Article pubs.acs.org/est
Chemical Composition of Microbe-Derived Dissolved Organic Matter in Cryoconite in Tibetan Plateau Glaciers: Insights from Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Analysis Lin Feng,† Jianzhong Xu,*,† Shichang Kang,† Xiaofei Li,† Yang Li,‡ Bin Jiang,§ and Quan Shi§ †
State Key Laboratory of Cryospheric Sciences, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences, Lanzhou 730000, China ‡ Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing 100101, China § State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Beijing 102249, China S Supporting Information *
ABSTRACT: Cryoconite in mountain glaciers plays important roles in glacial ablation and biogeochemical cycles. In this study, the composition and sources of dissolved organic matter (DOM) in cryoconite from the ablation regions of two Tibetan Plateau glaciers were determined using electrospray ionization (ESI) Fourier transform ion cyclotron resonance mass spectrometry (FT-ICRMS) and fluorescence spectrometry. A marked absorbance between 300 and 350 nm in the DOM absorption spectra was observed which was consistent with microbe-derived mycosporine-like amino acids. Fluorescence excitation− emission matrices showed that DOM had intense signals at protein-like substance peaks and weak signals at humic-like substance peaks. The highresolution mass spectra of FT-ICR-MS showed cryoconite DOM from both glaciers contained diverse lignins, lipids, proteins, and unsaturated hydrocarbons. The lipids and proteins were consistent with material from microbial sources, and the lignins and unsaturated hydrocarbons were probably from vascular plant material supplied in atmospheric aerosols and debris from around the glaciers. Almost one-third of the identified DOM molecules had low C/N ratios (≤20), indicating their high bioavailability. Using a conservative cryoconite distribution on Chinese mountain glacier surfaces (6%) and an average debris mass per square meter of cryoconite (292 ± 196 g m−2), we found that the amount of DOC produced in cryoconite on Chinese glaciers as much as 0.23 ± 0.1 Gg per cryoconite formation process. This dissolved organic carbon may absorb solar radiation, accelerate glacial melting, and be an important source of bioavailable DOM to proglacial and downstream aquatic ecosystems.
■
INTRODUCTION Glaciers and ice sheets on the earth have been under rapid melting in recent years because of global warming, and the meltwater is thought to contain amounts of organic matter that can affect downstream aquatic ecology.1 The biolabilities of dissolved organic matter (DOM) in glaciers around the world have therefore been quantified in several studies.2−4 Spencer et al.5 recently found that DOM in the Tibetan Plateau (TP) glaciers is high biolabile (20%−70%) and that the oldest carbon is the most biolabile. In order to evaluate the importance of organic matter in global glacier and ice sheet on biogeochemical cycle, the sizes of the carbon pools (including dissolved organic carbon (DOC) and particulate organic carbon) in glaciers and ice sheets (such as the Greenland Glacier Sheet and glaciers and ice sheets in Antarctica) have been estimated.4,6,7 Much less DOC is stored in glaciers and ice sheets (∼6 Pg) than that in permafrost soils (1600 Pg) around the world,7,8 but DOC in glaciers is highly active and labile. It has been shown in recent studies that glacial meltwater is also enrich in dissolved nitrogen species,9 which is an essential nutrient for microbial © XXXX American Chemical Society
community. For instance, nitrate concentrations can be up to 200 times higher in glacial-meltwater-fed lakes than in snowmelt-fed lakes in the Rocky Mountains of North America.10,11 These studies showed that glaciers and ice sheets are important sources of DOM and can influence nutrient cycles in aquatic ecosystems. The DOM in glaciers and ice sheets could originate from atmospheric wet and dry deposition,12 which are the dominant sources for remote polar ice sheets,13 and debris around glaciers, which can be transported onto glaciers as the glaciers melt (this is particularly evident on mountain glaciers). Singer et al.3 determined the chemical characteristics of organic matter in European alpine glaciers and found that the organic matter was very diverse in terms of peptides, phenolic compounds, and lipids which could also be derived from in situ microbial Received: Revised: Accepted: Published: A
August 9, 2016 November 10, 2016 November 16, 2016 November 16, 2016 DOI: 10.1021/acs.est.6b03971 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 1. Map of the locations of the glaciers at the Tibetan Plateau and the cryoconite samples on the glaciers.
UV−vis spectroscopy, fluorescence spectroscopy, and Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) with electrospray ionization (ESI) source, and estimate the production of cryoconite-producted DOC.
communities. Glacier biomes were mainly composed of algae, bacteria, protozoa, and viruses, and have distinct biological structures.15 Most microbial activity on glaciers occurs in cryoconite, which is a usually dark and powdery material deposited on snow or ice surfaces. Cryoconite on glaciers is usually found as cryoconite granules or in cryoconite holes. Cryoconite granules have been found to be common on Himalayan glaciers and Chinese mountain glaciers.14,15 Most of the interest in cryoconite has been focused on the physical properties and biological processes that occur in cryoconite.14,16,17 Biological parameter analyses and molecular rDNA sequencing have unambiguously revealed that cryoconite holes contain a wide range of microorganisms.16 The biogeochemical processes that occur in cryoconite increase the DOM and biomass, and may ultimately increase the amount of solar radiation absorbed and thus accelerate melting.18 However, little information is available on the molecular composition and sources of DOM in cryoconite. The TP is the largest and highest plateau in the world, and contains the greatest number of glaciers outside the Greenland and Antarctica, which supply the headwaters of many Asian rivers. However, the glaciers are currently shrinking rapidly, which could strongly influence both the amount of water being released and the ecological environments of the large rivers it feeds.19 Cryoconite is very common on the TP glaciers because the mountains around the glaciers are being eroded strongly and is thought to accelerate its ablation.14 It has recently been suggested that DOM will be released more efficient from the mountain glaciers on the TP than those in polar regions;6 nevertheless, the potential biogeochemical effects of the DOM released from the glaciers is poorly understood. The aim of this study is to improve our understanding of the chemical composition of DOM in the TP glaciers and its biogeochemical effects. For this purpose, we determined the molecular composition and sources of DOM in cryoconite via
■
EXPERIMENTAL SECTION Sites and Sampling. Cryoconite samples were collected from Laohugou Glacier No. 12 (LHG) at the northern edge of the TP and from the Dongkemadi Glacier in the Tanggula Mountains (TGL), in the central TP, in August 2015 (Figure 1). The mean air temperatures at the sampling sites were above 0 °C during the sample collection (4.7 °C at the LHG and 4.2 °C at the TGL in August 2015, measured from meteorology stations at the termini of the glaciers). The climates for the glaciers were described by Xu et al.20 Briefly, precipitation in summer is mainly influenced by the South Asian monsoon at the TGL and the East Asian monsoon at the LHG. Mineral dust at the surface of each glacier is mostly from debris around the glacier, released because of strong erosion during the melting season. Cryoconite samples were collected from the LHG at altitudes of 4350, 4400, 4450, and 4600 m on 28 August 2015. Cryoconite samples TGL1, TGL2, TGL3, and TGL4 were collected from different parts of the TGL but at similar altitudes (mean 5500 m) on 28 August 2015. Each sample was collected using a stainless steel scoop, then placed in a precleaned glass bottle (soaked in 1% alconox solution overnight, rinsed with Milli-Q water, and combusted at 400 °C for 6 h). The samples were frozen and kept in the dark to minimize chemical and photolytic reactions before measurements. Sample Preparation. Each cryoconite sample was freezedried at −50 °C in the dark for 24 h. An 8 g aliquot of each freeze-dried cryoconite sample was suspended in 150 mL water (LCMS grade) in a 1 L baffled flask, then the flask was shaken at 120 rpm overnight (>10 h) at room temperature. The B
DOI: 10.1021/acs.est.6b03971 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 2. UV−vis absorbance spectra of the dissolved organic matter (DOM) extracted from cryoconite samples from (a) the Laohugou Glacier No. 12 (LHG) and (b) the Dongkemadi Glacier (TGL). The LHG cryoconite samples were collected at different elevations, whereas the TGL cryoconite samples were collected at the same elevation but from different parts of the glacier.
supernatant was passed through a Whatman GF/F filter with 0.7 μm pores (GE Healthcare Bio-Sciences, Pittsburgh, PA, U.S.A.), then through a Whatman 0.3 μm glass fiber filter. The filters had been combusted at 500 °C for 6 h to remove particulate and organic matter before use. An ultrapure water sample (150 mL) was used as a procedural blank. The filtered samples were stored in the dark at −20 °C prior to analysis. DOC and DON Analysis. The DOC and total nitrogen concentrations in the samples were determined using a Vario EL CN analyzer (Elementar, Hanau, Germany). Each sample was diluted by a factor of 100 with LCMS grade water, then acidified by adding 60 μL of 10% hydrochloric acid to remove inorganic carbonates. Nonpurgeable organic carbon was then oxidized by combusting the sample at 850 °C using a carrier gas with a controlled O2 concentration. The evolved gases containing carbon were converted into CO2, which was determined using a nondispersive infrared analyzer. The system was calibrated using a potassium hydrogen phthalate standard. The total nitrogen concentration in each sample was determined by decomposing the nitrogen-containing species in the sample in the combustion tube at 850 °C to yield NO. The evolved gases were cooled and dehumidified using an electronic dehumidifier, and the NO concentration was measured using an electrochemical sensor. The total dissolved organic nitrogen (DON) concentration in each sample was determined by subtracting the inorganic nitrogen (including ammonium and nitrate) content, which was determined using two ion chromatography systems, from the total nitrogen content.21 Optical Analysis. Absorption spectra of DOM between 200 and 900 nm (at 1 nm intervals) were acquired using a UV−vis instrument (Shimadzu, Kyoto, Japan). An Milli-Q water sample was used as the reference. The absorption spectra were baseline corrected by subtracting the mean absorbance between 690 and 700 nm. Absorption coefficients (aCDOM) were obtained using the following equation
nm, respectively. Readings were collected in ratio mode, using a 5 nm interval for excitation, a 1 nm interval for emission, and a scanning speed of 2400 nm/min. The excitation and emission band passes were both 5 nm. A water blank EEM was subtracted to eliminate Raman scattering peaks of water.22 Previous studies defined several regions of interest in EEM spectra which were related with the different chemical components. Two major categories of fluorophores, i.e., humic-like and protein-like DOM, have been found in aquatic ecosystems.23 Three commonly designated humic-like peaks are A (λEx 230−260 nm, λEm 380−460 nm), C (λEx 320−360 nm, λEm 420−480 nm), and M (λEx 290−310 nm, 370−420 nm).24 There are two types of protein-like peak, tyrosine-like B (λEx 270−280 nm, λEm 300−315 nm) and tryptophan-like (T, λEx 270−280 nm, λEm 345−360 nm).24 ESI-FT-ICR-MS Analysis. Prior to ESI FT-ICR MS analysis, each sample extract (∼100 mL) was brought to pH 2 by adding LCMS grade HCl (Sigma-Aldrich), then passed through a 500 mg Bond Elut PPL cartridge (Agilent Technologies, Santa Clara, CA, U.S.A.).4,5,25 Each cartridge was preconditioned with 20 mL LCMS grade methanol, then rinsed with three cartridge volumes of acidified water (pH 2, achieved by adding LCMS grade HCl), then rinsed with one cartridge volume of LCMS grade water. Each sample was pumped through the extraction cartridge at 5 mL/min, and concentrated to ∼1 mL under a stream of N2, then kept at −20 °C in the dark. A blank sample that had been shaken and filtered was also extracted by a cartridge, and the extract was used as a procedural blank in the FT-ICR MS analyses. The extracts were analyzed using a Bruker Apex Ultra FTICR MS (Bruker, Billerica, MA, U.S.A.) equipped with a 9.4 T superconducting magnet and an Apollo II electrospray ion source.26 Each sample extract was diluted by a factor of 5 by adding 400 μL methanol to 100 μL extract, then the diluted extract was directly injected into the ESI source at 180 μL/h using a syringe pump. The spray shield voltage, capillary column introduction voltage, and capillary column end voltage were 3.0 kV, 4 kV, and −320 V, respectively, for negative-ion ESI, and 4.0 kV, 4.5 kV, and 320 V, respectively, for positive-ion ESI. Ions were accumulated in the hexapole for 0.1 s before being transferred to the ICR cell. The mass range was m/z 150−1000. A 4 M word size was selected for the time domain signal acquisition. The signal-to-noise ratio and dynamic range were enhanced by accumulating 128 time domain FT-ICR transients. Molecular Formula Assignment. It has been found that ESI+ and ESI− target different groups of compounds, resulting
aCDOM(λ) = 2.303·A CDOM (λ)/L
where ACDOM is the absorbance of chromophore-containing DOM at wavelength λ, L is the path length of the optical cell in meters (0.01 m for our measurements), and 2.303 is the common-to-natural logarithm conversion factor. Three-dimensional fluorescence excitation−emission matrices (EEMs) of DOM were acquired using a F-7000 fluorescence spectrometer with a 700 V xenon lamp (Hitachi High-Technologies, Tokyo, Japan). The excitation and emission scanning ranges were 200−450 nm and 250−600 C
DOI: 10.1021/acs.est.6b03971 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 3. Excitation−emission matrices for samples LHG4600, TGL2, and TGL3.
fluorescence intensities of emissions at 470 and 520 nm, both obtained from excitation at 370 nm.36 A fluorescence index of 1.4 or less indicates DOM of terrestrial origin, and a fluorescence index of 1.9 or higher indicates microbe-derived material.24 All three of our samples had fluorescence indices higher than 1.9 (Table 1) indicating that the DOM in the
in different mass spectra being acquired from the same sample.27,28 We therefore used both ESI− and ESI+ to allow as many molecules as possible to be detected. The ESI FT-ICR MS was calibrated using a series of alkylcarbazoles in a crude oil. Peaks in the range m/z 200−600 with relative abundances more than 10 times the standard deviation of the baseline noise were exported to a spreadsheet. Methodologies for FT-ICR MS mass calibration, data acquisition, and processing have been described elsewhere.29 In addition, the amount or number of peaks is a function of both amount and ionization efficiency.
Table 1. Fluorescence Indices (FIs), Humification Indices (HIXs), and Biological Indices (BIXs) for DOMs Extracted from the Cryoconite Samples in This Study and from Mat Water Samples (Raw Filamentous Microbial Mats Collected with Water Samples) from Other Areas24,a
■
RESULTS AND DISCUSSION Absorbance Spectroscopy of the DOM. The UV−vis absorbance spectra of the DOM in the LHG and TGL samples show a typical shape (Figure 2), which the absorbance efficient sharply decreases from 200 to 230 nm, and increases between 230 and 360 nm, but almost no absorption above 360 nm. Absorption between 200 and 230 nm was probably caused by humic-like substances which have been found in snow samples.30 Each sample contained small peaks between 260 and 270 nm. These peaks were consistent with strong absorbance by fluorescent amino acids, but could have been caused by a wide range of other compounds that absorb in this region.24 Distinct elevated DOM absorption between 290 and 360 nm occurred in all the samples, and the absorption peaks were centered between 300 and 350 nm (Figure 2). This feature was most likely related to the presence of mycosporinelike amino acids (MAAs), which are produced by organisms (such as ascomycetous and basidiomycetous fungi, cyanobacteria, heterotrophic bacteria, and microalgae) that live in environments with high sunlight levels, such as marine environments and high-elevation lakes.31−33 Liu et al.34 found cyanobacteria in snow samples in a TP glacier, and higher organic matter contents, including living cyanobacteria, were found in cryoconite than snow samples in Qiyi Glacier in the Qilian Mountains35 and Ü rümqi Glacier No. 1 in the Tien Shan Mountains.14 Fluorescence EEMs Classification and Fluorescence Indices of the DOM. The EEM spectra of our cryoconite samples are shown in Figure 3, which primarily contained tyrosine-like and tryptophan-like peaks (B and T), but less intense at humic-like peaks (A and C). Humic-like peaks of A and C probably represent high molecular weight DOM derived primarily from vascular plants.24 Protein-like components peaks of B and T have been used to fingerprint DOM derived from microbial sources.2 The EEM spectra therefore suggest that the DOM in the cryoconite was mainly derived from microbial sources, nevertheless, some was derived from vascular plants. Fluorescence intensity ratios can be used to estimate the relative contributions of autochthonous and allochthonous organic matter.24 The fluorescence index is the ratio of the
sampling site cryoconite samples LHG4600 TGL2 TGL3 mat water samples Pozzo di Crystali Sulfide surface well Grotta Bella
FI
HIX
BIX
3.44 3.17 3.12
1.37 1.32 1.11
0.83 0.65 0.93
1.99 1.62 2.29
0.25 0.29 0.4
1.52 1.95 0.99
Note that DOM with FI ≥ 1.9 is defined as microbially-derived DOM; DOM with HIX ≤ 1.5 is fresh DOM with low degree of humification; and DOM with 0.8 ≤ BIX ≤ 1.0 is freshly produced DOM by microbial communities.
a
cryoconite samples was a result of primary production by autochthonous microbes. The biological index is the ratio of the emission intensities at 380 and 430 nm, both obtained from excitation at 310 nm.37 A biological index of 0.8−1.0 indicates freshly produced DOM of biological or microbial origin, and a biological index below ∼0.6 is considered to indicate that little autochthonous organic matter is present.24,37 The biological indices for our cryoconite samples were mostly >0.8, except that the TGL2 biological index was ∼0.7 (Table 1). These values were consistent with autochthonous DOM derived predominantly from microbes.24 A modified humification index, calculated by taking the area under an emission spectrum (acquired with excitation at 254 nm) between 435 and 480 nm and dividing it by the area under the spectrum between 300 and 345 nm plus the area under the spectrum between 435 and 480 nm, has been used to allow the relative humification levels of DOM samples to be compared.38 The cryoconite samples had humification indices
