bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1 Linking the chemistry and reactivity of dissolved organic matter from low-latitude glaciers

2 and rock glaciers

3

4

5 Fegel, Timothy1, Boot, Claudia M.1,2, Baron, Jill S.1,3, Hall, Ed K.1,4

6 1. Natural Resource Ecology Laboratory, State University, 2. Department of

7 Chemistry, Colorado State University 3. U.S. Geological Survey 4. Department of Ecosystem

8 Science and Sustainability, Colorado State University

9

10

11

12

13

14

15

16

17

18

19

20

21 Keywords: glaciers, dissolved organic matter, chemodiversity, GC-MS, DOM, molecular

22 composition, lability, bacterial growth efficiency, BGE, subsidies, heteroptrophy, alpine

23

1 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

24 Abstract

25 As glaciers in the western United States thaw in response to warming, they release dissolved

26 organic matter (DOM) to alpine lakes and streams. Biological availability of DOM from small

27 mountain glaciers is unknown. Differences in DOM bioavailability between glacier types like

28 rock and ice glaciers remains undefined, yet rock glaciers outnumber ice glacier approximately

29 ten to one at low latitudes. To assess which components of aquatic DOM are most reactive and

30 the potential for glacial DOM from low latitude glaciers to subsidize heterotrophy in alpine

31 headwaters we evaluated reactivity and molecular composition of DOM from ice glaciers and

32 rock glaciers from four paired catchments (each with a glacier and rock glacier at their

33 headwaters). Biological reactivity was linked to molecular composition by evaluating the

34 chemical characteristics of each DOM pool pre- and post-incubation using common microbial

35 community laboratory assays paired with untargeted mass spectrometry-based metabolomics.

36 Glacier and rock glacier DOM was similar in concentration and chemodiversity, but differed in

37 composition. When incubated with a common microbial community, DOM from ice glacier

38 meltwaters contained a higher proportion of bioavailable DOM (BDOM), and resulted in greater

39 bacterial growth efficiency. Differences in DOM reactivity between glacier types was

40 determined by differences in the relative abundance of only a few dozen compounds. Though

41 BDOM was lower in rock glaciers, because rock glaciers are more abundant and are expected to

42 have greater longevity, we propose that both glacial types will be important sources of

43 bioavailable DOM to alpine headwaters over the coming years to decades.

44

2 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

45 Introduction

46 Glaciers bridge the atmospheric-terrestrial-aquatic interface, integrating atmospherically

47 deposited chemicals with weathering products and releasing reactive solutes to adjacent surface

48 waters (Williams et al. 2007; Dubnick et al. 2010; Fellman et al. 2010; Stibal et al. 2010;

49 Stubbins et al. 2012). Between continental and mountain glaciers, mountain glaciers release the

50 largest flux of carbon from melting ice annually (Hood et al. 2015). Dissolved organic matter

51 (DOM) from these large mountain glaciers of the European Alps and Alaska has been shown to

52 stimulate heterotrophic respiration (Hood et al. 2009; Singer et al. 2012). Glaciers and rock

53 glaciers, each with distinct geophysical attributes, are common in many of the alpine headwaters

54 of the western U.S. (Fegel et al. 2016). Glaciers are massive ice bodies that form and persist in

55 areas where annual snow accumulation is greater than annual snow ablation at decadal or longer

56 time spans. Rock glaciers are flowing bodies of permafrost, composed of coarse talus and

57 granular regolith both bound and lubricated by interstitial ice (Berthling, 2011). In the western

58 United States, rock glaciers are an order of magnitude more abundant than ice glaciers and are

59 more resistant to warming temperatures than ice glaciers (Fegel et al. 2016). Yet little is known

60 about the quantity or quality of DOM being released from smaller glaciers or rock glaciers in

61 mountain headwater ecosystems or their potential to affect low-latitude alpine ecosystems.

62 The chemical composition of DOM from natural aquatic systems is complex (Hedges et al.

63 2000; Kim et al. 2006; Hockaday et al. 2009), and glacier and rock glacier meltwaters are not

64 likely to be an exception. Previously, assessment of the bioavailability of DOM relied on

65 bioassays that measure the rate and amount of DOM consumed over time (e.g. Amon and Benner

66 1996; del Giorgio and Cole 1998; Guillemette and del Giorgio 2011). While some coarse

67 characterizations of molecular composition (Benner 2002; Berggren and del Giorgio 2015) or the

3 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

68 lability of individual compounds (del Giorgio and Cole 1998) have previously been applied to

69 characterize bioavailable DOM (BDOM), only recently has the research community begun to

70 apply high-resolution analytical chemistry to assess the molecular characteristics of

71 environmental DOM (e.g. Kellerman et al. 2014, 2015; Andrilli et al. 2015). Whereas no single

72 method can identify the entire spectrum of compounds present in an environmental DOM pool

73 (Derenne and Tu 2014), mass spectrometry allows for molecular identification of thousands of

74 specific DOM compounds, simultaneously yielding a more specific and broader metric for DOM

75 analysis compared to spectroscopic techniques such as fluorescence or ultraviolet absorption.

76 Simultaneous identification and quantification of total BDOM is not an easy task. Difficulties

77 in experimentally connecting molecular characterization of DOM pools to their bioavailability

78 are partly due to the highly diverse constituent compounds that compose natural DOM (Derenne

79 and Tu 2014). The effects of individual compounds identified to be bioavailable (e.g. amino

80 acids in glacial meltwaters (Fegel et al. 2016) may not be representative of the total BDOM pool.

81 It is also possible that DOM with a high diversity of labile compounds may have positive

82 feedbacks on the bioavailability of the total DOM pool by asserting priming effects on the total

83 pool (Guenet et al. 2010). In addition, certain metabolites within DOM pools may not be

84 bioavailable individually, but may act as cofactor metabolites that allow for a mutualistic

85 increase in bioavailability [Hilker 2014]. Thus, the relationship between DOM pool

86 characteristics and lability remains largely unknown in most natural systems. However, some

87 patterns are beginning to emerge, like the consistently high lability of proteinaceous DOM that

88 has been found in glacial, estuarine, and marine environments [Andrilli et al. 2015]. Defining

89 these fundamental relationships between chemical composition of DOM and lability has the

4 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

90 potential to provide a better understanding for how DOM pools contribute to heterotrophy across

91 a broad spectrum of aquatic ecosystems.

92 There are defining characteristics in the origin of DOM that differentiate glaciers from rock

93 glaciers. DOM derived from ice glaciers is structured by in situ microbial activity and can be an

94 important source of chemical energy to headwater ecosystems (Hood et al. 2009; Singer et al.

95 2012; Fellman et al. 2015; Fegel et al. 2016). DOM derived from rock glaciers is a

96 conglomeration of carbon compounds percolated through from vegetation growing on the rock

97 glacier surface and microbial processing within the rock glacier itself (Wahrhaftig and Cox,

98 1959; Williams et al. 2007; Fegel et al. 2016). Previous research has shown that DOM released

99 from glaciers and rock glaciers in the western United States differ in their optical properties

100 (Fegel et al. 2016). Whether or not these differences in the optical properties of DOM between

101 glacial types translates to differences in their reactivity or proportion of BDOM remains

102 unknown.

103 To address this, we asked whether differences in the composition of DOM between ice

104 glaciers and rock glaciers affected differences in microbial metabolism and whether DOM

105 chemistry in ice and rock glacier meltwater in the western United States is similar to what has

106 been reported for other glacial meltwaters. Here we present the results of laboratory incubations

107 of DOM from ice glacier and rock glacier meltwaters with a common microbial community.

108 Incubations were bookended (i.e. analyzed before and after incubation) with non-targeted

109 metabolomic analysis of DOM via gas chromatography mass spectrometry (GC-MS) to

110 determine differences in the specific chemical compounds metabolized by microbial processing.

111 By exposing DOM from different sources to a common microbial community we were able to

5 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

112 assess the lability of each DOM source and analyze which differences in the molecular

113 composition of DOM affected its lability and carbon use efficiency.

114 Methods

115 Site Description

116 Paired ice glaciers and rock glaciers within four watersheds on the Front Range of Northern

117 Colorado were selected based on their size (>0.5km2) and the proximity to each other within the

118 watershed, forming pairs of features with similar geographic parameters (Figure 1). We collected

119 samples of glacier meltwater in the late summer to capture the greatest contribution of ice melt

120 and minimize annual snowmelt contribution. We sampled the four pairs in September 2014.

121 Sample sites were: Isabelle Glacier (-105.640994,40.063373) and Navajo Rock Glacier (-

122 105.636092,40.061200), (-105.646351,40.023378) and Arapaho Rock Glacier

123 (-105.637699, 40.022482), Peck Glacier (-105.663810,40.068332) and Peck Rock Glacier (-

124 105.664310,40.071642) in the Indian Peaks Wilderness west of Boulder, CO; and Andrews

125 Glacier (-105.680639,40.288370) and Taylor Rock Glacier (-105.671197,40.275568) in the Loch

126 Vale Watershed in Rocky Mountain National Park (Figure 1a). A complete site description for

127 each site is given in Fegel et al. (2016).

128 Field extraction of DOM

129 At each glacier, meltwaters were collected in the early to mid-morning (0500-1000) to

130 minimize diurnal variability in ice melt from solar radiation. DOM was extracted from 20 L of

131 meltwater collected at the terminus of each feature in the field using a slightly modified protocol

132 established by Dittmar et al. (2008, Supplemental Information). Briefly, meltwater samples were

133 passed through pre-combusted (450° C, 5hr) Whatman GF/F filters (GE Whatman, Pittsburg,

134 PA, USA), acidified to ~ pH 2 with 32% HCl, and concentrated using Bond Elut PPL carbon

6 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

135 extraction cartridges (Agilent, Santa Clara, CA, USA). Concentrated DOM was eluted in the

136 field with 10mL HPLC grade methanol per cartridge into cleaned, combusted, and pre-weighed

137 120mL borosilicate bottles.

138 Metabolite Analyses

139 Once in the laboratory, pre-incubation DOM samples were dried under a clean N2 stream and

140 weighed (http://www.nrel.colostate.edu/projects/lvws/data.html). Each sample (n=8) was

141 prepared for metabolomic analysis by dissolving the dried OM into fresh HPLC-grade methanol

142 to a final concentration of 2 mg ml-1.

143 After the microbial bioassays (in which all samples were run in duplicate), post-incubation

144 DOM samples were collected separately from each individual microcosm (n=16). Water from

145 each microcosm was filtered through pre-leached 0.2 μm Millipore filters (EMD Millipore,

146 Billerica, MA, USA) to remove microbial biomass, freeze-dried, and the total remaining DOM

147 was weighed and re-dissolved into HPLC-grade methanol

148 (http://www.nrel.colostate.edu/projects/lvws/data.html). In order to increase the volatility of

149 molecules analyzed through GC-MS, samples were derivatized with trimethylsilane (TMS) using

150 standard protocols (Supplemental Materials) (Pierce, 1968).

151 Metabolomics

152 Both pre- and post- incubation OM samples were analyzed with inline gas chromatography-

153 mass spectroscopy (GC-MS) at the Proteomics and Metabolomics Facility at Colorado State

154 University. Metabolites were detected using a Trace GC Ultra coupled to a Thermo ISQ mass

155 spectrometer (Thermo Scientific, Waltham, MA, USA). Samples were injected in a 1:10 split

156 ratio twice in discrete randomized blocks. Separation occurred using a 30 m TG-5MS column

157 (0.25 mm i.d., 0.25 μm film thickness, Thermo Scientific, Waltham, MA, USA) with a 1.2 mL

7 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

158 min-1 helium gas flow rate, and the program consisted of 80°C for 30 sec, a ramp of 15°C per

159 min to 330°C, and an 8 min hold. Masses between 50-650 m/z were scanned at 5 scans sec-1 after

160 electron impact ionization (Broeckling et al. 2014).

161 DOM lability experiments

162 Concentrated DOM samples from each of the eight study sites were diluted to 4 mg L-1 C and

163 incubated in vitro with a natural microbial community collected from The Loch, a small sub-

164 alpine lake in Rocky Mountain National Park, CO, USA, (-105.6455, 40.2976). Unfiltered lake

165 water collected from The Loch was aged for 2 years at 5 °C in order to remove the majority of

166 the bioavailable carbon. At the start of the experiment DOM concentration of the lake water was

167 0.7 mg L-1 C. At the initiation of the incubations 2 L of aged lake water was filtered through a

168 pre-combusted (450° C, 5hr) Whatman GF/C filter (1.2μm nominal pore size) (GE Whatman,

169 Pittsburg, PA, USA) to remove bacterial grazers (e.g. protists and metazoans). Three aliquots of

170 filtered-aged lake water was preserved with 2% formalin (37% formaldehyde), and set aside for

171 enumeration of bacteria at the initiation of the experiment (i.e. t=0), and a second set of three

172 aliquots (7mL) was used for initial DOC/TN analysis on a Shimadzu TOC-VWS analyzer

173 (Shimadzu Corp., Kyoto, Japan). To create normalized concentrations (4 mg L-1 C), in each

174 incubation bottle we added 60.96 mL of unfiltered Loch water, between 3.80 and 9.06mL of

175 concentrated DOC solution (depending on the initial concentration), and filled to 70mL total

176 volume with DI; resulting in standardized concentrations of 4 mg L-1 C in each incubation bottle.

177 During the experiment, microcosms that received DOM from ice glacier and rock glacier sites

178 were incubated alongside control incubations that contained lake water and the common

179 microbial community but with no added DOM (i.e. experimental control). In addition, an

8 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

180 analytical blank of MilliQ water was incubated to correct for any instrumental drift that occurred

181 during the experiment.

182 All microcosms were incubated simultaneously at 15 °C for 10 weeks. To calculate microbial

183 respiration, we measured changes in dissolved oxygen (DO) at 1-minute intervals in each

184 microcosm using an Oxy-4 fiber-optic dissolved oxygen probe (PreSens, Regensburg, Germany).

185 The incubation was terminated when the fastest metabolizing microcosm approached 4mg L-1

186 DO to avoid hypoxia and the potential for anaerobic metabolism. All measurements with

187 amplitude less than 20000 amps were removed because of the potential for inaccurate readings of

188 DO. Absolute values from the raw fiber optic measurements were corrected for analytical drift

189 by subtraction of changes in signal from the MilliQ water analytical control over the course of

190 the experiment.

191 Bacterial Cell Counts and Bulk Chemistry

192 From each microcosm we collected a 2 mL aliquot post-incubation and preserved it with 2%

193 formalin (final concentration) to assess changes in cell abundance during the course of the

194 incubation. Aliquots were filtered onto 0.2 μm Millipore polycarbonate filters (EMD Millipore,

195 Billerica, MA, USA) and stained with Acridine Orange (Hobbie et al. 1977, Supplemental

196 Information) for enumeration. Post-incubation we assessed TOC and TDN

197 (http://www.nrel.colostate.edu/projects/lvws/data.html) for each microcosm. To ensure that

198 filtration did not contribute significant additional carbon to the dissolved carbon pool, a subset of

199 blank Millipore filters flushed with MilliQ was analyzed for TOC and TN. All filtration of

200 microcosm water for chemical analysis was conducted on pre-leached Millipore filters that

201 contributed insignificant amounts of carbon to the filtrate

202 (http://www.nrel.colostate.edu/projects/lvws/data.html).

9 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

203 Data Analysis

204 Metabolite Analysis

205 Compound annotation was prioritized based on order of normalized intensity and statistical

206 difference between glacier types. Putative identification (candidate compounds based on

207 physicochemical characteristics and spectral similarity with spectral libraries (Sumner et al.

208 2007), was made using spectral matching to in-house, NISTv12, Golm, Metlin, and Massbank

209 metabolite databases. Clustered features assigned candidate compounds in our study had high

210 similarity (>90%) to spectra from known standard compounds within the databases used

211 (Supplemental Information). Molecular rank was calculated by ordering candidate compounds

212 by their normalized ion intensity for pre- and post- incubation sample averages for all features.

213 Chemical diversity was calculated using the Shannon-Wiener diversity index (Shannon 1948) by

214 treating each unique chemical compound identified though GC-MS as a ‘species’. This was done

215 for DOM composition both before and after incubation in order to estimate changes in chemical

216 diversity through microbial metabolism.

217 Analysis of DOM Reactivity

218 Oxygen consumption rates were fit to a Berner-Multi-G two-pool decay model to estimate the

219 size of the labile pool (BDOM) and the recalcitrant pool. Oxygen consumption was averaged for

220 each glacier type and confidence intervals were calculated at α=0.05. Data were smoothed using

221 a third order polynomial (R2>0.999) and 95% confidence intervals were plotted. Berner’s Multi-

222 G model was used to model carbon pool bioavailability (Berner 1980; Guillemette and del

223 Giorgio 2011), using SAS. Dissolved oxygen curves generated from the incubation were fit to

224 the Equation:

225

10 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

푘푡 226 푌 = 퐵1 + 퐵0

227

228 where Y is the total carbon pool, B1 is the bioavailable carbon pool, k is the decay rate

229 constant of the bioavailable pool, t is time, and B0 is the recalcitrant carbon pool. We used a least

230 square means to test for statistical differences between ice glaciers and rock glaciers in the size

231 of the B1 and B0 pools. Total C consumed was calculated as the difference in pre- and post-

232 incubation DOC values.

233 To address the metabolic quality of consumed carbon we calculated Respiratory Quotient

-1 -1 234 (RQ) as the carbon consumed (in mg L ) divided by the oxygen (as O2) consumed (in mg L ).

235 Bacterial growth efficiency (BGE) was calculated to examine how efficiently each carbon

236 source was incorporated into bacterial biomass. To calculate BGE we divided bacterial

237 production (BP) by the sum of bacterial respiration and bacterial production (BR + BP)(del

238 Giorgio and Cole 1998). Thus, BGE is a ratio of carbon incorporated into biomass relative to

239 total carbon consumed for the bacterial community. BP was measured as change in cell number

240 over time and converted to units C using an estimate of 20 fg C per bacterial cell (Borsheim and

241 Bratbak 1987) as a conversion. Carbon consumption rate was measured as carbon consumed

242 during the course of the incubation by measuring DOC content in each filtered pre- and post-

243 incubation sample using a TOC/TDN analyzer.

244

245

246

247

248

11 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

249 Statistical Analyses

250 The DOM composition for each treatment (glaciers vs. rock glaciers) was compared with

251 ANOVA for each compound using the aov function in R, and p-values were adjusted for false

252 positives using the Bonferroni-Hochberg method in the p.adjust function. Post-incubation

253 samples were corrected for compounds added by the lake water by subtracting the peak

254 intensities for each chemical candidate within the lake water (i.e. experimental control) from

255 each post incubation sample.

256 We identified differences in compounds present in each sample using PCA conducted on

257 mean-centered and pareto variance-scaled data using the pcaMethods package in R. We analyzed

258 GC-MS spectra using principal component analysis (PCA) in R (R Core Team, 2014). For each

259 sample, raw data files were converted to .cdf format, and a matrix of molecular features as

260 defined by retention time and mass to charge ratio (m/z) was generated using XCMS package in

261 R for feature detection and alignment. Raw peak areas were normalized to total ion signal,

262 outlier features were detected based on total signal and PC1 of PCA, and the mean area of the

263 chromatographic peak was calculated among replicate injections (n=2). We grouped all spectral

264 features based on an in-house clustering tool, RAMClustR, which uses spectra based coelution

265 and covariance across the full dataset groups features (Broeckling et al. 2014). We used the t-test

266 configured for non-parametric Welch-Satterthwaite test to compare differences in RQ, BGE,

267 chemodiversity, and loss of chemodiversity between glacier types. Our use of a Welch-

268 Satterthwaite test allowed for comparison of samples of unequal variance and distribution.

269

270

271

12 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

272 Results

273 Ice glacier and rock glacier DOM composition

274 Neither the chemical diversity (mean=2.82), nor the C:N ratio (mean=2.1) differed between

275 glaciers and rock glaciers (Table 1). In addition, the molecular composition was similar in each

276 glacier type for the top 25 most abundant molecules. Combined, the glacier and rock glacier

277 DOM molecular composition consisted of 2033 compounds. Each individual compound

278 consisted of 3 to170 individual mass spectral features, for a total of 14,571 unique mass spectral

279 features in the complete dataset. Of the 2033 compounds, 328 were chosen for annotation based

280 either on high ion intensity (n=278) or significantly different relative abundances between glacier

281 and rock glacier sources (n=33). Annotated compounds consisted primarily of simple sugars,

282 amino acids and other organic acids (http://www.nrel.colostate.edu/projects/lvws/data.html).

283 There were significant differences between glacier types in the molecular composition of less

284 abundant compounds (Table 2), although the C:N ratio, chemical diversity, and compounds with

285 greatest ion intensity were similar. PCA analysis of the complete dataset suggested 33 key

286 compounds, ordinated on axes three and five of our PCA analysis (Figure 2A), were responsible

287 for differences in the ice and rock glacier DOM pools (Table 2). Of those 33 compounds, 5 were

288 assigned putative structures (Table 2, also see

289 http://www.nrel.colostate.edu/projects/lvws/data.html) with ice glacier meltwaters significantly

290 enriched in the simple sugar maltose and the amino acid glutamate, and rock glacier meltwaters

291 enriched in the primary organic acids, glycolate, threonate and quinate (Table 2). The other 28

292 compounds represented components of the DOM pool that differentiated DOM composition

293 from each glacier type, but were not present in the databases we employed and had no known

294 structural analogs.

13 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

295 Incubations

296 The bacterial respiration rates of DOM from each glacier type were not significantly different,

297 as estimated by the decay constant k (Table 3). However, a significantly larger portion of ice

298 glacier DOM was bioavailable, (i.e. BDOM, B1 = 58.8 ± 9.7%) compared to rock glacier DOM

299 (B1= 37.3 ± 10.2%, p <0.01, Table 3). Bacterial growth efficiency (BGE) was higher for

300 microbial communities incubated with ice glacier DOM compared to rock glacier DOM (G =

301 0.26 ± 0.13, RG = 0.16 ± 0.16, Table 3). In general, more oxygen was consumed per mg organic

302 carbon metabolized in incubations that contained glacial DOM compared to those that contained

303 rock glacier DOM, even though the amount of carbon consumed between treatments was similar

304 (Figure 4, Table 3). This resulted in a lower respiratory quotient (RQ) for glacial derived DOM

305 compared to rock glacier derived DOM overall.

306 Post-incubation DOM analysis

307 Analysis of DOM after the incubation period allowed us to assess how microbial metabolism

308 altered the composition of DOM. Incubation with a common microbial community both rarified

309 and homogenized DOM between glacier types, resulting in fewer compounds with high ion

310 intensities, and more compounds with low ion intensities (Figure 3). This reorganized the

311 molecular rank-abundance curve, resulting in a different set of compounds with the highest

312 relative abundance in the pre vs. post-incubation dataset (Table 4). However, similar to pre-

313 incubation DOM composition, post-incubation ice glacier and rock glacier derived DOM shared

314 the same abundant compounds (Figure 2B)

315 (http://www.nrel.colostate.edu/projects/lvws/data.html). Interestingly, the chemical diversity of

316 DOM between glacier types diverged during the course of the incubation. At the end of the

317 incubation glacial DOM had significantly higher diversity and rock glacier DOM had

14 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

318 significantly lower diversity compared to pre-incubation DOM for the same glacial type (Table

319 1, Figure 2C). Combined with molecular rank abundance data (ordered on ion intensity), this

320 change in chemical diversity resulted in an increase in the richness of glacier DOM pools by

321 microbial metabolism and a decrease in richness in rock glacier DOM during the incubation. For

322 both DOM sources, many of the organic acids and sugars present pre-incubation were consumed

323 while amino acids that were present at lower intensities in pre-incubation DOM pools increased

324 in relative abundance in post-incubation DOM pools. While there were differences in the

325 chemical diversity of rock and ice glacier derived DOM post-incubation, PCA suggested were no

326 differences in the overall molecular composition between glacier types (Figure 2B). This trend

327 did not follow for all rock glaciers, as one replicate of Taylor Rock Glacier (lowest left point,

328 Figure 2A & 2B) saw few changes in its principal components after the incubation. Interestingly,

329 while no significant difference was seen in GC-MS compounds between glacier type post

330 incubation (Figure 2B), both replicates of Peck Rock Glacier (lower rightmost points Figure 2B)

331 showed separation along the y-axis (PC1) of their PCA, a separation that was not seen before

332 microbial metabolism (Figure 2A). The reasons for this remain unclear, but may be due

333 geographical separation (climatological differences) between Peck Rock Glacier and all other

334 sites.

335 Discussion

336 Our results demonstrate that chemically complex DOM released from Rocky Mountain ice

337 glaciers and rock glaciers stimulated bacterial respiration and productivity. Each DOM source

338 had unique molecular characteristics that reflected their quality as an energy source for microbial

339 metabolism. Glacier and rock glacier DOM shared many of the same organic compounds, but

340 differences in the relative concentrations of just 33 compounds appeared to drive differences in

15 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

341 the bioavailability of DOM between glacier types. Enrichment in simple sugars of glacier DOM

342 contributed to higher bioavailability of glacier DOM as measured by size of the BDOM pool and

343 bacterial growth efficiency (BGE). The DOM from rock glaciers, was enriched in less

344 bioavailable, primary organic acids. In spite of these differences, the bacteria homogenized

345 DOM from both sources during the incubations, resulting in more similar DOM composition at

346 the end of the incubation compared to the composition of the DOM before the incubation. In our

347 experiment, the source of the DOM affected its composition and reactivity and interaction with

348 the microbial community metabolism re-structured its chemistry.

349 Glaciers are a source of labile DOM

350 The results from our study of glaciers in Colorado showed similar patterns to previous studies

351 of glacial DOM in the Arctic and the European Alps, where bioavailable carbon from glaciers

352 also stimulated microbial production (Hood et al. 2009; Singer et al. 2012; Fellman et al. 2015).

353 Carbon concentrations in meltwaters from ice and rock glaciers in our study were low (0.5- 1.5

354 mg C L- 1), but similar to the global average of 0.97 mg C L-1 that has been previously reported

355 for glaciers (Dubnick et al. 2010; Stubbins 2012; Singer et al. 2012; Hood et al. 2015). The

356 percentage of DOM that was bioavailable (BDOM) from ice glacier DOM in our study (~50%)

357 was comparable to values of BDOM seen in DOM derived from glaciers in the European Alps

358 (58%), and within the range of values reported from glaciers in the Alaskan Range (23-66%)

359 (Hood et al. 2009; Singer et al. 2012). In all of these studies, including the results reported here,

360 glaciers had higher BDOM concentrations than other surficial freshwaters (Volk et al. 1997).

361 Previous reports on freshwater noted 16.5-34.5% of BDOM, which more closely resembled the

362 amount of BDOM we report from rock glacier BDOM (37% on average).

16 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

363 We propose that ice glaciers have higher proportions of BDOM compared to rock glaciers for

364 two reasons. First, rock glaciers host mosses, lichens, and vascular plants, including woody

365 shrubs and trees (Wahrhaftig and Cox, 1959; Cannone and Gerdol 2003; Burga et al. 2004).

366 Dissolved organic compounds released from these allochthonous phototrophs include a wide

367 variety of complex organic acids and polymers that are potentially difficult for microbes in

368 freshwaters to metabolize, resulting in a higher proportion of recalcitrant DOM within rock

369 glaciers (Wetzel 1992, Rovira and Vallejo 2002). The potential for the development of protosoils

370 within the rock glacier matrix may promote the presence of liquid pore water. DOM from rock

371 glaciers may already be partially processed by microbes within this protosoil environment by the

372 time it reaches the sub-rock-glacial surface water. This pre-processing may result in DOM with a

373 level of lability more similar to what is delivered to inland surface waters and derived from the

374 soil SOM matrix (Fellman et al. 2010). Conversely, BDOM in ice glaciers comes from

375 autocthonous sources (i.e. microbial production and processing on the ice surface and within the

376 ice interstitial space (Fellman et al. 2009) and atmospherically deposited aerosols (Hood et al.

377 2012), both of which appear to be more biologically available than terrestrial sources. Ice glacier

378 BDOM may be preserved in the ice matrix and physically inaccessible to microbial degradation.

379 In contrast, rock glacier DOM may be in contact with intraglacial sediments and liquid water,

380 allowing for further processing through microbial metabolism. In addition, sub-glacial

381 environments below ice glaciers are often anoxic (Tranter et al. 2005), allowing some

382 compounds to remain in a chemically reduced state until released with ice melt. This would

383 allow for rapid metabolism of these energetically-preferable, reduced compounds once unlocked

384 from the ice and released to an oxic environment (Hood et al. 2009). Thus the production of

385 DOM within each feature differs and pathways by which that DOM is delivered to the glacial

17 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

386 meltwater differ. These mechanisms would result in the differences in the proportion of BDOM

387 we observed in our bioassays of DOM lability.

388 The molecular structure of glacial DOM from large glaciers has been mechanistically linked to

389 its lability and shown to be more labile than marine or freshwater derived material due to the

390 presence of protein and amino-sugar components of microbial origin (Andrilli et al. 2015). This

391 is consistent with studies from other inland waters that show that higher DOM bioavailability

392 correlates with enrichment of amino acids and simple sugars (Lafreniere and Sharp 2004;

393 Williams et al. 2007; Dubnick et al. 2010). Taken together with the results presented here, we

394 suggest that DOM enriched in amino acids and simple sugars may be a key trait that defines

395 BDOM and be characteristic of glacial meltwaters worldwide.

396 Glacier DOM Diversity

397 The diversity of compounds in organic matter released from glaciers in this study was similar

398 to the DOM diversity found in other freshwater systems (Dubnick et al. 2010, Guillemette and

399 del Giorgio 2011, Kellerman et al. 2014). We found that chemical diversity was very similar

400 between ice glaciers and rock glaciers. However, the diversity of DOM between glacial types

401 was altered by microbial metabolism in different ways when incubated. Chemical diversity in

402 incubations with glacial DOM increased during the course of the incubation, whereas chemical

403 diversity decreased in incubations with rock glacier DOM. This suggests that when BDOM is a

404 smaller fraction of total DOM, microbial communities may use previously processed DOM to

405 fuel their own heterotrophic metabolism. This would reduce chemical diversity through the

406 consumption of primary metabolites (i.e. simple organic acids).

407 Our study expands the understanding of DOM complexity in inland waters by assessing the

408 glacial contribution of DOM to alpine headwaters and moving beyond broad functional groups

18 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

409 and compound classes identification to identification of specific compounds (Dubnick et al.

410 2010; Fellman et al. 2010; Singer et al. 2012, Fegel et al. 2016). By using a mass spectrometry

411 based technique we were able to identify individual candidate compounds through metabolomics

412 paired with total chemical pool complexity. Our results are more detailed, but consistent with

413 previous analyses that used a fluorescence index (Fegel et al. 2016). Results from Fegel et al.

414 (2016) showed higher fluorescence index (FI) values in ice glacier meltwaters compared to that

415 of rock glaciers, indicating a higher presence of proteinaceous (i.e. nitrogen containing)

416 compounds in ice glacier meltwaters. Here, using metabolomics, we identified higher

417 proportions of specific amino acids within ice glacier meltwaters compared to rock glaciers that

418 are likely the cause of the increased FI values observed (Fegel et al. 2016). Application of a

419 community level analytical approach such as mass spectrometry-based metabolomics provides

420 the opportunity to disentangle the critically important components of complex DOM pools for

421 aquatic C cycling, however only a few studies have used metabolomic approaches to address the

422 bioavailability of DOM as a heterotrophic subsidy (see Kujawinski 2011 for a review; Logue et

423 al. 2015).

424 Whereas our results provide unambiguous evidence that differences in bioavailability are due

425 to chemical differences between DOM that differ in origin from glacier types, some components

426 of the DOM pool are likely not assessed by the GC-MS we employed. GC-MS measures

427 molecules up to 1200 amu with electron impact ionization, leaving larger molecules unidentified.

428 Further, some of the components important for glacial C processing could not be identified using

429 the most current databases. While we were unable to determine exactly what those compounds

430 were, we are able to say what they were not. Many of the unannotated compounds that were

431 significantly different between glaciers and rock glaciers had m/z (mass to charge ratio) values

19 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

432 greater than 200 (Table 2), indicative of secondary metabolites with large molecular weights

433 relative to primary metabolites, most of which have mass to charge (m/z) values between 60 and

434 205. Products of primary metabolism, e.g. amino acids, organic acids, sugars and peptides are

435 well represented in mass spectrometry databases and therefore more easily identified with a GC-

436 MS approach. Less well represented are larger molecular weight products of secondary

437 metabolism such as terpenes, alkaloids, polyketides, aromatic structures, and products of mixed

438 biosynthesis. Known metabolites are often a small portion of data obtained through mass

439 spectrometry (<10%), with much of the data reflecting unknown metabolites or those yet to be

440 verified with standards (Jansson et al. 2009). Yet, the quality of mass spectrometry databases and

441 representation is rapidly improving, and this will likely be a critically important source of

442 information for understanding the relationship between molecular composition of DOM and its

443 lability in future research. The high proportion of compounds that could not be annotated with

444 candidate assignments in our study reflects the infancy of using mass spectrometry databases in

445 environmental applications. It exposes the need for more environmentally derived spectra and

446 standards from secondary metabolism to be added to the current metabolite databases. With the

447 expansion of metabolic databases, and the verification of environmentally-derived spectra to

448 known standard compounds, metabolomic techniques will provide the ability to directly assess

449 the functionality of in situ metabolic pathways from natural systems.

450 Distribution of Glaciers and Rock Glaciers

451 Complete mapping of rock glaciers has only been completed for the contiguous United States

452 and portions of South America, yet early results suggest rock glaciers may be exceptionally more

453 abundant than ice glaciers in headwater ecosystems (Falaschi et al. 2013; Rangecroft et al. 2015;

454 Fegel et al. 2016). In addition to their ubiquity, rock glacier derived DOM may contribute to

20 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

455 ecosystem productivity for much longer than ice glaciers due to the slower recession of rock

456 glaciers compared to ice glaciers (Woo 2012). At similar carbon concentrations and with only

457 ~15% less BDOM (~37% in rock glaciers compared to 52% in glacier derived DOM), as

458 observed in our study, rock glaciers may play a role in heterotrophic metabolism of alpine

459 headwaters. Current glacial carbon modeling neglects the contribution of rock glacial carbon

460 (Hood et al. 2015), yet our results suggest that rock glaciers supply a significant source of

461 organic matter for metabolism and will continue to do so for decades to come.

462 In conclusion, the work presented here suggests that as long as the cryosphere is present,

463 DOM entering alpine streams and lakes from both glacier and rock glaciers is likely to continue

464 to contribute to ecosystem heterotrophy during summer melt. We found clear differences in the

465 proportion of BDOM and microbial growth efficiency between glacier and rock glaciers DOM.

466 This suggests a higher potential for secondary production in systems with glacial headwaters

467 compared to those with rock glacier headwaters. The applicability of metabolomics to the

468 analysis of natural DOM provides the potential to expand the application of our results for the

469 prediction of BDOM beyond glaciated ecosystems to a broad range of inland waters. At the same

470 time our approach exposes the need for better metabolite database development for ecological

471 metabolomic approaches. In the coming decades, the DOM inputs to alpine lakes and streams

472 will increasingly be dominated by rock glacier-like DOM inputs, as ice glaciers are lost and rock

473 glaciers continue to contribute to alpine hydrology due to their slower ablation rates. Despite

474 differences in chemical compounds and BDOM pool sizes between glacier types, both glaciers

475 and rock glaciers represent an important heterotrophic subsidy of organic carbon to alpine

476 headwaters that will fuel ecosystems processes from the bottom up for years to come.

477

21 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

478 References

479 Amon, Rainer MW, and Ronald Benner. "Bacterial utilization of different size classes of

480 dissolved organic matter." Limnology and Oceanography 41.1 (1996): 41-51.

481

482 Andrilli, Juliana, et al. "An ultrahigh‐resolution mass spectrometry index to estimate natural

483 organic matter lability." Rapid Communications in Mass Spectrometry 29.24 (2015): 2385-2401.

484

485 Benner, Ronald. "Chemical composition and reactivity." Biogeochemistry of marine dissolved

486 organic matter 3 (2002): 56-90.

487

488 Berggren, Martin, and Paul A. Giorgio. "Distinct patterns of microbial metabolism associated to

489 riverine dissolved organic carbon of different source and quality." Journal of Geophysical

490 Research: Biogeosciences (2015).

491

492 Berner, Robert A. Early diagenesis: A theoretical approach. No. 1. Princeton University Press,

493 1980.

494

495 Berthling, Ivar. "Beyond confusion: Rock glaciers as cryo-conditioned

496 landforms." Geomorphology 131.3 (2011): 98-106.

497

498 Borsheim, Knut Yngve, and Gunnar Bratbak. "Cell volume to cell carbon conversion factors for

499 a bacterivorous Monas sp. enriched from seawater." Mar. Ecol. Prog. Ser 36.17 (1987): ll.

500

22 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

501 Broeckling, Corey David, et al. "RAMClust: A Novel Feature Clustering Method Enables

502 Spectral-Matching-Based Annotation for Metabolomics Data."Analytical chemistry 86.14

503 (2014): 6812-6817.

504

505 Burga, Conradin A., et al. "Vegetation on Alpine rock glacier surfaces: a contribution to

506 abundance and dynamics on extreme plant habitats." Flora-Morphology, Distribution,

507 Functional Ecology of Plants 199.6 (2004): 505-515.

508

509 Cannone, Nicoletta, and Renato Gerdol. "Vegetation as an ecological indicator of surface

510 instability in rock glaciers." Arctic, Antarctic, and Alpine Research 35.3 (2003): 384-390.

511

512 Del Giorgio, Paul A., and Jonathan J. Cole. "Bacterial growth efficiency in natural aquatic

513 systems." Annual Review of Ecology and Systematics (1998): 503-541.

514

515 Derenne, Sylvie, and Thanh Thuy Nguyen Tu. "Characterizing the molecular structure of organic

516 matter from natural environments: An analytical challenge." Comptes Rendus Geoscience 346.3

517 (2014): 53-63.

518

519 Dittmar, Thorsten, et al. "A simple and efficient method for the solid‐phase extraction of

520 dissolved organic matter (SPE‐DOM) from seawater." Limnology and Oceanography:

521 Methods 6.6 (2008): 230-235.

522

23 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

523 Dubnick, Ashley, et al. "Characterization of dissolved organic matter (DOM) from glacial

524 environments using total fluorescence spectroscopy and parallel factor analysis." Annals of

525 Glaciology 51.56 (2010): 111-122.

526

527 Falaschi, Daniel, et al. "First glacier inventory and recent changes in glacier area in the Monte

528 San Lorenzo region (47 S), Southern Patagonian Andes, South America." Arctic, Antarctic, and

529 Alpine Research 45.1 (2013): 19-28.

530

531 Fegel, T. S., Baron, J. S., Fountain, A. G., Johnson, G. F., & Hall, E. K. (2016). The differing

532 biogeochemical and microbial signatures of glaciers and rock glaciers. Journal of Geophysical

533 Research: Biogeosciences.

534

535 Fellman, Jason B., Eran Hood, and Robert GM Spencer. "Fluorescence spectroscopy opens new

536 windows into dissolved organic matter dynamics in freshwater ecosystems: A

537 review." Limnology and Oceanography 55.6 (2010): 2452-2462.

538

539 Fellman, Jason B., et al. "The impact of glacier runoff on the biodegradability and biochemical

540 composition of terrigenous dissolved organic matter in near-shore marine ecosystems." Marine

541 Chemistry 121.1 (2010): 112-122.

542

543 Fellman, Jason B., et al. "Evidence for the assimilation of ancient glacier organic carbon in a

544 proglacial stream food web." Limnology and Oceanography (2015).

545

24 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

546 Guenet, Bertrand, et al. "Priming effect: bridging the gap between terrestrial and aquatic

547 ecology." Ecology 91.10 (2010): 2850-2861.

548

549 Guillemette, François, and Paul A. del Giorgio. "Reconstructing the various facets of dissolved

550 organic carbon bioavailability in freshwater ecosystems. “Limnology and Oceanography 56.2

551 (2011): 734-748.

552

553 Hedges, John I., et al. "The molecularly-uncharacterized component of nonliving organic matter

554 in natural environments." Organic Geochemistry 31.10 (2000): 945-958.

555

556 Hilker, Monika. "New synthesis: parallels between biodiversity and chemodiversity." Journal of

557 chemical ecology 40.3 (2014): 225-226.

558

559 Hobbie, J. El, R. Jasper Daley, and STTI977 Jasper. "Use of nuclepore filters for counting

560 bacteria by fluorescence microscopy." Applied and environmental microbiology 33.5 (1977):

561 1225-1228.

562

563 Hockaday, William C., et al. "Electrospray and photoionization mass spectrometry for the

564 characterization of organic matter in natural waters: a qualitative assessment." Limnology and

565 Oceanography: Methods 7.1 (2009): 81-95.

566

567 Hood, Eran, et al. "Glaciers as a source of ancient and labile organic matter to the marine

568 environment." Nature 462.7276 (2009): 1044-1047.

25 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

569

570 Hood, Eran, et al. "Storage and release of organic carbon from glaciers and ice sheets." Nature

571 Geoscience 8.2 (2015): 91-96.

572

573 Jansson, Janet, et al. "Metabolomics reveals metabolic biomarkers of Crohn’s disease." PloS

574 one 4.7 (2009): e6386.

575

576 Kellerman, Anne M., et al. "Chemodiversity of dissolved organic matter in lakes driven by

577 climate and hydrology." Nature communications 5 (2014).

578

579 Kellerman, Anne M., et al. "Persistence of dissolved organic matter in lakes related to its

580 molecular characteristics." Nature Geoscience 8.6 (2015): 454-457.

581

582 Kim, Sunghwan, Louis A. Kaplan, and Patrick G. Hatcher. "Biodegradable dissolved organic

583 matter in a temperate and a tropical stream determined from ultra-high resolution mass

584 spectrometry." Limnology and Oceanography 51.2 (2006): 1054-1063.

585

586 Kujawinski, Elizabeth B. "The impact of microbial metabolism on marine dissolved organic

587 matter." Annual review of marine science 3 (2011): 567-599.

588

589 Lafrenière, Melissa J., and Martin J. Sharp. "The concentration and fluorescence of dissolved

590 organic carbon (DOC) in glacial and nonglacial catchments: interpreting hydrological flow

591 routing and DOC sources." Arctic, Antarctic, and Alpine Research 36.2 (2004): 156-165.

26 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

592

593 Logue, Jürg B., et al. "Experimental insights into the importance of aquatic bacterial community

594 composition to the degradation of dissolved organic matter." The ISME journal (2015).

595

596 Pierce, Alan Eugene. "Silylation of organic compounds." (1968).

597

598 R Core Team (2014). R: A language and environment for statistical

599 computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-

600 project.org/.

601

602 Rangecroft, S., S. Harrison, and K. Anderson. "Rock glaciers as water stores in the Bolivian

603 Andes: an assessment of their hydrological importance." Arctic, Antarctic, and Alpine

604 Research 47.1 (2015): 89-98.

605

606 Rovira, Pere, and V. Ramón Vallejo. "Labile and recalcitrant pools of carbon and nitrogen in

607 organic matter decomposing at different depths in soil: an acid hydrolysis

608 approach." Geoderma 107.1 (2002): 109-141.

609

610 Shannon, C. E. "A Mathematical Theory of Communication-An Integrated Approach." (1948).

611

612 Singer, Gabriel A., et al. "Biogeochemically diverse organic matter in Alpine glaciers and its

613 downstream fate." Nature Geoscience 5.10 (2012): 710-714.

27 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

614 Stibal, Marek, et al. "Organic matter content and quality in supraglacial debris across the ablation

615 zone of the Greenland ice sheet." Annals of Glaciology51.56 (2010): 1-8.

616

617 Stubbins, Aron, et al. "Anthropogenic aerosols as a source of ancient dissolved organic matter in

618 glaciers." Nature Geoscience 5.3 (2012): 198-201.

619

620 Sumner, Lloyd W., et al. "Proposed minimum reporting standards for chemical

621 analysis." Metabolomics 3.3 (2007): 211-221.

622

623 Tranter, Martyn, Mark Skidmore, and Jemma Wadham. "Hydrological controls on microbial

624 communities in subglacial environments." Hydrological Processes 19.4 (2005): 995-998.

625

626 Volk, Christian J., Catherine B. Volk, and Louis A. Kaplan. "Chemical composition of

627 biodegradable dissolved organic matter in streamwater."Limnology and Oceanography 42.1

628 (1997): 39-44.

629

630 Wahrhaftig, Clyde, and Allan Cox. "Rock glaciers in the Alaska Range." Geological Society of

631 America Bulletin 70.4 (1959): 383-436.

632

633 Wetzel, Robert G. "Gradient-dominated ecosystems: sources and regulatory functions of

634 dissolved organic matter in freshwater ecosystems." Dissolved organic matter in lacustrine

635 ecosystems. Springer Netherlands, 1992. 181-198.

636

28 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

637 Williams, M. W., et al. "Nitrate content and potential microbial signature of rock glacier outflow,

638 Colorado Front Range." Earth Surface Processes and Landforms 32.7 (2007): 1032-1047.

639

640 Woo, Ming-ko. Permafrost hydrology. Springer Science & Business Media, 2012.

641

642

643

644 Acknowledgements

645 The authors would like to thank Edward Stets and Ann Hess for their for help with DOM pool

646 modeling, Corey Broeckling and Sarah Lyons for their help interpreting metabolomic data, and

647 Gunnar Johnson for his help in the field and with figures. This project is a product of the

648 Western Mountain Initiative (WMI) and is funded by the United States Geological Survey.

649 Edward K Hall was supported by NSF award DEB#1456959 during the preparation of this

650 manuscript. All data are available to the public and reside with Loch Vale Watershed long-term

651 ecological research and monitoring program database at

652 http://www.nrel.colostate.edu/projects/lvws/data.html.

653

29 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

654 Figure Legends 655 656 Figure 1 A) Stars note the approximate location of each glacier feature that was sampled for this

657 study within the state of Colorado. In total four pairs of glaciers and rock glaciers were sampled

658 along the Front Range of Colorado. (B) Isabelle Glacier and Navajo Rock Glacier form a pair

659 of a glacier and a rock glacier from the same watershed are shown here to illustrate the

660 differences between the two types of features.

661

662

663 Figure 2 PCA GC-MS Analysis DOM compounds A) PCA plots showing separation along PC5

664 between glaciers (blue) and rock glaciers (red) before incubation and B) no significant difference

665 between glacier types after incubation (p-value of 0.119 and 0.0681 for PC1 and PC2,

666 respectively). C) The Shannon-Wiener Index (SW) for chemical diversity was similar between

667 glaciers and rock glaciers before incubation, however microbial metabolism increased chemical

668 diversity in glacier DOM and decreased chemical diversity in rock glacier DOM.

669

670

671 Figure 3 Molecular distribution of GC-MS identified compounds A) Distribution of

672 compounds by ion intensity pre- (orange) and post-incubation (brown). Many of the compounds

673 present before incubation that were of intermediate abundance were metabolized. The most

674 abundant molecules were different between pre- and post- incubation metabolomics analysis (see

675 Table 4).

676

677

30 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

678 Figure 4 Dissolved oxygen consumption curves from glaciated watersheds (n=8). Values were

679 averaged for each of the glaciers (blue) and rock glaciers (red), and the analytical control (dotted

680 black). All curves were smoothed using a third order polynomial regression function (R2=0.999).

681 Here we show 95% Confidence Intervals in light blue for glaciers and in pink for rock glaciers.

682 Glaciers consumed a larger portion of available DOM during the course of the incubation.

683 684 685 686 687 688 689

31 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

690 A. B.

691 692 Figure 1 A) Stars note the approximate location of each glacier feature that was sampled for this

693 study within the state of Colorado. In total four pairs of glaciers and rock glaciers were sampled

694 along the Front Range of Colorado. (B) Isabelle Glacier and Navajo Rock Glacier form a pair

695 of a glacier and a rock glacier from the same watershed are shown here to illustrate the

696 differences between the two types of features.

697 698 699

32 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

700 Scores Scores Loadings Loadings 701 Scores Scores Loadings Loadings Scores Scores Loadings Loadings

Scores Scores Loadings ScoresLoadings Scores Loadings Loadings 0

Scores Scores0 Scores ScoresLoadings Loadings Loadings Loadings

4

4

0

0 0

0 blank blank

0

0

4

4

7

7

0

0 0

0 blank blank 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl− 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl− 0

0 freezeBlank freezeBlank

0

0

4

4

0

0 7

7 blank blank

3

3

0

0 5

5 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl− 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl− 4

4 Scores Scores Loadings Loadings 0

0 iceGlacier( Rock(Glacier( iceGlacier( Rock(Glacier( 3.4freezeBlank freezeBlank

0

.

0

0 0

. ice ice

4

4

0

0 7

b7 lank blank 0

0 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl− 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl− 3

A. B. A. B. 3 5

C. C. freez5 eBlank freezeBlank 1

1 Disiloxane, hexamethyl− Disiloxane, hexamethyl− 0

0 B.

4

4

4

4 .

. ice ice

0

0

0

0 0

Succinic acid, 3−chlorophenyl 4−methoxybenzyl ester Succinic acid, 3−chlorophen0 yl 4−methoxybenzyl ester 7

rock rock 7 blank rock blank blank rock blank 0

0 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl− 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl−

3

3

5

5 2 A. B. freezeBlank C. freezeBlank 2

Disiloxane, hexamethyl− Disiloxane, hexamethyl− 1

1 Maltose( Threonic(acid( Maltose( Threonic(acid(

.

2 2

. ice Serineice (3TMS) Serine (3TMS)

7

7

7

7

.

.

0

0

0

0 0

0 1,2,4−Triazole, 3−mercapto−4−phen1,2,4yl−T−r5iaz−methole, 3yl−−mercapto−4−phenyl−5−methyl− =

= 1,2,4−Triazole, 3−mercapto−4−phen1,2,4yl−T−r5iaz−methole, 3yl−−mercapto−4−phenyl−5−methyl− 3

3 rock rock

5

5 .

. freezeBlank freezeBlank freezeBlank freezeBlank

2

2 0

0 Disiloxane, hexamethyl− Disiloxane, hexamethyl− .

. ice ice BenzSerineenecarbothioic (3TMS) acid, 2,4,6pyroglutamate−triethyl−, S−(2−phenylethyl) ester BenzSerineenecarbothioic (3TMS) acid, 2,4,6pyroglutamate−triethyl−, S−(2−phenylethyl) ester

0

0

.

.

3

3

3

3 0

3rock.2 rock 0

5

5

= 5

5 glycine glycine

= 2

2 aspartate aspartate

2

2

2

2 0

0 Disiloxane, hexamethyl− Disiloxane, hexamethyl−

.

. .

ice . ice ice ice 4

Diiodoacetylene Diiodoacetylene 4 0

0 Serine (3TMS) Serine (3TMS) 0 0 blank blankBenzenecarbothioic acid, 2,4,6 −triethyl−, S−(2−phenylethyl) ester Benzenecarbothioic acid, 2,4,6 −triethyl−, S−(2−phenylethyl) ester

Glutamate( Quinic(acid( . Glutamate( Quinic(acid( pyroglutamate.

. pyroglutamate . rock betarock−alanine beta−alanine

= Sucrose, 8TMS derivative Sucrose, 8TMS derivative

=

0

0 ^

^ glycine glycine

2

2 2

aspar 2 tate aspartate 0

Pentadecan−1−ol, n− (1TMS) Pentadecan−1−ol, n− (1TMS) 0 Disiloxane, hexamethyl− Disiloxane, hexamethyl− Disiloxane, hexamethyl− Disiloxane, hexamethyl−

=

=

0

0 0

0 Serine (3TMS) 7

Homosalate, TMS derivative Homosalate, TMS derivative 7 pBenzyroglutamateenecarbothioiclactate acid, 2,4,6 −triethyl−, S−(2−phenSerylethineyl) (3TMS) esterpBenzyroglutamateenecarbothioiclactate acid, 2,4,6 −triethyl−, S−(2−phenylethyl) ester

r

r

.

. 0

rock 0 rock rock rockpyroglutamate1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl− pyroglutamate1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl−

=

=

0

0 0

0 beta−alanine glycine beta−alanine 2

freezeBlank 2 Sucrose, 8TMS derivative glycine 2

freezeBlank 2 Sucrose, 8TMS derivative

2

2 ^ ^ asparNorthreonineleucinetate (3TMS) asparthreoninetate

Norleucine (3TMS)

0

0 BenzSerineenecarbothioic (3TMS) acid, 2,4,6glutamate−Sertriethineyl −(3TMS), S−ser(2−inephen3,5yleth−Diyl)−ter estert−butylBenzSer−4ineenecarbothioic−h ydro(3TMS)xybenzoic acid, acid, 2,4,6 glutamateTMS−Sertr deriethineivylativ −(3TMS), Se−ser(2−inephen3,5yleth−Diyl)−ter estert−butyl−4−hydroxybenzoic acid, TMS der ivative

2

.

. . Diiodoacetylene .

2 Diiodoacetylene pyroglutamatepalmitelaidicpyroglutamate acid 02 pyroglutamatepalmitelaidicpyroglutamate acid 02

3

=

=

3 =

= lactate r

r lactate

;

;

0

0

0

0 5

C996( Glycolic(acid(( C996( Glycolic(acid(( 5 beta−alanine glycinephosphoric acid Sucrose, 8TMSbeta− alaninederivative glycinephosphoric acid Sucrose, 8TMS derivative

2

2 ^ ^ aspartate HydroScopoletin,xylamine TBDMS xylose/lyxose(3TMS)glucosedisacchar der ivatividee aspartate Hydroxylamine xylose/lyxose(3TMS)glucosedisaccharide

0 Scopoletin, TBDMS der ivative 0 threonine . threonine . ice ice Norleucine (3TMS) Norleucine (3TMS) 3 Benzenecarbothioic acid,1,2 2,4,6−Ethenediol,−Benztriethenecarbothioicyl− ,2TMS S−(2− derpheniv 3,5acid,ativyleth−eDi 2,4,6yl)−ter estert−−trbiethutylBenzyl−4−enecarbothioic−, Shydro−(2−xybenzphenyleth oicacid, yl)acid,1,2 2,4,6ester− TMSEthenediol,−Benztr iethderenecarbothioicyliv−ativ ,2TMS Se−(2− derpheniv 3,5acid,ativyleth−eDi 2,4,6yl)−ter estert−−trbiethutylyl−4−−, Shydro−(2−xybenzphenylethoic yl)acid, ester TMS der ivative

^ pyroglutamate pyroglutamate

^ Dibenz[d,f]cycloheptane , 5−amino−1,2,3,8−tetramethoxy− Dibenz[d,f]cycloheptane , 5−amino−1,2,3,8−tetramethoxy− pyroglutamate lactateglutamate serine pyroglutamate lactatepglutamateyroglutamate serine pyroglutamate

r

r

0

0 0

Methanone, (2−amino−5−chlorophenyl)(2−fluorophenyl)− Methanone, (2−amino−5−chlorophenyl)(2−fluorophenyl)− beta−alanine glycinepalmitelaidic acid 02 betaglycine−alanine 0 glycinepalmitelaidic acid 02 glycine

2

2 2 2 Boric acid, 3TMSSucrose der ivativ, 8TMSe derivative

; Boric acid, 3TMSSucrose der ivativ, 8TMSe derivative

; trisiloxane, 1,1,1,5,5,5 −hexamethyl−3−[(trimethylsilyl)oxy]− trisiloxane, 1,1,1,5,5,5−hexamethyl−3−[(trimethylsilyl)oxy]−

9

^ 9

C349 C349 ^ aspartate phosphoric acid aspartate aspartate phosphoric acid aspartate 1 Silane, diethyl(2−phenylethoxy)tetradecyloxy− Silane, diethyl(2−phenylethoxy)tetradecyloxy− threonine 1 xylose/lyxoseglucosedisaccharDisiloxaneide , hexamethyl− xylose/lyxoseglucosedisaccharDisiloxaneide , hexamethyl− Terephthalic acid, octyl 2,4,40 −trimethylpentyl ester threonine Terephthalic acid, octyl 2,4,40 −trimethylpentyl ester r Norleucine (3TMS) r Succinic acid, 2−fluorophenyl 2,4−dimethylpent−3−yl ester Norleucine (3TMS) Succinic acid, 2−fluorophenyl 2,4−dimethylpent−3−yl ester HydroScopoletin,xylamine TBDMS (3TMS) der ivative HydroScopoletin,xylamine TBDMS (3TMS) der ivative

0 3,5−Di−tert−butyl−4−hydroxybenzoic acid, TMS der ivative 3−pyrazolidinone, 4,4−dimethyl−1−phenyl−2−(trimethylsilyl)− 3−pyrazolidinone, 4,4−dimeth0 yl−1−phenrockyl−2−(trimethylsilyl)− pyroglutamate rocklactate glutamate1,2−Ethenediol,ser ine2TMS der ivative pyroglutamate lactate glutamate1,2−Ethenediol,ser ine2TMS3,5 der−Di−ivterativt−ebutyl−4−hydroxybenzoic acid, TMS der ivative Silane, diethyldodecylohomoserxypentyloiner xy−

SuccinicCyclodisilazaneSilane acid,, di(2dimeth, 2,2,4,4−fluorophenyl(dimeth−tetramethyl)yldecylo esteryl−xysilylo1,3−diphenxy)decyloyl− xy− SuccinicSilaneCyclodisilazaneSilane, diethacid,,yldodecylo di(2dimethhomoser, 2,2,4,4−fluorophenyl(dimethxypentyloine−r tetramethyl)yldecylo esterxy−yl−xysilylo1,3−diphenxy)decyloyl− xy− beta−alanine betapalmitelaidic−alanine acid 02 Sucrose, 8TMSbeta−alanine derivative Sucrose, 8TMSbetapalmitelaidic− alaninederivativ acide 02 Sucrose, 8TMS derivative Sucrose, 8TMS derivative

;

^

^

;

^

^ 2

phosphortrBorisiloicxane acid,ic acid, 1,1,1,5,5,5 3TMS2 der −ivheativxamethe yl−3−[(trimethylsilyl)oxy]− phosphortrBorisiloicxane acid,ic acid, 1,1,1,5,5,5 3TMS der−ivheativxamethe yl−3−[(trimethylsilyl)oxy]−

9 9

1 Phenol, 2,4,6−tri−terxylose/lyxoset−butyl−

1 3−pyrazglucosedisaccharolidinone,ide 4,4−dimethyl−1−phenyl−2−(trimethPhenol,ylsilyl)− 2,4,6−tri−terxylose/lyxoset−butyl−

0

0 0 C884( C589( C884( C589( 0 threonine 3−pyrazglucosedisaccharolidinone,ide 4,4−dimethyl−1−phenyl−2−(trimethylsilyl)−

1 threonine

Norleucine (3TMS) 1 Serine (3TMS) HydroScopoletin,xylaminepalmitic3,5 TBDMS− (3TMS)Di acid−ter dert−bivutylativ−e4−hydroxybenzoic acid,Serine TMSNor (3TMS)leucine der ivativ Hydro(3TMS)Scopoletin,e xylaminepalmitic TBDMS (3TMS) acid der ivative ; 3,5−Di−tert−butyl−4−hydroxybenzoic acid, TMS der ivative ; pyroglutamate 1,2−Ethenediol,pyroglutamate 2TMS der ivative 1,2−Ethenediol, 2TMS der ivative

. serine pyroglutamate pyroglutamate

glutamatelactate . lactate glutamate serine r

r lactate lactate

r r

similar to Cyclohexasiloxane, dodecamethyl C Silane, diethyldecylosimilarxyoctylo to xyCyclohe− xasiloxane, dodecamethyl Silane, diethyldecyloxyoctyloxy− palmitelaidic acid 02 C

= palmitelaidic acid 02

Siloxane hydroxybenzoic acid 3−Methyl−3−(N−methylSilo−2−xanepyrrolyl)−1,2−diphenhydroylcyclopropenexybenzoic acid = 3−Methyl−3−(N−methyl−2−pyrrolyl)−1,2−diphenylcyclopropene ;

similar to Cyclopentasiloxane, decamethyl similar to Cyclopentasiloxane, decameth; yl phosphorictr Boracidisiloicxane acid,, 1,1,1,5,5,5 3TMS der−ivheativxamethe yl−3−[(trimethylsilyl)oxy]phosphor− ictr Boracidisiloicxane acid,, 1,1,1,5,5,5 3TMS der−ivheativxamethe yl−3−[(trimethylsilyl)oxy]− 9

Cyclooctasiloxane, heSilanexadecameth, diethyldodecyloyl− xy(2−methoxyethoxy)− Cyclooctasiloxane, heSilanexadecameth, diethyldodecyloyl− 9 xy(2−methoxyethoxy)− threonine xylose/lyxoseglucosedisaccharide threonine threonine xylose/lyxoseglucosedisaccharide threonine

Norleucine (3TMS) Norleucine (3TMS)

Norleucine (3TMS) Norleucine (3TMS) 1 Hydroxylamine (3TMS) 1 1 Phenol, 2,4,6−3−trpi−yrterazt−olidinonebutyl− , 4,4−dimethyl−1−phenylHydro−2−(trxylamineimethylsilyl) (3TMS)−

1 Scopoletin, TBDMS der ivative Phenol, 2,4,63−−trpi−yrterazt−olidinonebutyl− , 4,4−dimethyl−1−phenyl−2−(trimethylsilyl)− 1 1 Benzenecarbothioic acid, 2,4,6palmitic3,5−−trDiieth −acidteryl−t−, Sbutyl−(2−4phen−hydroylethxybenzyl) 3,5esteroic−Di Benzacid,−terScopoletin,enecarbothioict −TMSbutyl der−4 −ivTBDMShativydro eacid,xybenz der 2,4,6ivpalmiticoicativ3,5− acid,e−trDiieth −acid terTMSyl−t−, Sb derutyl−(2iv−4ativphen−heydroylethxybenzyl) 3,5esteroic−Di acid,−tert −TMSbutyl der−4−ivhativydroexybenzoic acid, TMS der ivative

0 serine serine

0 1,2−glutamateEthenediol, 2TMS derpyroglutamateivative glutamate 1,2−glutamateEthenediol, 2TMSser derinepyroglutamateivative glutamate serine

0 8

0 palmitelaidic acid 02 palmitelaidic acid 02 palmitelaidic acid 02 palmitelaidic acid 02

8

0

2

0

2

3

; ;

Silane, dimethyl(octadecyloxy)propyl− C

3

; ;

phosphoric acid phosphorSilaneic acid, dimethyl(octadecyloxy)propyl− glycine C glycine 2

2 phosphoric acid phosphoric acid

. phosphoric acid phosphoric acid

. trBorisiloicxane acid,, 1,1,1,5,5,5 aspar3TMState der−ivheativxamethe yl−3−[(trimethylsilyl)oxy]− trBorisiloicxane acid,, 1,1,1,5,5,5 aspar3TMState der −ivheativxamethe yl−3−[(trimethylsilyl)oxy]−

9 .

9 xylose/lyxosedisaccharide xylose/lyxosedisaccharide . P glucose glucose xylose/lyxose xylose/lyxose 2−Phenoxathiinamine−10,10P −dioxide 2−Phenoxathiinamineglucosedisacchar−10,10ide −dioxide glucosedisaccharide

2.8 1 Hydroxylamine (3TMS) Hydroxylamine (3TMS) 1 Hydroxylamine (3TMS) Hydroxylamine (3TMS) 1 Scopoletin,Phenol, TBDMS 2,4,63−−tr pderi−yrterazivtativ−olidinonebutyle Scopoletin,− , 4,4−dimeth TBDMSyl−1 −derphenivativyl−e2−(trimethScopoletin,ylsilyl)− TBDMS der ivative Scopoletin, TBDMS der ivative

1 Phenol, 2,4,63−−trpi−yrterazt−olidinonebutyl− , 4,4−dimethyl−1−phenyl−2−(trimethylsilyl)− .

5H−Pyrano[2,3,4,5−lmn]phenanthridin−5−one 5H−Pyrano[2,3,4,5−lmn]phenanthridin−5−one 1,2−Ethenediol, 2TMSpalmitic der ivacidativ1,2e −. Ethenediol, 2TMS der ivative 1,2−Ethenediol, 2TMSpalmitic der ivacidativ1,2e −Ethenediol, 2TMS der ivative 1 C815( C531( Scopoletin,C815 TBDMS( der ivativCe 531( Scopoletin, TBDMS der1 ivative

C beta−alanine NA C beta−alanine

0 NA 0

2 Sucrose, 8TMS derivative

2 3−Dimethyl(phenyl)silyloxypentadecane Sucrose, 8TMS derivative 2

3−Dimethyl(phenyl)silyloxypentadecane 2 ^

^ trBorisiloicxane acid,, 1,1,1,5,5,5 3TMS der −ivheativxamethe trBorisiloyl−ic3xane −acid,[(tr,imeth 1,1,1,5,5,5 3TMSylsilyl)o der −ivxy]heativxameth−e yltr−Borisilo3−[(tricxane acid,imeth, 1,1,1,5,5,5 3TMSylsilyl)o derxy]−iv−heativxamethe trBorisiloyl−ic3xane −acid,[(tr,imeth 1,1,1,5,5,5 3TMSylsilyl)o der −ivxy]heativxameth−e yl−3−[(trimethylsilyl)oxy]−

9

9

9

9

.

.

0

0

1

1

1

1

0

0

P

P

1 0 1 Phenol, 2,4,623−−trPhenopi−yrterazt−olidinonexathiinaminebutyl− , 4,4−−dimeth10,10−yldio−1xide−phenyl−2−(trimethPhenol,ylsilyl)− 2,4,6 −tri−tert−butyl−

0 23−Phenopyrazolidinonexathiinamine, 4,4−−dimeth10,10−yldio−1xide−phenyl−2−(trimethylsilyl)− C

C palmitic acid palmitic acid

1

1

.

. 0 pyroglutamate 0 pyroglutamate

r lactate lactate

r

:

:

C

C 2

Phenylethylmalonamide, 2TMS derivative Phenylethylmalonamide, 2TMS derivative 2

0

0

3

3 .

. threonine

1

1

l 1

1 Phenol, 2,4,6 −tri−tert−butyl− Phenol, 2,4,6 −tri−tert−butyl− threonine l

P 3−Norpyrleucineazolidinone (3TMS), 4,4−dimethyl−31−−pphenyrazolidinoneyl−2−(trimeth, 4,4Phenol,−ylsilyl)dimeth− 2,4,6yl−1−−phentri−teryl−t−2b−utyl(trimeth− Phenol,ylsilyl)− 2,4,6 −tri−tert−butyl−

P 3−pyrazolidinone, 4,4−dimethyl−31−−pphenyrazolidinoneyl−2−(trimeth, 4,4−ylsilyl)dimeth−yl−1−phenyl−2−(trimethylsilyl)− 2−Phenoxathiinamine−10,10−dioxide Norleucine (3TMS)

0 2−Phenoxathiinamine−10,10−dioxide

0 palmitic acid 3,5−Di−palmitictert−butyl acid−4−hydroxybenzoic acid,palmitic TMS der acidivative 3,5−Di−palmitictert−butyl acid−4−hydroxybenzoic acid, TMS der ivative

1

1 . C716 glutamate 1−Ethoserxyine−2−. propanol, TMS der ivative glutamate serine

. 1−Dibenzofurancarboxylic acid, 3,9b−dihydro−4,9b−dimethyl−, methylC716 ester 1−Ethoxy−2−propanol, TMS der ivative

. 1−Dibenzofurancarboxylic acid, 3,9b−dihydro−4,9b−dimethyl−, methyl ester palmitelaidic acid 02

0 palmitelaidic acid 02

0

C

C C

Silane1−Propanol,, dimeth TMSyl(3 −dermethivativylbeut−2−enyloxy)tridecyloxy− Silane1−Propanol,, dimeth TMSyl(3 −dermethivativylbeut−2−enyloxy)tridecyloxy− C

P

P − C1890( C35( NA ; NA

C1890( C35( ;

2

: 2

: phosphoric acid phosphoric acid .

FuranC2146−Phthalic2−carbo xylicacid,Pentadecan acid isobpalmitic [2utyl−−(2,2,6,6 14− acidol,octyl n− ester tetr(1TMS)amethyl−piperidin−4−yl)−ethFuryl]anC2146−amide−Phthalic2−carbo xylicacid,Pentadecan acid isobpalmitic [2utyl−−(2,2,6,6 14−− . acidol,octyl n− ester tetr(1TMS)amethyl−piperidin−4−yl)−ethyl]−amide xylose/lyxoseglucosedisaccharide xylose/lyxoseglucosedisaccharide

P 2 2−Phenoxathiinamine−10,10P −dioxide 2 Hydroxylamine (3TMS) 0 2−PhenoxathiinamineHydro−10,10xylamine−dioxide (3TMS)

0 Scopoletin, TBDMS der ivative Scopoletin, TBDMS der ivative

l

l

1

1

a

1 1 C407 C407 a . 1,2−Ethenediol, 2TMS der ivative

1,3−Dimethyl−2,6−dioxo−2,3,6,7−tetrahydro−1H−purine−8−carbaldeh1,3yde−Dimeth−oximeyl−2,6−dioxo−2,3,6,7−tetrahydro−1H−purine−8−carbaldehyde−oxime . 1,2−Ethenediol, 2TMS der ivative 0

1−Ethoxy−0 2−propanol, TMS der ivative 2

2 1−Ethoxy−2−propanol, TMS der ivative 2

4−Hydroxy−4,5−dihydro−2−methyl−3,3−bis(ethoxycarbonyl)−4−ethoxy−3H−pyrrole 2 0

: 4−Hydroxy−4,5−dihydro−2−methyl−3,3−bis(ethoxycarbonyl)−4−ethoxy−3H−pyrrole 0

stearic acid stearic acid : −

2.6 Boric acid, 3TMS der ivative − Boric acid, 3TMS der ivative .

. trisiloxane, 1,1,1,5,5,5 −hexamethyl−3−[(trimethylsilyl)oxy]− trisiloxane, 1,1,1,5,5,5 −hexamethyl−3−[(trimethylsilyl)oxy]−

.

. 9

C2166 C2166 9 P

Phosphetane, 2,2,3,4,4−pentamethyl−1−phenoxy−, 1−oxide Phosphetane, 2,2,3,4,4−pentamethyl−1−phenoxy−, 1−oxide P

P P

1 2−Phenoxathiinamine−10,10−dio2−xidePhenoxathiinamine−10,10−dioxide

1 2−Phenoxathiinamine−10,10−dio2−xidePhenoxathiinamine−10,10−dioxide

0

0

l

l

.

.

.

.

v

:

v :

− Phthalic acid, octyl 2 −propylpentyl ester

− Phthalic acid, octyl 2 −propylpentyl ester

a

a 0

1−0 Ethoxy−2−propanol, TMS der ivative 1−Ethoxy−2−propanol, TMS der ivative

:

:

l l

C1870( C1942( C2165 GMDCA1360021870( C1942( C2165 − 0

0 GMDA136002

0 0 1 Phenol, 2,4,6 −tri−tert−butyl−

Decan−2−yl trimethylsilyl phthalate Decan−2−yl trimethylsilyl1 phthalate 3−pyrazolidinone, 4,4−dimethyl−1−phenyl−2−(trimethPhenol,ylsilyl)− 2,4,63−−trpi−yrterazt−olidinonebutyl− , 4,4−dimethyl−1−phenyl−2−(trimethylsilyl)−

2

2 l

l palmitic acid palmitic acid

0

0

a

a

v

0

0

v

0 0

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

C 1−Ethoxy−2−propanol, TMS der ivative

C 1−Ethoxy−2−propanol, TMS der ivative

:

:

:

:

.

.

a

a

l

l

l

l

1

1

a

a

v

v

0

0 0 0 1−Ethoxy−2−propanol, TMS der1iv−ativEthoexy−2−propanol, TMS der ivativ1−Ethoe xy−2−propanol, TMS der1iv−ativEthoexy−2−propanol, TMS der ivative

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

;

2

;

2

.

.

0

0

v

v P

2−Phenoxathiinamine−10,10P −dioxide 2−Phenoxathiinamine−10,10−dioxide

.

. a

C1840( C1439( C1840( C1439(a

a

a

v

v 0

0 2.4

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

0

0

0

0

;

;

1

1

0

0

p

p

v

v v v Glycerol (3TMS)

0 Glycerol (3TMS)

0

0 0

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

0

:

0

:

;

;

0

0

0

0

1

1

6

6

l

l

0

0

;

0

0

;

C

C p

p 2−Dimethyl(trimethylsilyl)silyloxytetr2−Dimethadecaneyl(trimethylsilyl)silyloxytetradecane p C1739( C1376( p 1Glycerol−Ethoxy −(3TMS)2−propanol, TMS der ivative 1Glycerol−2Etho−Dimethxy −(3TMS)2−yl(trpropanol,imethylsilyl)silylo TMS der ivxytetrativ2−Dimetheadecaneyl(trimethylsilyl)silyloxytetradecane

C17390 ( C1376(

0

.

.

;

;

0 0

NA 1

NA 1

0

0

1

1

6

6 a

a Glycerol (3TMS)

0 C

0 Glycerol (3TMS)

3

0

C

0

3

;

; ;

2.2 ;

0

P

0

.

P

.

0

0

1

1

0

0

6

1 6

Silane, diethylisohexyloxy(2−methylpentyloxy)− Silane, diethylisohexyloxy(2−methylpentylo1 xy)−

v

− v − Glycerol (3TMS)

C Glycerol (3TMS)

C

4

4

0

0

0

0

.

. 1

C1655( C1226( 1 1

C1655( C1226( 1

P

0

0

P

0

0

0

0

C

C

.

.

6

1

6

1 0

0 Glycerol (3TMS) Glycerol (3TMS) Glycerol (3TMS) Glycerol (3TMS)

− C

− 2−Dimethyl(trimethylsilyl)silyloxytetradecane p

C 2−Dimethyl(trimethylsilyl)silyloxytetradecane

p

.

.

0

0

P

0

0

P

0

0

6

6

6

6

0

1

0

1

0

0

C

C

C

C

P

0

P

0

;

.

.

;

.

.

P

P

0

0

1

1 1

C1414( C1125( C1414( C1125( 1

0

0

− −

0 2

0

P

P

P

P

1

0

0

1 0

−5000 0 5000 −5000 0−0.4 5000−0.2 0.0 0.2 −0.4 0 −0.2 0.0 0.2

− −

− Glycerol (3TMS) Glycerol (3TMS)

6

6 0

−5000 0 0 5000 −5000 0−0.4 5000−0.2 0.0 0.2 −0.4 −0.2 0.0 0.2

6

6

− C

! ! C

.

. −

− −5000 0 5000 −5000 0−0.4 5000−0.2 0.0 0.2 −0.4 −0.2 0.0 0.2 1

−2000 0 2000 −2000 0 −0.12000 0.1 0.3 −0.1 0.1 0.3 1

P

P 0 −5000PC2; pval:0.0681;0 5000 r^2=0.193−5000 0−0.4 5000−0.2 0.0PC2 0.2 −0 0.4 −0.2 0.0 0.2 − PC2; pval:0.0681; r^2=0.193 PC2

−5000 0−50005000 − 0 −50005000 0−−0.450005000−0.2 0−0.40.0 5000−0.20.2 −0.00.4 −0.20.2 −0.40.0 −0.20.2 0.0 0.2 − PC2; pval:0.0681;Scores r^2=0.193 PC2; pval:0.0681; r^2=0.193PC2 Loadings− PC2 PC5; pval:0.0101; r^2=0.066 PC5; pval:0.0101; r^2=0.066PC5 PC2; pval:0.0681;PC5 Scores r^2=0.193 PC2; pval:0.0681; r^2=0.193PC2 Loadings PC2 PC2; pval:0.0681;−5000 r^2=0.193Scores0 5000PC2; pvPral:0.0681;−e5000-Incubat ir^2=0.193on0−0.4Post-PC2I5000−n0.2cubatio0.0n Loadings0.2 −0.4 PC2−0.2 0.0 0.2 Scores Loadings PC2; pval:0.0681;ScoresPC2; r^2=0.193 pval:0.0681;PC2; r^2=0.193 pval:0.0681;PC2; r^2=0.193 pval:0.0681;PC2 r^2=0.193LoadingsPC2 PC2 PC2

702 0 Scores Scores Loadings Loadings

4 0

C. D. 0 PC2; pval:0.0681; r^2=0.193 PC2;b lankpval:0.0681; r^2=0.193PC2 PC2

0

4

7 0 0 blank 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl−

0 703 3.4 freezeBlank

0

4

0 7

3 blank 0

5 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl−

4 Scores Loadings

0 iceGlacier( Rock(Glacier( freezeBlank

0 0

. ice

4 0

7 blank

0 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl− A. B. 3 C. 5 freezeBlank

1 Disiloxane, hexamethyl−

0

4 4

. ice

0 0 704 Figure 2 PCA GC-MS Analysis DOM compounds0 Succinic acid, 3 −A)chlorophen PCAyl 4−metho plotsxybenzyl ester showingblank separationblank along PC5 rock 7 rock

0 1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl−

3 5 freezeBlank 2 Disiloxane, hexamethyl−

1 Maltose( Threonic(acid( .

2 ice Serine (3TMS)

7

7

.

0

0 0

3.2 = 1,2,4−Triazole, 3−mercapto−4−phen1,2,4yl−T−r5iaz−methole, 3yl−−mercapto−4−phenyl−5−methyl−

3 rock 5

. freezeBlank freezeBlank 2

0 Disiloxane, hexamethyl−

. ice BenzSerineenecarbothioic (3TMS) acid, 2,4,6pyroglutamate−triethyl−, S−(2−phenylethyl) ester

0

.

3 3

rock 0 5

5 glycine =

2 aspartate

2 2

0 Disiloxane, hexamethyl− . . ice ice Diiodoacetylene 4 0 Serine (3TMS) 0 blank Benzenecarbothioic acid, 2,4,6 −triethyl−, S−(2−phenylethyl) ester Glutamate( Quinic(acid( . pyroglutamate . rock beta−alanine

= Sucrose, 8TMS derivative 0

^ glycine 2 2 aspartate

Pentadecan−1−ol, n− (1TMS) 0 Disiloxane, hexamethyl− Disiloxane, hexamethyl−

=

0 0

Homosalate, TMS derivative 7 Serine (3TMS)pBenzyroglutamateenecarbothioiclactate acid, 2,4,6 −triethyl−, S−(2−phenylethyl) ester

r .

0 rock rock pyroglutamate1,2,4−Triazole, 3−mercapto−4−phenyl−5−methyl−

= 0

0 beta−alanine glycine 2

freezeBlank 2 Sucrose, 8TMS derivative 2

^ asparNorthreonineleucinetate (3TMS)

0 BenzSerineenecarbothioic (3TMS) acid, 2,4,6glutamate−Sertriethineyl −(3TMS), S−ser(2−inephen3,5yleth−Diyl)−ter estert−butyl−4−hydroxybenzoic acid, TMS der ivative

. .

2 Diiodoacetylene 3 pyroglutamatepalmitelaidicpyroglutamate acid 02

3 = = lactate between glaciers (blue) and rock glaciers (red)r before incubation and B) no significant difference

705 ;

0 0

C996( Glycolic(acid(( 5 beta−alanine glycinephosphoric acid Sucrose, 8TMS derivative 2 ^ aspartate HydroScopoletin,xylamine TBDMS (3TMS)xylose/lyxoseglucosedisacchar der ivatividee

0 threonine . ice Norleucine (3TMS) Benzenecarbothioic acid,1,2 2,4,6−Ethenediol,−Benztriethenecarbothioicyl− ,2TMS S−(2− derpheniv 3,5acid,ativyleth−eDi 2,4,6yl)−ter estert−−trbiethutylyl−4−−, Shydro−(2−xybenzphenylethoic yl)acid, ester TMS der ivative

^ Dibenz[d,f]cycloheptane , 5−amino−1,2,3,8−tetramethoxy− pyroglutamate lactatepglutamateyroglutamate serine pyroglutamate

r 0

Methanone, (2−amino−5−chlorophenyl)(2−fluorophenyl)− beta−alanine 0 glycinepalmitelaidic acid 02 glycine 2 2 Boric acid, 3TMSSucrose der ivativ, 8TMSe derivative

; trisiloxane, 1,1,1,5,5,5−hexamethyl−3−[(trimethylsilyl)oxy]− 9 C349 ^ aspartate phosphoric acid aspartate Silane, diethyl(2−phenylethoxy)tetradecyloxy− 1 xylose/lyxoseglucosedisaccharDisiloxaneide , hexamethyl−

Terephthalic acid, octyl 2,4,40 −trimethylpentyl ester threonine r Norleucine (3TMS) Succinic acid, 2−fluorophenyl 2,4−dimeth ylpent−3−yl ester HydroScopoletin,xylamine TBDMS (3TMS) der ivative 3−pyrazolidinone, 4,4−dimeth0 yl−1−phenyl−2−(trimethylsilyl)− rock pyroglutamate lactate glutamate1,2−Ethenediol,ser ine2TMS3,5 −derDi−ivterativt−ebutyl−4−hydroxybenzoic acid, TMS der ivative

SuccinicSilaneCyclodisilazaneSilane, diethacid,,yldodecylo di(2dimethhomoser, 2,2,4,4−fluorophenyl(dimethxypentyloine−r tetramethyl)yldecylo esterxy−yl−xysilylo1,3−diphenxy)decyloyl− xy− beta−alanine betapalmitelaidic−alanine acid 02 Sucrose, 8TMS derivative Sucrose, 8TMS derivative

;

^ ^

2 phosphortrBorisiloicxane acid,ic acid, 1,1,1,5,5,5 3TMS der−ivheativxamethe yl−3−[(trimethylsilyl)oxy]− 9

1 xylose/lyxose 0 C884( C589( 0 threoninePhenol, 2,4,63−−trpi−yrterazglucosetdisacchar−olidinonebutyl− ,ide 4,4−dimethyl−1−phenyl−2−(trimethylsilyl)−

1 Norleucine (3TMS) Serine (3TMS) HydroScopoletin,xylaminepalmitic3,5 TBDMS− (3TMS)Di acid−ter dert−bivutylativ−e4−hydroxybenzoic acid, TMS der ivative ; pyroglutamate 1,2−Ethenediol,pyroglutamate 2TMS der ivative

. glutamatelactate serine lactate r similar to Cyclohexasiloxane, dodecamethyl r Silane, diethyldecyloxyoctyloxy− C palmitelaidic acid 02 Siloxane hydroxybenzoic acid = 3−Methyl−3−(N−methyl−2−pyrrolyl)−1,2−diphenylcyclopropene similar to Cyclopentasilo xane, decameth; yl phosphorictr Boracidisiloicxane acid,, 1,1,1,5,5,5 3TMS der−ivheativxamethe yl−3−[(trimethylsilyl)oxy]−

Cyclooctasiloxane, heSilanexadecameth, diethyldodecyloyl− 9 xy(2−methoxyethoxy)− threonine xylose/lyxoseglucosedisaccharide threonine

Norleucine (3TMS) Norleucine (3TMS) 1

1 HydroScopoletin,xylaminePhenol, TBDMS (3TMS) 2,4,6 der iv3−ativ−trpi−yreterazt−olidinonebutyl− , 4,4−dimethyl−1−phenyl−2−(trimethylsilyl)− 2.8 1 Benzenecarbothioic acid, 2,4,6palmitic3,5−−trDiieth −acidteryl−t−, Sbutyl−(2−4phen−hydroylethxybenzyl) 3,5esteroic−Di acid,−tert −TMSbutyl der−4−ivhativydroexybenzoic acid, TMS der ivative 0 1,2−glutamateEthenediol, 2TMSser derinepyroglutamateivative glutamate serine

0 palmitelaidic acid 02 palmitelaidic acid 02

8

0

2

3

; ; phosphorSilaneic acid, dimethyl(octadecyloxy)propyl− C glycine

2 phosphoric acid phosphoric acid . trBorisiloicxane acid,, 1,1,1,5,5,5 aspar3TMState der−ivheativxamethe yl−3−[(trimethylsilyl)oxy]−

9 xylose/lyxose xylose/lyxose . P 2−Phenoxathiinamineglucosedisacchar−10,10ide −dioxide glucosedisaccharide

1 Hydroxylamine (3TMS) Hydroxylamine (3TMS) 1 Scopoletin,Phenol, TBDMS 2,4,63−−tr pderi−yrterazivtativ−olidinonebutyle Scopoletin,− , 4,4−dimeth TBDMSyl−1 −derphenivativyl−e2−(trimethylsilyl)− 5H−Pyrano[2,3,4,5−lmn]phenanthridin−5−one . 1,2−Ethenediol, 2TMSpalmitic der ivacidativ1,2e −Ethenediol, 2TMS der ivative C815( C531( Scopoletin, TBDMS der1 ivative

NA C beta−alanine 0

2 3−Dimethyl(phenyl)silyloxypentadecane Sucrose, 8TMS derivative 2

^ trBorisiloicxane acid,, 1,1,1,5,5,5 3TMS der −ivheativxamethe trBorisiloyl−ic3xane− acid,[(tr,imeth 1,1,1,5,5,5 3TMSylsilyl)o der−ivxy]heativxameth−e yl−3−[(trimethylsilyl)oxy]− 9

706 between glacier types after incubation (p-value9 of 0.119 and 0.0681 for PC1 and PC2,

.

0

1

1

0 P

1 Phenol, 2,4,6−tri−tert−butyl− 0 32−pPhenoyrazolidinonexathiinamine, 4,4−−dimeth10,10−yldio−1xide−phenyl−2−(trimethylsilyl)−

C palmitic acid

1 .

0 pyroglutamate lactate

r

: C

Phenylethylmalonamide, 2TMS derivative 2

0 3

. threonine

1 1

l Phenol, 2,4,6−tri−tert−butyl− Phenol, 2,4,6−tri−tert−butyl−

P 3−pyrazolidinone, 4,4−dimethyl−31−−pphenyrazolidinoneyl−2−(trimeth, 4,4−ylsilyl)dimeth−yl−1−phenyl−2−(trimethylsilyl)− 2−PhenoNorxathiinamineleucine (3TMS)−10,10−dioxide

0 palmitic acid 3,5−Di−palmitictert−butyl acid−4−hydroxybenzoic acid, TMS der ivative 1 1−C716Dibenzofurancarboxylic acid, 3,9b−dihydro−4,9b−dimethyl−, methyl ester . glutamate 1−Ethoserxyine−2−propanol, TMS der ivative

. palmitelaidic acid 02

0 C

Silane1−Propanol,, dimeth TMSyl(3 −dermethivativylbeut−2−enyloxy)tridecyloxy− C

− P

C1890( C35( NA ; 2 : phosphoric acid

FuranC2146−Phthalic2−carbo xylicacid,Pentadecan acid isobpalmitic [2utyl−−(2,2,6,6 14− . acidol,octyl n− ester tetr(1TMS)amethyl−piperidin−4−yl)−ethyl]−amide xylose/lyxoseglucosedisaccharide P 2 2−Phenoxathiinamine−10,10−dioxide 0 HydroScopoletin,xylamine TBDMS (3TMS) der ivative

2.6 l

1 1 C407 a 1,3−Dimethyl−2,6−dioxo−2,3,6,7−tetrahydro−1H−purine−8−carbaldehyde−oxime . 1,2−Ethenediol, 2TMS der ivative

0 1−Ethoxy−2−propanol, TMS der ivative 2

4−Hydroxy−4,5−dihydro−2−methyl−3,3−bis(ethoxycarbonyl)−4−ethoxy−3H−pyrrole 2 0 stearic acid :

− Boric acid, 3TMS der ivative . . trisiloxane, 1,1,1,5,5,5 −hexamethyl−3−[(trimethylsilyl)oxy]−

C2166 9 P Phosphetane, 2,2,3,4,4−pentamethyl−1−phenoxy−, 1−oxide P

1 2−Phenoxathiinamine−10,10−dio2−xidePhenoxathiinamine−10,10−dioxide

0

l

.

.

v :

− Phthalic acid, octyl 2 −propylpentyl ester a

0 1−Ethoxy−2−propanol, TMS der ivative

: l

C1870( C1942( C2165 GMDA136002 −

0 0

Decan−2−yl trimethylsilyl1 phthalate Phenol, 2,4,63−−trpi−yrterazt−olidinonebutyl− , 4,4−dimethyl−1−phenyl−2−(trimethylsilyl)− 2

l palmitic acid

0

a

v

0 0

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

C 1−Ethoxy−2−propanol, TMS der ivative

:

:

.

a

l respectively). C) The Shannon-Wiener Index (SW)l for chemical diversity was similar between

707 1

a

v 0

0 1−Ethoxy−2−propanol, TMS der1iv−ativEthoexy−2−propanol, TMS der ivative

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

;

2

.

0 v

2.4 P 2−Phenoxathiinamine−10,10−dioxide

C1840( C1439( .

a

a

v 0

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

0

0

;

1

0

p

v

v Glycerol (3TMS)

0 0

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

0

:

;

0

0

1

6

l

0

0

;

C p

C1739( C1376( p 1Glycerol−2Etho−Dimethxy −(3TMS)2−yl(trpropanol,imethylsilyl)silylo TMS der ivxytetrativ2−Dimetheadecaneyl(trimethylsilyl)silyloxytetradecane

0

.

; 0

NA 1

0

1

6

a Glycerol (3TMS)

0

C

0

3

;

;

0

.

P

0 1

2.2 0 6

Silane, diethylisohexyloxy(2−methylpentylo1 xy)− v

− Glycerol (3TMS) C

708 glaciers and rock4 glaciers before incubation, however microbial metabolism increased chemical

0

0

. 1

C1655( C1226( 1

P

0

0

0

C

.

1 6

0 Glycerol (3TMS) Glycerol (3TMS)

− C

p 2−Dimethyl(trimethylsilyl)silyloxytetradecane

.

0

0

P

0

6

6

0

1

0

C

C

P

0

;

.

.

P

0 1

C1414( C1125( 1

0

0

P

P

1 0

2 −5000 0 5000 −0.4 0 −0.2 0.0 0.2

− −

− Glycerol (3TMS) 6

0 −5000 0 5000 −0.4 −0.2 0.0 0.2

6 − Pre-Incubation Post-Incubation −

! C . − 709 diversity in glacier DOM and decreased chemical diversity−5000 in rock0 glacier5000 DOM. −0.4 −0.2 0.0 0.2

−2000 0 2000 −0.1 0.1 0.3 1 P −5000PC2; pval:0.0681;0 5000 r^2=0.193 −0 0.4 −0.2 0.0PC2 0.2 − −5000 0−50005000 0 5000 −0.4 −0.2 −0.40.0 −0.20.2 0.0 0.2 PC2; pval:0.0681; r^2=0.193 − PC2 PC5; pval:0.0101; r^2=0.066 PC5 PC2; pval:0.0681; r^2=0.193 PC2 710 PC2;PC2; pv pval:0.0681;−al:0.0681;5000 PC2; r^2=0.193 r^2=0.193 0pval:0.0681;5000 r^2=0.193 −0.4 PC2−PC20.2 0.0 PC20.2 711 PC2; pval:0.0681; r^2=0.193 PC2

33 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Figure712 713 Figure 3 Molecular distribution of GC-MS identified compounds A) Distribution of

714 compounds by ion intensity pre- (orange) and post-incubation (brown). Many of the compounds

715 present before incubation that were of intermediate abundance were metabolized. The most

716 abundant molecules were different between pre- and post- incubation metabolomics analysis (see

717 Table 4).

718 719 720 721

34 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

722

35 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

723 Figure 4 Dissolved oxygen consumption curves from glaciated watersheds (n=8). Values were

724 averaged for each of the glaciers (blue) and rock glaciers (red), and the analytical control (dotted

725 black). All curves were smoothed using a third order polynomial regression function (R2=0.999).

726 Here we show 95% Confidence Intervals in light blue for glaciers and in pink for rock glaciers.

727 Glaciers consumed a larger portion of available DOM during the course of the incubation.

728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763

36 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

764 Tables 765 766 Table 1 Characteristics of DOM and microbial metabolism fed DOM from glaciers and rock

767 glaciers. C:N is the ratio of carbon to nitrogen for pre-incubation DOM. D.O (dissolved oxygen)

768 and DOC (dissolved organic carbon) represent the change in concentration of each during the

769 course of the incubation (mg L-1). RQ = respiratory quotient, SW = Shannon Wiener Diversity

770 Index, and SW = change is SW of DOM before and after the incubation. Bold values represent

771 significant differences between glaciers and rock glaciers (p<0.05), standard deviations of mean

772 values are listed in parentheses.

773 Site Type C:N D.O. DOC RQ SW SW Isabelle G 1.65 6.03 2.50 0.42 2.92 0.19 Peck G 2.02 3.95 1.89 0.48 2.89 0.88 Andrews G 3.00 4.11 2.64 0.64 2.68 0.13 Arapaho G 2.73 3.97 2.84 0.72 2.84 0.52 Peck RG 2.63 3.52 2.09 0.59 2.80 0.54 Navajo RG 1.48 2.81 1.71 0.61 2.83 -0.32 Arapaho RG 2.45 4.12 3.00 0.73 2.79 -1.10 Taylor RG 0.86 2.69 2.25 0.83 2.80 -0.01 Mean G 2.35 (0.62) 4.51 (1.02) 2.47 (0.41) 0.56 (0.14) 2.83 (0.11) 0.43 Mean RG 1.85 (0.83) 3.29 (0.67) 2.26 (0.54) 0.69 (0.11) 2.80 (0.02) -0.22 774 775

37 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

776 777 Table 2 DOM compounds that defined the differences between glaciers and rock glaciers before

778 incubation along PC5 from Figure 3A. Mass to charge ratios (m/z) ratios are the largest ion

779 observed for each compound. Normalized intensities (N.I.), an indicator of compound

780 concentration, are given for glaciers and rock glaciers. Bold values indicate which glacier type

781 had a higher normalized intensity (N.I) on average.

782 783 DOM Peak m/z N.I. Glacier N.I. Rock Glacier Maltose 568.3 4817 2201 Glutamate 365.3 5393 2704 C996 94.2 28959 3589 C884 565.3 14584 1468 C815 419.1 2486 569 C1890 613.2 805 346 C1870 642 1516 1229 C1840 640.1 384 245 C1739 489.1 2407 1376 C1655 480.2 17313 9597 C1414 507.2 550 479 Threonic Acid 411 117 2730 Quinic Acid 409.2 2136 7891 Glycolic Acid 279.2 2920 16696 C589 649.5 1605 1868 C531 343.2 7446 22845 C35 348.2 1422 2188 C1942 535.1 846 1062 C1439 406.2 598 861 C1376 440.2 2623 3610 C1226 634.1 708 1602 C1125 468.1 3550 7504 784 785 786

38 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

787 Table 3 Results from carbon decay model where BO represents the size of the recalcitrant pool of

-1 -1 788 DOM (mg C L ), B1 is the size of the bioavailable pool (mg C L ), k is the decay constant (mg

789 C L-1 h-1), and BGE = bacterial growth efficiency. Glaciers had a larger percentage of BDOM

790 compared to rock glaciers and rock glaciers had a larger proportion of recalcitrant C compared to

791 glaciers. BGE was also higher for glaciers than rock glaciers. Bold values represent significant

792 differences between glaciers and rock glaciers (p<0.05) and standard deviations of mean values

793 are listed in parentheses.

794 795 796 Site Type B0 B1 k % BDOM % Recalcitrant BGE 3.20 5.59 0.001 48.23 51.77 0.37 Arapaho Glacier 3.72 5.20 0.001 47.74 52.26 0.39 Isabelle Glacier 1.35 7.03 0.001 60.90 39.10 0.39 Peck Glacier 3.78 5.24 0.001 46.06 53.94 0.17 Arapaho Rock Glacier 5.32 3.80 0.002 32.05 67.95 0.13 Navajo Rock Glacier 5.64 3.42 0.001 29.89 70.11 0.03 Peck Rock Glacier 3.31 5.24 0.001 45.74 54.26 0.14 Taylor Rock Glacier 4.69 3.83 0.001 36.59 63.41 0.07 Mean G Glacier 3.38 (1.09) 6.35 (1.06) 0.001 52.83 (9.73) 47.17 (7.03) 0.263 (0.13) (0.000) Mean RG Rock Glacier 4.74 (1.26) 4.07 (1.12) 0.001 37.18 (10.23) 62.82 (6.84) 0.157 (0.16) (0.000) 797 798

39 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

799 Table 4 The top 25 compounds present in pre- and post-incubation samples organized by

800 normalized intensity. Glaciers and rock glaciers had the same top 25 compounds before and after

801 incubation. Microbial metabolism altered compounds present, resulting in a different set of

802 compounds post incubation.

803 Pre-Incubation Post-Incubation Succinic Acid Glycerol Octyl 2-propylpentyl ester-phthalic acid 2,4,6-triethyl-, S-(2-phenylethyl) ester- benzenecarbothioic acid Octyl 2,4,4-trimethylpentyl ester-terephthalic acid Pyroglutamate 3-chlorophenyl 4-methoxybenzyl ester-succinic acid 10-dioxide-2-phenoxathiinamine-10 pentadecan-1-ol Serine Hydroxylamine 1-Ethoxy-2-propanol Homosalate Palmitic Acid p-cyanophenyl 4'-heptyl-4-biphenylcarboxylate Sucrose Scopoletin Beta-alanine Diiodoacetylene-1 2,4,6-tri-tert-butyl-phenol Palmitic Acid Aspartate Diiodoacetylene Boric Acid Phosphoric Acid Stearic Acid (2-amino-5-chlorophenyl)(2-fluorophenyl)- Norleucine methanone Stearic Acid Homoserine phenyl ester-2-thiophenecarboxylic acid Glycine phenylethylmalonamide n-pentadecan-1-ol 3-Methyl-3-(N-methyl-2-pyrrolyl)-1,2- D-pinitol diphenylcyclopropene 5-amino-1,2,3,8-tetramethoxy- N-methoxycarbonyl-,isohexyl ester-l-valine deibenz[d,f]cycloheptane Furan-2-carboxylic acid [2-(2,2,6,6-tetramethyl- 3,5-Di-tert-butyl-4-hydroxybenzoic acid piperidin-4-yl)-ethyl]-amide 1,3-Dimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H- 1-ethyl-1H-Benzimidazole purine-8-carbaldehyde-oxime p-Cyanophenyl p-(2-butoxyethoxy)benzoate Phosphoric acid 2-fluorophenyl 2,4-dimethylpent-3-yl ester-succinic Tyrosine acid Homoserine Sarcosine Pre-Incubation Post-Incubation 804

40 bioRxiv preprint doi: https://doi.org/10.1101/115808; this version posted March 10, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

805 806 807 808 809 810 811 812 813

41