (> 10 μg N m−2 h−1) remained high for approximately one week and one month during the 2nd and 3rd freeze-thaw periods, respectively, and accounted for

2.4. Statistical analysis Nonlinear fits and stepwise regression analysis were applied to identify the key environmental regulators of the gas fluxes or the relationship between fluxes and environmental factors. The Van’t Hoff equation (y = a·ebx) was fitted to obtain the coefficient of soil temperature (5 cm in depth) sensitivity (Q10) of CH4 uptake. The significance of nonlinear regression was determined using an F-test. The software packages SPSS Statistics Client 16.0 (SPSS Inc.), Origin 8.0 (OriginLab Ltd.), and SigmaPlot V12 (Systat Software Inc.) were used for the statistical analysis. 29

Agriculture, Ecosystems and Environment 255 (2018) 27–36

Y. Fu et al.

Fig. 1. Daily averages of air (1.5 m above ground) and soil (5 cm in depth) temperatures (a), soil (0–6 cm in depth) moisture (water-filled pore space, WFPS) and daily precipitation (b), and soil (0–10 cm in depth) nitrate (NO3−) and ammonium (NH4+) concentrations (c) during the entire experiment period. The soil moisture and inorganic nitrogen data points in (b) and (c) are the means of four and three spatial replicates, respectively, with vertical bars showing the standard errors. GS: growing season. NGS: non-growing season. STP: spring thaw period.

significantly higher cumulative emissions (0.06 and 0.14 kg N ha−1, respectively). The cumulative N2O emissions in the growing seasons (0.06–0.10 kg N ha−1 over 161–171 days) were similar to the emissions during the short spring thaw periods (0.03–0.14 kg N ha−1 over 48–50 days). The cumulative emissions during the spring thaw periods accounted for about 33% and 52% of the annual total emissions in the 2nd and 3rd year-round periods, respectively. The annual total emissions were 0.18 ± 0.03 and 0.27 ± 0.01 kg N ha−1 yr−1, respectively, in the alpine shrub meadow (average ± standard error, Fig. 2d and Table 4).

Table 3 Averaged air temperature (Ta), soil temperature (Ts, 5 cm), water-filled pore space (WFPS, 0–6 cm), nitrate (NO3−), and ammonium (NH4+, 0–10 cm) concentrations, and total precipitation (P). Year

Period

1st

GS NGS

2nd

1st year

2nd year

3rd year

Starting date Ending date days

≥0 ≤5

2012–04–21 2012–09–28 161

2013–04–14 2013–10–01 171

2014–04–22 2014–10–09 171

NGS

Starting date Ending date days

<5 >0

2012–09–29 2013–04–13 197

2013–10–02 2014–04–21 202

2014–10–10 2015–04–21 194

STP

Starting date Ending date days

> −10 >0

2013–02–23 2013–04–13 50

2014–03–05 2014–04–21 48

2015–03–05 2015–04–21 48

GS

(mm)

(%)

(mg N kg−1 dry soil)

NH4+

62.3 (1.2) –

0.9 (0.2) –

2.2 (0.2) –

7.1

42.5 (2.8)

1.4 (0.4)

2.7 (1.1)

370.6

–

–

–

GS

8.1 (0.3)

386.6

62.7 (1.2)

0.9 (0.1)

2.0 (0.3)

NGS

−6.9 (0.4) −1.6 (0.5) −0.04 (0.5)

10.6 (0.3) −1.8 (0.3) 0.7 (0.3)

99.3

51.7 (2.6)

1.6 (0.3)

1.4 (0.2)

76.2

61.1 (4.2)

0.8 (0.1)

2.0 (0.7)

3.9 (0.4)

485.9

57.7 (1.5)

1.2 (0.2)

1.7 (0.2)

9.7 (0.3) −2.7 (0.3) 0.6 (0.2)

507.5 87.4

66.7 (2.1) 77.6 (2.4)

0.3 (0.1) 0.7 (0.1)

2.0 (0.2) 2.9 (0.9)

33.0

86.8 (1.9)

1.0 (0.2)

4.8 (1.7)

3.1 (0.4)

594.9

72.1 (1.8)

0.5 (0.1)

2.4 (0.4)

Annual 3rd

NO3−

337.2 33.4

STP

Table 2 Definition of growing season (GS), non-growing season (NGS), and spring thaw period (STP) by air temperature (Ta).

WFPS

9.8 (0.3) −2.0 (0.3) −0.04 (0.2) 3.2 (0.4)

Annual

Simple regression analysis showed that the CH4 uptake displayed a significantly (p < 0.01) exponential dependency on soil temperature (5 cm in depth). The dependencies of CH4 uptake on soil temperature

P

7.9 (0.3) −7.2 (0.4) −2.6 (0.3) −0.5 (0.5)

STP

Ta (°C)

Ts

(°C)

3.3. Effects of environmental factors on the fluxes

Period

Ta

GS NGS STP Annual

7.1 (0.3) −7.0 (0.4) −1.3 (0.5) −0.4 (0.4)

GS: growing season; NGS: non-growing season; STP: spring thaw period. The values in parentheses indicate the standard error.

30

Agriculture, Ecosystems and Environment 255 (2018) 27–36

Y. Fu et al.

Fig. 2. Time series of CH4 and N2O fluxes (a, c) and cumulative exchanges (b, d) during the entire experiment period. Individual data points are the means of four spatial replicates with the vertical bars showing the standard errors. GS: growing season. NGS: non-growing season. STP: spring thaw period.

N2O fluxes during the entire experimental period (r2 = 0.32, p < 0.01). The N2O emissions were stimulated (≥10 μg N m−2 h−1) by the relatively high soil WFPS (≥75%) and inorganic nitrogen concentrations (≥6 mg N kg−1 dry soil, Fig. 4b).

Table 4 Cumulative CH4 and N2O fluxes. CH4 (kg C ha−1)

N2O (kg N ha−1)

Year

1st

2nd

3rd

1st

2nd

3rd

GS

−0.84 (0.04) –

−0.87 (0.06) −0.46 (0.03) −0.11 (0.01) −1.33 (0.08)

−0.96 (0.08) −0.39 (0.02) −0.10 (0.01) −1.35 (0.10)

0.10 (0.02) –

0.06 (0.01) 0.12 (0.02) 0.06 (0.02) 0.18 (0.03)

0.07 (0.01) 0.20 (0.01) 0.14 (0.01) 0.27 (0.01)

NGS STP Annual

−0.11 (0.01) –

0.03 (0.004) –

4. Discussion 4.1. Effects of gas chromatographic set-ups on nitrous oxide flux measurements Year-round flux measurements were carried out in a few types of grassland in China using the static opaque chamber-gas chromatography method (Table 5). Among these measurements, four gas chromatographic set-ups were adopted: (I) pure N2 as a carrier gas (hereinafter referred to as the pure N2 method); (II) a mixture gas of Ar and CH4 as a carrier gas (the Ar-CH4 method); (III) ascarite as a CO2 trap for gas samples while using pure N2 as a carrier gas (the N2-ascarite method); and (IV) a mixed gas of CO2 and N2 (e.g. 10% CO2 in N2) as a make-up gas while using pure N2 as a carrier gas (the N2-CO2 method). Zheng et al. (2008) evaluated the effects of three of the four gas chromatographic set-ups (the pure N2, Ar-CH4, and N2-ascarite methods) on N2O flux measurements in upland croplands. They concluded that the pure N2 method significantly overestimated N2O emissions compared to the Ar-CH4 and N2-ascarite methods, due to the interference of CO2 on N2O detection by the ECD of the gas chromatograph. However, the degree to which N2O emissions were overestimated by the pure N2 method was highly dependent on the flux levels and ecosystem types. The effects of the pure N2 method on the measurements of N2O fluxes in grassland ecosystems have not been evaluated so far. Wang et al. (2010) recommended a new gas chromatographic set-up for the accurate detection of N2O concentrations by

GS: growing season; NGS: non-growing season; STP: spring thaw period. –: missing observation due to the absence of laboratory heating system in the winter. The values in parentheses indicate the standard error of four spatial replicates.

were obviously different between the growing and non-growing seasons (Fig. 3a–c). The exponential fit showed a Q10 value of 1.79 ± 0.17 in the annual scale. The stepwise regression further proved that soil temperatures were the primary factor regulating the CH4 uptake among the observed environmental factors (soil temperature, moisture, and inorganic nitrogen concentrations). However, the CH4 uptake could be facilitated (≤–20 μg C m−2 h−1) by the high soil temperature (10–15 °C) and the appropriate soil moisture (WFPS: 40–75%, Fig. 4a). Simple regression analysis indicated that soil moisture significantly (p < 0.01) affected N2O emissions from the alpine shrub meadow. The dependency of N2O fluxes on soil WFPS could be described by an exponential function in the non-growing season and annual scale (Fig. 3d–f). The stepwise regression analysis suggested that soil WFPS and NH4+ concentrations jointly regulated the seasonal variation of 31

Agriculture, Ecosystems and Environment 255 (2018) 27–36

Y. Fu et al.

Fig. 3. Correlation between soil temperature (Ts, 5 cm in depth) and CH4 fluxes, and soil water-filled pore space (WFPS, 0–6 cm in depth) and N2O fluxes in the annual scale (a, d), growing (b, e) and non-growing seasons(c, f).

et al. (2006) conducted year-round N2O flux measurements in an ungrazed Leymus chinensis steppe using the static chamber-gas chromatography method with pure N2 as a carrier gas in the years 1995, 1998, and 2001–2003 (sampling frequency: one to three times per month). The averaged N2O emission was 0.73 ± 0.52 kg N ha−1 yr−1 during the measurement period. Although no freeze–thaw event was measured, the estimated total annual emission was 1.62 ± 0.53 kg N ha−1 yr−1 in 2001. Liu et al. (2010b) measured the N2O fluxes in the same plot using the Ar-CH4 method in 2005 and 2006 (sampling frequency: one to two times per month). The annual N2O emissions from the same plot were calculated to be only 0.24 ± 0.17

the ECD of the gas chromatograph (the N2-CO2 method). To date, more and more researchers have adopted the new N2-CO2 method to quantify N2O fluxes over terrestrial ecosystems in China. It is therefore important to answer: (I) whether or not the pure N2 method also significantly overestimated N2O emissions from grassland ecosystems; and (II) whether or not the N2O fluxes observed by the N2-CO2 method were comparable to the Ar-CH4 and N2-ascarite methods. To answer these two questions, we reviewed the year-round N2O flux measurements over natural grasslands in China (Table 5). The highest N2O emissions (> 1 kg N ha−1 yr−1) were exclusively obtained using the pure N2 method (Du et al., 2006, 2008; Li et al., 2012a, b). Du

Fig. 4. Effects of soil temperature (Ts, 5 cm in depth), water-filled pore space (WFPS, 0–6 cm in depth) and inorganic nitrogen concentrations (Inorganic N, 0–10 cm in depth) on (a) CH4 uptake (μg C m−2 h−1) and (b) N2O emissions (μg N m−2 h−1).

32

Agriculture, Ecosystems and Environment 255 (2018) 27–36

Y. Fu et al.

Table 5 Review of annual CH4 and N2O flux measurements over grasslands in China. a

Grassland

Treatment

CH4

LC steppe

Summer-grazed Winter-grazed Ungrazed Ungrazed 1979 Ungrazed 1979 Grazed Ungrazed 1999 Winter-grazed Ungrazed 2004 Lightly grazed Moderately grazed Heavily grazed Ungrazed 1979 Grazed Grazed Ungrazed Lightly grazed Grazed Herbaceous block Shrub community Herbaceous + shrub

1.4 2.4 1.1 / / / 3.8 (0.1) 2.3 (0.1) 4.2 (0.1) 4.5 (0.1) 4.0 (0.1) 3.3 (0.1) 4.3 (0.1) 3.5 (0.1) / / / 1.6 (0.7) 2.6 (0.7) 1.9 (0.8) 1.33–1.35

SG steppe Desert steppe Alpine SP grassland Alpine KH meadow Alpine PF meadow

NGSCH4

28–43% / / / 35% 23% 35% 32% 36% 36% 40% 40% / / / 42% 47% 43% 25–39%

b

Q10

c

1.50 1.80 1.60 / / / 1.84 2.10 / 1.74 1.52 1.58 1.54 1.48 / / / 1.11 1.23 1.37 1.79

N2O

d

NGSN2O

0.19 (0.03) 0.15 (0.03) 0.43 (0.07) 0.25–1.26 0.27 (0.03) 0.23 (0.09) 0.22 (0.07) 0.01 (0.03) 0.23 (0.05) 0.15 (0.05) 0.13 (0.02) 0.15 (0.02) / / 0.22 (0.06) 1.11 (0.02) 0.73 (0.01) 2.20 (0.16) 2.88 (0.15) 2.22 (0.21) 0.18–0.27

58% 47% 79% / / / 73% 0% 85% 78% 64% 43% / / / 24% 27% 26% 23% 29% 67–74%

e

GC set-up

f

Literature

N2-CO2

Rong et al. (2015); Yang et al. (2015); Liu et al. (2017)

Pure N2 Ar-CH4

Du et al. (2006) Liu et al. (2010b); Peng et al. (2011)

N2-ascarite

Wolf et al. (2010); Chen et al. (2011)

/

Chen et al. (2011)

Ar-CH4 Pure N2

Wang et al. (2011) Li et al. (2012b) Du et al. (2008); Guo et al. (2015)

N2-CO2

This study

a. Annual CH4 uptake (kg C ha−1 yr−1); b. Contribution of cumulative CH4 uptake during non-growing seasons to annual totals; c. Q10 represents the exponential change in CH4 uptake due to a temperature change (10 °C); d. Annual N2O emission (kg N ha−1 yr−1); e. Contribution of cumulative N2O emission during non-growing seasons to annual totals; f. Gas chromatographic set-ups./: no measurement or report. LC: Leymus chinensis; SG: Stipa grandis; SP: Stipa purpurea; KH: Kobresia humilis; PF: Potentilla fruticose.

and 0.30 ± 0.21 kg N ha−1 yr−1 in 2005 and 2006, respectively (Liu et al., 2010b). The other contrasting case is that of Du et al. (2008), who measured the N2O fluxes in a Potentilla fruticosa meadow, close to the present study area, using the pure N2 method in the years 2003–2006 (sampling frequency: one to four times per month). They reported that the annual N2O emissions from the alpine shrub meadow were up to 1.85–2.61 kg N ha−1 yr−1, which is one order of magnitude higher than the emissions observed in this study. Although the differences could partly be ascribed to the inter-annual variations, the multi-year flux measurements showed that the inter-annual variations did not normally reach up to one order of magnitude in natural grasslands (Liu et al., 2010b; Rong et al., 2015; Yang et al., 2015; this study). The extremely high N2O emissions from unfertilized natural grasslands obtained using the pure N2 method were already as high as the background emissions from fertilized croplands (Gu et al., 2007), which is highly unlikely considering the general limitation of nitrogen substrates in natural grasslands of China (Liu et al., 2008). The gas chromatographic set-up with pure N2 as the carrier gas was very likely significantly overestimating the N2O emissions from natural grasslands in China. Meanwhile, the flux measurements conducted by the Ar-CH4, N2-ascarite, and N2-CO2 methods consistently showed that the unfertilized natural grasslands functioned as weak N2O sources (0.01–0.43 kg N ha−1 yr−1, Table 5).

contributing proportions of non-growing seasons to the full year were calculated to be 25–39% in the winter-grazed alpine Potentilla fruticosa meadow in the Qinghai-Tibetan Plateau. These proportions (25–39%) fell within the range of 23–43% (with a mean of 35% for 11 cases) obtained from the year-round measurements in semi-arid Leymus chinensis and Stipa grandis steppes, subject to different grazing management practices (Table 5), and were slightly lower than the proportions (42–47%) of alpine Kobresia humilis and Potentilla fruticosa meadows reported by Guo et al. (2015). Overall, the CH4 uptake in non-growing seasons contributed between 23% and 47% of annual total uptake, with an average of 36% (4 grassland types and 15 treatments; see Table 5). The CH4 uptake of natural grasslands in China was generally low in the non-growing seasons compared to the growing seasons. However, due to the long duration of non-growing seasons (190–210 days) in the northern temperate zone, the cumulative uptake of non-growing seasons accounted for a considerable proportion (around a third) of the annual uptake. The soil temperatures were found to play the primary role in regulating the seasonal variation of CH4 uptake, being able to explain 59% of the flux variances in the present study. The Q10 value (1.25 for the growing seasons) of the present Potentilla fruticosa shrub meadow was much lower than the values of the alpine Stipa purpurea steppe and Kobresia pygmae meadow at higher altitudes of the Qinghai-Tibetan Plateau (Q10 = 1.67 and 1.72, respectively, for growing seasons; altitude: 4700–5200m, Wei et al., 2015). Overall, the Q10 values for alpine grasslands (range: 1.1–1.8, average: 1.4) were slightly lower than the values for temperate semi-arid grasslands (range: 1.5–2.1, average: 1.7, Table 5). The lower Q10 values and annual CH4 uptake (range: 1.3–2.6 kg C ha−1 yr−1, average: 1.8 kg C ha−1 yr−1) in alpine grasslands indicated that global warming at the same extent would stimulate less of an increase in CH4 uptake by alpine grasslands on the QinghaiTibetan Plateau than those in the temperate semi-arid grasslands (range: 1.1–4.5 kg C ha−1 yr−1, average: 3.2 kg C ha−1 yr−1, Table 5). As already discussed, the gas chromatographic set-up using pure N2 as a carrier gas may significantly overestimate N2O emissions by up to one order of magnitude in natural grassland ecosystems. The other factor in determining the magnitude of N2O emissions from unfertilized natural grasslands is the freeze–thaw events occurring during the transition period between non-growing and growing seasons. A few studies performing full-year measurements have found a significant

4.2. Contributions of non-growing season fluxes to annual exchanges The natural grasslands of China (nearly 80% of the grassland area) are mainly distributed in the northern temperate and high-altitude zones. Accordingly, conducting year-round flux measurements, especially in the cold climate conditions of the Qinghai-Tibetan Plateau, are relatively difficult, and such data are scarce. Wei et al. (2015) conducted CH4 flux measurements in an alpine meadow and steppe of the Qinghai-Tibetan Plateau over six growing seasons. Due to the missing measurements in non-growing seasons, they estimated the annual CH4 uptake assuming that the cumulative uptake in the non-growing seasons accounted for 24% of total annual uptake. For the same reason, Wang et al. (2014) assumed a similar factor of 30% to extrapolate measurements in the growing season to annual fluxes in the arid and semi-arid regions in northern China, and in the Qinghai-Tibetan Plateau. In this study, performing intensive year-round measurements, the 33

Agriculture, Ecosystems and Environment 255 (2018) 27–36

Y. Fu et al.

4.3. Necessity of year-round and multi-year flux measurements

contribution from spring thaw periods (35–81%) and non-growing seasons (55–95%) to the annual total N2O emission in the semi-arid steppes (Table 5). Heavy grazing significantly inhibited the freeze–thaw effect, and thereby reduced the annual N2O emissions from grazed steppes compared to ungrazed steppes (Wolf et al., 2010; Yang et al., 2015). However, Li et al. (2012b) did not observe any pulse N2O emissions during the spring thaw period, or any grazing effects at the grazing management treatments of Tianshan alpine grasslands in central Asia. The N2O emissions during the spring thaw period only accounted for 6.6% of the annual total emissions. The remarkable differences in the magnitude of N2O emissions and the importance of nongrowing season emissions might be explained by the differences in precipitation (multi-year average: 330–380 mm yr−1 reported by Wolf et al., 2010 and Yang et al., 2015; 266 mm yr−1 by Li et al., 2012b). The flux measurements were conducted in three successive spring thaw periods from 2013 to 2015 in the present study. The pulse N2O emissions were not observed during the 1st spring thaw period but lasted for approximately one week during the 2nd spring thaw period, and one month during the 3rd. The cumulative emissions during the 2nd and 3rd spring thaw periods (48 days) contributed to 33–52% of the annual total emissions. We can clearly see that the increased precipitation from 1st to 3rd year-round period steadily enhanced the duration of pulse N2O emissions (0, 7 and 30 d in the 1st, 2nd and 3rd years) during the spring thaw period and the contributions from the spring thaw period to annual emissions (33% and 52% in the 2nd and 3rd years, respectively). Since 80–90% of the total annual precipitation concentrated on the growing seasons, the precipitation of the previous growing season might be a major determinant of the magnitude of N2O emissions during the spring thaw period. The precipitation during the nongrowing seasons were quite similar (99.3 and 84.7 mm), but the duration of pulse N2O emissions increased from one week to one month in the 2nd and 3rd year-round periods. The phenomenon further indicated the importance of the precipitation in the previous growing season to the N2O emissions during the spring thaw period. The high availability of substrates and anaerobic conditions were two requisite factors in triggering pulse N2O emissions during the spring thaw periods. The soil nitrogen mineralization rates in natural grasslands were at the same level or even higher during the spring thaw periods compared to the growing seasons (Wu et al., 2012), and thus there was more inorganic nitrogen available during the spring thaw periods, since the inactive plant roots could not utilize the mineralized nitrogen. Indeed, we observed higher inorganic nitrogen concentrations (> 5 mg N kg−1 dry soil) during the spring thaw periods. However, the relatively high inorganic nitrogen concentrations failed to trigger the pulse N2O emissions during the 1st spring thaw period. The anaerobic environment, as the other essential factor in triggering pulse N2O emissions, was missing during the 1st spring thaw period as a result of lower precipitation and soil moistures in the 1st year than in the 2nd and 3rd years (Table 3). During the spring thaw period, higher soil moisture allowed more ice to block the surface soil pores, and thus prevent infiltration of unfrozen water to deeper soil layers during the daytime. Subsequently, anaerobic microsites or even waterlogged environments were formed in the topsoil, where the unfrozen water enabled the mass transfer of abundant carbon and nitrogen substrates (Regina et al., 2004). As a result, microbial denitrification was enhanced, whereby N2O production was stimulated as an intermediate product. Our data, as well as earlier studies (Koponen and Martikainen, 2004; Holst et al., 2008; Wolf et al., 2010), demonstrated that pulse N2O emissions during the spring thaw periods relied heavily on high soil moisture conditions. The tremendous inter-annual variations of N2O emissions during the spring thaw periods could well be explained by the differences in soil moisture and precipitation in the present study (Table 3 and Figs. 3d–f). The total annual precipitation, especially the precipitation of the previous growing season, was the primary regulator of the soil moisture during the spring thaw period, and could therefore be regarded as a predictor of N2O emissions during the spring thaw period.

The year-round flux measurements over natural grasslands in China showed that the non-growing season fluxes contributed 23–47% and 0–85% of the annual CH4 uptake and N2O emissions, respectively. Once the freeze–thaw event had occurred, the N2O emissions in the nongrowing seasons dominated (> 50%) the annual emissions (Table 5). Although the soil temperatures and moistures might be the crucial factors regulating the seasonal variation of gas fluxes, the present study showed that the correlations of environmental factors and gas fluxes were quite different between the growing and non-growing seasons. Thus, the correlations obtained by the single factor analysis of environmental factors and fluxes in the growing seasons could not be used to predict the CH4 and N2O fluxes in the non-growing seasons. Furthermore, the pulse N2O emissions during the spring thaw periods lasted from a few days (as in the 2nd spring thaw period) to one month (as in the 3rd spring thaw period). The sampling frequency in the nongrowing seasons should be higher than a weekly frequency. Otherwise, the freeze-thaw effects on N2O emissions might not be detected (like the pulse N2O emissions during the 2nd spring thaw period in the present study). Once the pulse N2O emissions occurred, the sampling frequency should be increased to a daily frequency. However, the bi-weekly or monthly measurements were generally adopted by most of the yearround flux measurements in natural grasslands of China, which might not be enough to capture the pulse emissions and quantify the contributions from non-growing seasons to annual cumulative emissions. Based on the flux data obtained by this study, we analyzed the effects of extremely low sampling frequencies (bi-weekly or monthly) on the estimations of annual CH4 uptake and N2O emissions. If the bi-weekly and monthly sampling frequencies were adopted, the annual CH4 uptake might be underestimated by −3.9 (bi-weekly) and −5.4% (monthly) or overestimated by 5.5 (bi-weekly) and 7.6% (monthly), respectively. The annual N2O emissions might be largely underestimated up to −31.4 (bi-weekly) and −44.7% (monthly) or overestimated up to 27.4 (bi-weekly) and 40.0% (monthly). Obviously, the year-round and nongrowing season flux measurements with a suitable sampling frequency (daily to weekly measurements) were essential for the quantification of annual gas exchanges, especially for the annual N2O emissions, in natural grasslands. The natural grasslands in China mainly distributed in the semi-arid (200–400 mm yr−1) and semi-humid areas (400–800 mm yr−1). The precipitation generally presented huge inter-annual variations in these areas. The total annual precipitation fluctuated between 326 and 850 mm from 1980 to 2012 in the present experimental area (Zhang et al., 2014). The remarkably inter-annual variation of precipitation significantly influenced the magnitude of soil moisture during the spring thaw period. The pulse N2O emissions during the spring thaw periods were highly dependent on the soil moisture conditions. The N2O emissions might present a tremendous inter-annual variation with the fluctuation of precipitation. It is arbitrary to draw a conclusion about the effects of freeze–thaw events on annual N2O emission based on a single year of measurements, due to the large inter-annual variation of precipitation in northern China. Thus, the multi-year flux measurements were essential for the evaluation of freeze-thaw effects on annual N2O emissions. 4.4. Importance of alpine meadows for grassland greenhouse gas exchanges Alpine meadows, as the largest grassland type (64 million ha), account for nearly a fifth of the grassland area in China (Zhang et al., 2010). Year-round flux measurements remain scarce in alpine meadows considering their vast area. The few year-round measurements in alpine meadows showed that the CH4 uptake (1.3–2.6 kg C ha−1 yr−1 in the present study and Guo et al., 2015) was in the low range compared to the Eurasian semi-arid steppes (1.1–4.5 kg C ha−1 yr−1; Table 5). The relatively low soil moisture and high temperature conditions (averaged 34

Agriculture, Ecosystems and Environment 255 (2018) 27–36

Y. Fu et al.

Acknowledgements

WFPS and air temperature: < 50% and > 0 °C, respectively) in semiarid steppes were beneficial to the CH4 diffusion to the soil and the activity of methanotrophs compared to alpine meadows (annual means of WFPS and air temperature: > 50% and < 0 °C, respectively). However, overgrazing and the water stress of methanotrophs induced by extreme drought in semi-arid regions could lower the CH4 uptake of Eurasian steppes to the same level of alpine meadows. Even with the very cold climate in the Qinghai-Tibetan Plateau, the annual CH4 uptake of alpine meadows by far exceeded the uptake rates calculated by a process-based ecosystem model (0–1 kg C ha−1 yr−1 in the Qinghai-Tibetan Plateau; Tian et al., 2011). Considering the vast areas (64 million ha) and considerable CH4 uptake (1.3–2.6 kg C ha−1 yr−1; Table 5), the CH4 uptake by alpine meadows in China might reach up to the level of 85–166 Gg C yr−1. Tian et al. (2011) estimated the net CH4 uptake of grasslands in China to be 465 Gg C yr−1 in 1961–2005, based on the Dynamic Land Ecosystem Model. Therefore, alpine meadows might contribute approximately one-fifth to one-third of the CH4 uptake by grasslands in China, and thus might be the most important CH4 sink. If the N2O flux data obtained by the pure N2 method were excluded, these year-round measurements showed that the N2O emissions from natural grasslands of China were very weak. The annual N2O emissions from the unfertilized natural grasslands ranged from 0.01 to 0.43 kg N ha−1 yr−1, with an average of 0.20 ± 0.03 kg N ha−1 yr−1 for 13 cases (Table 5). The N2O emissions from natural grasslands in China were calculated to be 67 ± 10 Gg N yr−1 based on the product of averaged flux (0.20 ± 0.03 kg N ha−1 yr−1) and grassland area (337 million ha), which was close to the values estimated by the processbased models (77 ± 13 and 71 Gg N yr−1 by Zhang et al., 2010 and Tian et al., 2011). The field measurements and model estimations together demonstrated that the unfertilized natural grasslands in China were very weak N2O sources. A general lack of nitrogen substrate availability for microbial nitrification and denitrification limited N2O production and emission in unfertilized natural grasslands in China (Liu et al., 2008). The previously reported high emissions obtained by the pure N2 method needs to be carefully reappraised.

This study was funded by the National Natural Science Foundation of China (41375152, 41603075) and by the Ministry of Science and Technology (2016YFA0602303, 2016YFA0600804). We thank the Haibei Alpine Meadow Ecosystem Research Station, Chinese Academy of Sciences, for providing infrastructure and help. References Chen, W.W., Wolf, B., Zheng, X.H., Yao, Z.S., Butterbach-Bahl, K., Brüggemann, N., Liu, C.Y., Han, S.H., Han, X.G., 2011. Annual methane uptake by temperate semiarid steppes as regulated by stocking rates, aboveground plant biomass and topsoil air permeability. Glob. Change Biol. 17, 2803–2816. Chen, W.W., Zheng, X.H., Chen, Q., Wolf, B., Butterbach-Bahl, K., Brüggemann, N., Lin, S., 2013. Effects of increasing precipitation and nitrogen deposition on CH4 and N2O fluxes and ecosystem respiration in a degraded steppe in Inner Mongolia, China. Geoderma 192, 335–340. Du, R., Lu, D., Wang, G.C., 2006. Diurnal, seasonal, and inter-annual variations of N2O fluxes from native semi-arid grassland soils of inner Mongolia. Soil Biol. Biochem. 38, 3474–3482. Du, Y.G., Cui, Y.G., Xu, X.L., Liang, D.Y., Long, R.J., Cao, G.M., 2008. Nitrous oxide emissions from two alpine meadows in the Qinghai-Tibetan Plateau. Plant Soil 311, 245–254. Gu, J.X., Zheng, X.H., Wang, Y.H., Ding, W.X., Zhu, B., Chen, X., Wang, Y.H., Zhao, Z.C., Shi, Y., Zhu, J.G., 2007. Regulatory effects of soil properties on background N2O emissions from agricultural soils in China. Plant Soil 295, 53–65. Guo, X.W., Du, Y.G., Han, D.H., Xu, X.L., Zhang, F.W., Li, L., Li, Y.K., Liu, S.L., Ouyang, J.Z., Cao, G.M., 2015. Effects of landuse change on CH4 soil-atmospheric exchange in alpine meadow on the Tibetan Plateau. Pol. J. Environ. Stud. 24, 1593–1602. Holst, J., Liu, C.Y., Brüggemann, N., Butterbach-Bahl, K., Zheng, X.H., Wang, Y.S., Han, S.H., Yao, Z.S., Yue, J., Han, X.G., 2007. Microbial N turnover and N-oxide (N2O/NO/ NO2) fluxes in semi-arid grassland of Inner Mongolia. Ecosystems 10, 623–634. Holst, J., Liu, C.Y., Yao, Z.S., Brüggemann, N., Zheng, X.H., Giese, M., Butterbach-Bahl, K., 2008. Fluxes of nitrous oxide, methane and carbon dioxide during freezingthawing cycles in an Inner Mongolian steppe. Plant Soil 308, 105–117. Hu, Y.G., Chang, X.F., Lin, X.W., Wang, Y.F., Wang, S.P., Duan, J.C., Zhang, Z.H., Yang, X.X., Luo, C.Y., Xu, G.P., Zhao, X.Q., 2010. Effects of warming and grazing on N2O fluxes in an alpine meadow ecosystem on the Tibetan Plateau. Soil Biol. Biochem. 42, 944–952. Hutchinson, G.L., Mosier, A.R., 1981. Improved soil cover method for field measurement of nitrous oxide fluxes. Soil Sci. Soc. Am. J. 45, 311–316. Jiang, C.M., Yu, G.R., Fang, H.J., Cao, G.M., Li, Y.N., 2010. Short-term effect of increasing nitrogen deposition on CO2, CH4 and N2O fluxes in an alpine meadow on the QinghaiTibetan Plateau. China. Atmos. Environ. 44, 2920–2926. Koponen, H.T., Martikainen, P.J., 2004. Soil water content and freezing temperature affect freeze-thaw related N2O production in organic soil. Nutr. Cycl. Agroecos. 69, 213–219. Li, Y.N., Zhao, X.Q., Cao, G.M., Zhao, L., Wang, Q.X., 2004. Analyses on climates and vegetation productivity background at Haibei alpine meadow ecosystem research station. Plateau Meteoro. 23, 558–567 (in Chinese). Li, Y.N., Sun, X.M., Zhao, X.Q., Zhao, L., Xu, S.X., Gu, S., Zhang, F.W., Yu, G.R., 2006. Seasonal variations and mechanism for environmental control of NEE of CO2 concerning the Potentilla fruticosa in alpine shrub meadow of Qinghai-Tibet Plateau. Sci. China Ser. D. 49, 174–185. Li, K.H., Gong, Y.M., Song, W., He, G.X., Hu, Y.K., Tian, C.Y., Liu, X.J., 2012a. Responses of CH4, CO2 and N2O fluxes to increasing nitrogen deposition in alpine grassland of the Tianshan Mountains. Chemosphere 88, 140–143. Li, K.H., Gong, Y.M., Song, W., Lv, J.L., Chang, Y.H., Hu, Y.K., Tian, C.Y., Christie, P., Liu, X.J., 2012b. No significant nitrous oxide emissions during spring thaw under grazing and nitrogen addition in an alpine grassland. Glob. Change Biol. 18, 2546–2554. Li, X.L., Perry, G.L.W., Brierley, G., Sun, H.Q., Li, C.H., Lu, G.X., 2014. Quantitative assessment of degradation classifications for degraded alpine meadows (heitutan), Sanjiangyuan, Western China. Land Degrad. Dev. 25, 417–427. Li, Y.Y., Dong, S.K., Liu, S.L., Zhou, H.K., Gao, Q.Z., Cao, G.M., Wang, X.X., Su, X.K., Zhang, Y., Tang, L., Zhao, H.D., Wu, X.Y., 2015. Seasonal changes of CO2, CH4 and N2O fluxes in different types of alpine grassland in the Qinghai-Tibetan Plateau of China. Soil Biol. Biochem. 80, 306–314. Lin, X.W., Wang, S.P., Ma, X.Z., Xu, G.P., Luo, C.Y., Li, Y.N., Jiang, G.M., Xie, Z.B., 2009. Fluxes of CO2, CH4, and N2O in an alpine meadow affected by yak excreta on the Qinghai-Tibetan Plateau during summer grazing periods. Soil Biol. Biochem. 41, 718–725. Liu, C.Y., Holst, J., Brüggemann, N., Butterbach-Bahl, K., Yao, Z.S., Han, S.H., Han, X.G., Zheng, X.H., 2008. Effects of irrigation on nitrous oxide,methane and carbon dioxide fluxes in an inner mongolian steppe. Adv. Atmos. Sci. 25, 748–756. Liu, C.Y., Zheng, X.H., Zhou, Z.X., Han, S.H., Wang, Y.H., Wang, K., Liang, W.G., Li, M., Chen, D.L., Yang, Z.P., 2010a. Nitrous oxide and nitric oxide emissions from an irrigated cotton field in Northern China. Plant Soil 332, 123–134. Liu, X.R., Dong, Y.S., Qi, Y.C., Li, S.G., 2010b. N2O fluxes from the native and grazed semi-arid steppes and their driving factors in Inner Mongolia. China. Nutr. Cycl. Agroecosyst. 86, 231–240. Liu, C.Y., Wang, K., Zheng, X.H., 2013. Effects of nitrification inhibitors (DCD and DMPP) on nitrous oxide emission, crop yield and nitrogen uptake in a wheat–maize cropping

5. Conclusion The annual CH4 and N2O fluxes were quantified in a winter-grazed alpine Potentilla fruticosa shrub meadow in the Qinghai-Tibetan Plateau, using the static chamber method with pure N2 as the carrier gas and a mixture of CO2 and N2 as the make-up gas of the gas chromatograph. The annual CH4 uptake was in the low range of the reported uptakes in grasslands of China. Soil temperatures were identified as the primary factor regulating the seasonality of CH4 uptake. The low Q10 values and CH4 uptake indicated that global warming at the same extent would result in less of an increase in the CH4 sink of alpine meadows in the Qinghai-Tibetan Plateau, compared to other grassland areas. The N2O emissions during the spring thaw period showed a tremendous interannual variation. The high availability of substrates and soil moisture during the spring thaw periods together provided the conditions for pulse N2O emissions. The total annual precipitation or the precipitation during the previous growing season could be an indicator for N2O emissions during the spring thaw period. The cumulative CH4 and N2O fluxes during the non-growing seasons contributed significantly to the total annual exchanges, and these measurements were therefore essential to quantify annual exchanges in natural grasslands. Improper measuring methods, such as the pure N2 method of gas chromatograph, and very low measuring frequencies (bi-weekly to monthly) might result in a significant overestimate or underestimate of the annual N2O emissions from natural grasslands. Excluding the flux data obtained by the pure N2 method, the natural grasslands of China were generally characterized by weak N2O emissions.

35

Agriculture, Ecosystems and Environment 255 (2018) 27–36

Y. Fu et al.

Wei, D., Xu, R., Tarchen, T., Wang, Y.S., Wang, Y.H., 2015. Considerable methane uptake by alpine grasslands despite the cold climate: in situ measurements on the central Tibetan Plateau, 2008–2013. Glob. Change Biol. 21, 777–788. Wolf, B., Zheng, X.H., Brüggemann, N., Chen, W.W., Dannenmann, M., Han, X.G., Sutton, M.A., Wu, H.H., Yao, Z.S., Butterbach-Bahl, K., 2010. Grazing-induced reduction of natural nitrous oxide release from continental steppe. Nature 464, 881–884. Wu, H.H., Dannenmann, M., Wolf, B., Han, X.G., Zheng, X.H., Butterbach-Bahl, K., 2012. Seasonality of soil microbial nitrogen turnover in continental steppe soils of Inner Mongolia. Ecosphere 3, 34. Yang, X.M., Chen, H.Q., Gong, Y.S., Zheng, X.H., Fan, M.S., Kuzyakov, Y., 2015. Nitrous oxide emissions from an agro-pastoral ecotone of northern China depending on land uses. Agr. Ecosyst. Environ. 213, 241–251. Yashiro, Y., Shizu, Y., Hirota, M., Shimono, A., Ohtsuka, T., 2010. The role of shrub (Potentilla fruticosa) on ecosystem CO2 fluxes in an alpine shrub meadow. J. Plant Eco. 3, 89–97. Zhang, Y.M., Liu, J.K., 2003. Effects of plateau zokors (Myospalax Fontanierii) on plant community and soil in an alpine meadow. J. Mammal. 84, 644–651. Zhang, F., Qi, J., Li, F.M., Li, C.S., Li, C.B., 2010. Quantifying nitrous oxide emissions from Chinese grasslands with a process-based model. Biogeosciences 7, 2039–2050. Zhang, W., Liu, C.Y., Zheng, X.H., Fu, Y.F., Hu, X.X., Cao, G.M., Butterbach-Bahl, K., 2014. The increasing distribution area of zokor mounds weaken greenhouse gas uptakes by alpine meadows in the Qinghai-Tibetan Plateau. Soil Biol. Biochem. 71, 105–112. Zhao, Z.Z., Dong, S.K., Jiang, X.M., Liu, S.L., Ji, H.Z., Li, Y., Han, Y.H., Sha, W., 2017. Effects of warming and nitrogen deposition on CH4, CO2 and N2O emissions in alpine grassland ecosystems of the Qinghai-Tibetan Plateau. Sci. Total Environ. 592, 565. Zheng, X.H., Mei, B.L., Wang, Y.S., Xie, B.H., Wang, Y.S., Dong, H.B., Xu, H., Chen, G.X., Cai, Z.C., Yue, J., Gu, J.X., Su, F., Zou, J.W., Zhu, J.G., 2008. Quantification of N2O fluxes from soil?plant systems may be biased by the applied gas chromatograph methodology. Plant Soil 311, 211–234. Zhou, H.K., Tang, Y.H., Zhao, X.Q., Zhou, L., 2006. Long-term grazing alters species composition and biomass of a shrub meadow on the Qinghai-Tibet Plateau. Pak. J. Bot. 38, 1055–1069. Zhu, X.X., Luo, C.Y., Wang, S.P., Zhang, Z.H., Cui, S.J., Bao, X.Y., Jiang, L.L., Li, Y.M., Li, X.N., Wang, Q., Zhou, Y., 2015. Effects of warming: grazing/cutting and nitrogen fertilization on greenhouse gas fluxes during growing seasons in an alpine meadow on the Tibetan Plateau. Agr. Forest Meteoro. 214–215, 506–514.

system. Biogeosciences 10, 2427–2437. Liu, J., Chen, H.Q., Yang, X.M., Gong, Y.S., Zheng, X.H., Fan, M.S., Kuzyakov, Y., 2017. Annual methane uptake from different land uses in an agro-pastoral ecotone of northern China. Agr. Forest Meteoro. 236, 67–77. Pei, Z.Y., Ouyang, H., Zhou, C.P., Xu, X.L., 2003. Fluxes of CO2, CH4, N2O from alpine grassland in the Tibetan Plateau. J. Geogr. Sci. 13, 27–34. Peng, Q., Qi, Y.C., Dong, Y.S., Xiao, S.S., He, Y.T., 2011. Soil nitrous oxide emissions from a typical semiarid temperate steppe in inner mongolia: effects of mineral nitrogen fertilizer levels and forms. Plant Soil 342, 345–357. Regina, K., Syväsalo, E., Hannukkala, A., Esala, M., 2004. Fluxes from N2O from peat soils in finland. Eur. J. Soil Sci. 55, 591–599. Rong, Y.P., Ma, L., Johnson, D.A., 2015. Methane uptake by four land-use types in the agro-pastoral region of northern China. Atmos. Environ. 116, 12–21. Tian, H.Q., Xu, X.F., Lu, C.Q., Liu, M.L., Ren, W., Chen, G.S., Melillo, J., Liu, J.Y., 2011. Net exchanges of CO2, CH4, and N2O between China's terrestrial ecosystems and the atmosphere and their contributions to global climate warming. J. Geophys. Res. 116, G02011. Wang, Q.J., Zhou, X.M., Zhang, Y.Q., Zhao, X.Q., 1991. Structure characteristics and biomass of Potentilla fruticosa shrub in Qinghai-Xizang plateau. Acta Bot. Boreal.occident. Sin. 11, 333–340 (in Chinese). Wang, Y.S., Xue, M., Zheng, X.H., Ji, B.M., Du, R., Wang, Y.F., 2005. Effects of environmental factors on N2O emission from and CH4 uptake by the typical grasslands in the Inner Mongolia. Chemosphere 58, 205–215. Wang, Y.H., Wang, Y.S., Ling, H., 2010. A new carrier gas type for accurate measurement of N2O by GC-ECD. Adv. Atmos. Sci. 27, 1322–1330. Wang, Z.W., Hao, X.Y., Shan, D., Han, G.D., Zhao, M.L., Willms, W.D., Wang, Z., Han, X., 2011. Influence of increasing temperature and nitrogen input on greenhouse gas emissions from a desert steppe soil in Inner Mongolia. Soil Sci. Plant Nutr. 57, 508–518. Wang, Y.F., Chen, H., Zhu, Q., Peng, C.H., Wu, N., Yang, G., Zhu, D., Tian, J.Q., Tian, L.X., Kang, X.M., He, Y.X., Gao, Y.H., Zhao, X.Q., 2014. Soil methane uptake by grasslands and forests in China. Soil Biol. Biochem. 74, 70–81. Wei, D., Xu, R., Wang, Y.H., Wang, Y.S., Liu, Y.W., Yao, T.D., 2012. Responses of CO2, CH4 and N2O fluxes to livestock exclosure in an alpine steppe on the Tibetan Plateau, China. Plant Soil 359, 45–55. Wei, D., Xu, R., Liu, Y.W., Wang, Y.H., Wang, Y.S., 2014. Three-year study of CO2 efflux and CH4/N2O fluxes at an alpine steppe site on the central Tibetan Plateau and their responses to simulated N deposition. Geoderma 232–234, 88–96.

36