Spatial Variation in Concentration and Sources of Organic Carbon in the Lena

1 Spatial variation in concentration and sources of organic carbon in the Lena

2 River, Siberia

3

4 Liselott Kutscher1,2*, Carl-Magnus Mörth1, Don Porcelli3, Catherine Hirst1,2, Trofim C.

5 Maximov4,5, Roman E. Petrov 4,5, Per S. Andersson2

6 1Department of Geological Sciences, Stockholm University, SE-10691 Stockholm, Sweden.

7 2Department of Geosciences, Swedish Museum of Natural History, SE-10405 Stockholm, Sweden

8 3Department of Earth Sciences, University of Oxford, Oxford, UK

9 4Institute for Biological Problems of Cryolithozone of Siberian Branch of Russian Academy of Science

10 5Institute for Natural Sciences of North-Eastern Federal University, Yakutsk, Russia

11 *Corresponding author: Liselott Kutscher, Deperatment of Geological Sciences, Stockholm University,

12 SE-106 91 Stockholm, Sweden ([email protected])

13

14 Key points:

15  Spatial differences in OC concentrations and export due to catchment topography and

16 regional hydroclimate in Lena River basin

17  Sources of DOM were fresh plant material and SOM with seasonal patterns reflecting

18 differences in water flow pathways

19  Sources of POM were aquatic primary production and SOM, and were related to

20 topography

1 21 Abstract

22 Global warming in permafrost areas is expected to change fluxes of riverine organic carbon (OC)

23 to the Arctic Ocean. Here OC concentrations, stable carbon isotope signatures (δ13C) and carbon-

24 nitrogen ratios (C/N) are presented from 22 sampling stations in the Lena River and 40 of its

25 tributaries. Sampling was conducted during two expeditions: the first in July 2012 in the south

26 and southeastern region and the second in June 2013 in the northern region of the Lena basin.

27 The data showed significant spatial differences in concentrations and major sources of OC. Mean

28 sub-catchment slopes were correlated with OC concentrations, implying that mountainous areas

29 in general had lower concentrations than lowland areas. δ13C and C/N data from tributaries

30 originating in mountainous areas indicated that both dissolved and particulate OC (DOC and

31 POC) were mainly derived from soil organic matter (SOM). In contrast, tributaries originating in

32 lowland areas had larger contributions from fresh vegetation to DOC, while aquatically produced

33 OC was the major source of POC. We suggest that these differences in dominant sources

34 indicated differences in dominant flow pathways. Tributaries with larger influence of fresh

35 vegetation probably had surficial flow pathways, while tributaries with more SOM influence had

36 deeper water flow pathways. Thus, the future export of OC to the Arctic Ocean will likely be

37 controlled by changes in spatial patterns in hydroclimatology and the depth of the active layers

38 influencing the dominant water flow pathways in Arctic river basins.

39

2 40 1 Introduction

41 Permafrost soils are considered to be among the largest terrestrial reservoirs of carbon [Hugelius

42 et al., 2014]. During the last century, large areas underlain by permafrost in the Northern

43 Hemisphere have experienced environmental changes due to climate warming. Permafrost

44 temperatures have risen [Romanovsky et al., 2010], the active layer has become deeper during

45 summer months [Frey and McClelland, 2009; Zhang et al., 2005], and freshwater runoff from

46 rivers to the Arctic Ocean has increased [Costard et al., 2007; Peterson et al., 2002; Rachold et

47 al., 1996; Shiklomanov and Lammers, 2009]. Some of the largest Arctic Rivers have also had

48 increased sediment export [Rachold et al., 1996]. According to climate models, climate warming

49 in permafrost regions will continue during the next century and the Arctic is predicted to

50 experience one of the largest temperature increases in the world [Collins and Krinner, 2013]. It is

51 expected that permafrost thawing will change fluxes of natural organic matter (NOM) from the

52 terrestrial environment to freshwater and marine systems due to changes in discharge, transport

53 pathways, residence times in soils and water, plant species composition, and increased erosion

54 [Costard et al., 2007; Frey and Smith, 2005; Frey and McClelland, 2009; Guo et al., 2007; Neff

55 and Hooper, 2002; Striegl et al., 2005]. However, there is no consensus regarding the direction of

56 these changes [Frey and McClelland, 2009]. Whatever the direction of changes, it will ultimately

57 have implications for the carbon cycling and ecosystem functioning in Arctic freshwater systems

58 as well as in the Arctic Ocean [Frey and McClelland, 2009; Keller et al., 2010], since NOM

59 plays an important role in the aquatic food web as nutrient and energy source as well as transport

60 media of essential metals and dissolved nutrients [Jansson et al., 2007; Tipping, 2002].

3 61 Riverine NOM is a terrestrially (allochtonous) or aquatically (autochthonous) produced

62 heterogeneous mixture of organic compounds with a generally complex structure containing

63 carbon as its major constituent [Finlay and Kendall, 2008; Lam et al., 2007]. Due to the high

64 content of carbon, NOM is often equated with organic carbon (OC) and can be divided into two

65 size fractions: dissolved organic carbon (DOC) and particulate organic carbon (POC). DOC in

66 stream ecosystems is generally derived from terrestrial sources, e.g. litterfall and soil leachate,

67 while POC primarily is derived from eroded soil and sedimentary rocks as well as from

68 autochthonous sources, e.g. macrophytes, algae and heterotrophic bacteria [Galy et al., 2015;

69 Meybeck, 1993]. However, the division between DOC and POC is arbitrary, since DOC can

70 flocculate and form POC; also, both DOC and POC can be mineralized along river networks [del

71 Giorgio and Pace, 2008; Drake et al., 2015; Kerner et al., 2003; Meybeck, 1993].

72 Several different factors controlling OC concentrations in rivers have been suggested. Mean OC

73 concentrations have been correlated to spatial variables such as topography and landscape

74 characteristics, e.g. amount of upstream wetland cover, lake cover, and average channel slope

75 [Eckhardt and Moore, 1990; Mulholland, 1997; Mulholland, 2003]. Factors controlling temporal

76 DOC dynamics have been shown to include discharge [Eckhardt and Moore, 1990; Winterdahl et

77 al., 2014], soil temperature, local climate (air temperature and precipitation) and antecedent flow

78 [Winterdahl et al., 2011]. OC concentrations in rivers often peak in connection with high

79 discharge events; this is commonly explained to be an effect of shifting flow pathways from

80 deeper soil horizons to shallower organic-rich soil layers [Mulholland, 2003]. For example, in

81 rivers in permafrost areas and in many boreal rivers, OC concentrations often increase during

82 snowmelt and decrease with decreasing flow so that the lowest concentrations commonly occur

83 during base flow [Holmes et al., 2012; Le Fouest et al., 2013; Raymond et al., 2007; Winterdahl

4 84 et al., 2014]. A warming climate is expected to shift dominating flow paths to deeper soil layers

85 [Frey and McClelland, 2009]. However, the subsurface hydrology in permafrost is not fully

86 understood [Watson et al., 2013; Woo et al., 2008] and limits our understanding of element

87 transport from areas underlain by permafrost.

88 For a better understanding of the influence of permafrost thawing on export and fluxes of riverine

89 carbon it is necessary to unravel the main contributing areas and sources of OC in permafrost

90 regions. Among the world’s deepest layers of permafrost are found in the Lena River drainage

91 basin, in central Siberia, where the major part of the basin is underlain by continuous permafrost

92 [Chevychelov and Bosikov, 2010]. It is the ninth largest drainage basin in the world and has the

93 largest transport of OC to the Arctic Ocean [Holmes et al., 2012; Le Fouest et al., 2013]. Holmes

94 et al. [2012] estimates that the annual export from the six largest Arctic rivers draining directly

95 into the Arctic Ocean is 25 Tg DOC year-1, of which the Lena River contributes approximately

96 5.8 Tg DOC year-1 and 0.8 Tg POC year-1 [Holmes et al., 2012; Le Fouest et al., 2013;

97 McClelland et al., 2016; Raymond et al., 2007]. The export of OC from the Lena varies

98 seasonally and OC concentrations peak during snowmelt in the beginning of June and then

99 decrease to reach annual lows during winter baseflow [Amon et al., 2012; Holmes et al., 2012; Le

100 Fouest et al., 2013; McClelland et al., 2016; Raymond et al., 2007]. Previous studies in the Lena

101 River indicate that the dominant sources of DOC in the main channel are relatively young soil

102 and plant material [Amon et al., 2012; Lara et al., 1998; Lobbes et al., 2000]. POC on the other

103 hand is suggested to be older degraded soil organic matter (SOM) [Lobbes et al., 2000;

104 McClelland et al., 2016]. The amount of aquatically produced OC in the Lena River is believed

105 to be low [Lobbes et al., 2000; McClelland et al., 2016; Sorokin and Sorokin, 1996]. The

106 contribution of old, highly biologically available OC from thawing permafrost to fluvial networks

5 107 increase aquatic microbial metabolism and CO2 emissions within Arctic drainage basins [Mann et

108 al., 2015]. This permafrost-derived OC is also suggested to result in loss of ancient OC in river

109 networks and a selective export of younger OC to the oceans [Mann et al., 2015; Vonk et al.,

110 2010].

111 Stable carbon isotopes values (δ13C) and carbon/nitrogen atomic ratios (C/N) can be used to

112 identify sources of freshwater OC, since these could be indicative of different sources [Cleveland

113 and Liptzin, 2007; Elser et al., 2000; Finlay and Kendall, 2008; McGroddy et al., 2004; Sterner

114 and Elser, 2002]. Freshwater primary producers assimilate carbon from CO2, which can originate

115 from a variety of sources and result in a wide range in δ13C signatures of aquatically produced

116 OC (-47 to -8‰) [Farquhar et al., 1989; Finlay and Kendall, 2008]. In addition, the isotope

117 fractionation factor of freshwater primary production is more varied than in terrestrial systems

118 [O'Leary, 1988]. C/N ratios in aquatically produced material are often narrow in range (4-15)

119 [Elser et al., 2000; Finlay and Kendall, 2008], which could be due to the low supply of nitrogen

120 in aquatic systems [Sardans et al., 2012]. In contrast, terrestrial sources generally have a

121 narrower range in δ13C and a wider range in C/N [Elser et al., 2000; Finlay and Kendall, 2008;

122 McGroddy et al., 2004; O'Leary, 1988]. Terrestrial plants mainly assimilate carbon from the

123 atmosphere, which is a well-mixed reservoir with a δ13C of -8‰. The isotope fractionation factor

124 in C3 plants, the most common plant type [Ridge, 2002], is about -20‰, resulting in an average

125 δ13C of approximately -28‰ in C3 plants [Farquhar et al., 1989; Finlay and Kendall, 2008;

126 O'Leary, 1988]. C/N in plants and litter worldwide varies substantially and could range from

127 around 15 to several hundred, but reported mean values of C/N are between 36 and 66 [Elser et

128 al., 2000; Finlay and Kendall, 2008; McGroddy et al., 2004]. It is commonly observed that δ13C

129 increase while C/N decrease with depth in mineral soils in boreal and arctic ecosystems [Boström

6 130 et al., 2007; Cleveland and Liptzin, 2007; Kaiser et al., 2007; McGroddy et al., 2004; Palmtag et

131 al., 2015; Zech et al., 2007], and is believed to be due to isotope fractionation and selective

132 removal during SOM decomposition [Boström et al., 2007; Ehleringer et al., 2000; Palmtag et

133 al., 2015].

134 Previous studies of OC in the Lena River have mainly focused on the main channel of the river

135 and the delta [Amon et al., 2012; Cooper et al., 2008; Dudarev et al., 2006; Holmes et al., 2012;

136 Karlsson, 2015; Lara et al., 1998; Le Fouest et al., 2013; Lobbes et al., 2000; Pipko et al., 2010;

137 Raymond et al., 2007; Semiletov et al., 2011]. OC data from tributaries of the Lena River are

138 scarce and include only the largest sub-basins and then only for a few occasions [Kuzmin et al.,

139 2009; Lara et al., 1998]. Spatially defined OC fluxes from the terrestrial environment to the Lena

140 River are therefore not well understood, limiting our ability to adequately predict future changes

141 in OC concentrations and fluxes to the Arctic Ocean.

142 In this study, we present a spatial data set of OC concentrations, C/N ratios, and δ13C values of

143 DOC and POC, from the Lena River main channel and 40 of its tributaries sampled during two

144 field expeditions conducted in July 2012 and June 2013. These catchments cover different

145 landscape types with varying vegetation, lithology, and hydrology, as well as permafrost depth

146 and distribution. The specific objectives of this study were to enhance our understanding of

147 spatial variability in OC concentration and sources in the Lena basin by (i) quantifying spatial

148 patterns in OC concentrations and composition, (ii) making a preliminary spatial export estimate

149 of OC from the largest sub-basins of Lena and (iii) identifying major sources of OC in rivers in

150 the Lena basin.

7 151 2 Methods

152 2.1 Study area

153 The area of the Lena River basin is about 2.4 million km2 and extends from its southernmost

154 headwater at latitude 52-53°N in the mountain areas of Baikal, Stanovoi and Dzhugdzhur to its

155 delta at the shallow shelf in the Laptev Sea at 73°N (Figure 1a). The eastern part of the drainage

156 basin is dominated by the mountain ranges Verkhoyansk and Suntar-Khaiata, with its easternmost

157 headwaters at longitude 141°E. The central and northwestern areas are characterized by plains

158 and plateaus, which make the western water divide diffuse in comparison to the mountain ranges

159 in the south and the east. The westernmost headwaters extend to the Viliuskoe plateau as far west

160 as longitude 104°E.

161 The Lena River basin is characterized by a continental climate, with long, cold winters and short,

162 warm summers [Chevychelov and Bosikov, 2010]. The mean annual air temperatures are in

163 general low and decrease with increasing latitudes [Fedorov et al., 2014], from approximately

164 -5°C in the southern parts of the Lena basin, -9°C at Viliuisk in the central part of the basin, and

165 around -13°C in the north at Kiuisur [Fedorov et al., 2014]. The major part of the Lena basin is

166 underlain by continuous permafrost, except for patches of discontinuous and sporadic permafrost

167 in the southern parts of the basin and some areas along the main channel of the Lena River

168 [Chevychelov and Bosikov, 2010]. In the central part of the Lena basin, the Central Siberian

169 Plateau (CSP), the deepest layers of permafrost are found [Chevychelov and Bosikov, 2010]. Both

170 air and permafrost temperatures in the basin show increasing trends since the late 1950s

171 [Romanovsky et al., 2007], which have resulted in an estimated deepening of the active layer of

172 0.35 cm/year [Frey and McClelland, 2009].

8 173 The permafrost is primarily overlaid by podzols and leptosols in the south, cambisols in the

174 central basin, and gleysols in the northernmost region [Stolbovoi and McCallum, 2002]. The

175 vegetation is divided into three zones, where the most productive forest is found in the

176 southernmost area. Most of the basin is dominated by larch forest, while alpine regions have bare

177 rock terrain [Chevychelov and Bosikov, 2010].

178 The mean annual precipitation in the southern and southwestern part of the Lena basin is ~350-

179 500 mm/year and ~200-250 mm/year elsewhere [Chevychelov and Bosikov, 2010]. Monthly

180 discharge has been measured at several hydrologic stations since the 1930s by the Russian

181 Hydro-meteorological Services. These data show a strong seasonal flow regime in the Lena River

182 and its tributaries, with the highest discharge peak at snow melt in May to June (Figure 1c) [Ye et

183 al., 2009]. Discharge gradually decreases after spring flood and reaches the annual low during

184 winter. The average in-stream water travel time is 1-2 months from the uppermost regions of the

185 basin to the outflow of Lena [Smith and Pavelsky, 2008; Ye et al., 2003].

186 2.2 Field methods and sampling

187 River water samples were collected at 77 locations across the Lena River basin during two field

188 expeditions conducted with the ships R/V Geokryolog and R/V Akademic in July 2012 and with

189 R/V Merelotoved in June 2013 (Figure 1, Table S1). The first expedition covered the south and

190 southeastern region of the Lena basin, upstream of the confluence of Lena and Aldan rivers, and

191 the second expedition covered the region downstream of Yakutsk. The samples were collected at

192 22 stations in the Lena River main channel, between the Toulba River (60.6°N) and Dzhardzhan

193 River (68.7°N), and at 55 stations in tributaries (Figure 1 and Table S1). In most tributaries the

194 outflow of the rivers were sampled, but in the largest tributaries (the Aldan and Viliui Rivers)

9 195 additional samples were taken along their main channels. In total 40 different tributaries were

196 sampled. Sampling was repeated in the second expedition in the outflow of Aldan and some of

197 the other tributaries by the confluence of Aldan and Lena River as well as in the Lena River

198 between Yakutsk and the outflow of Aldan. In addition to water samples, plant samples were

199 collected at three different locations along the Lena River. At the end of the 2012 expedition a

200 flood took place in the upstream areas of the Lena basin resulting in elevated water levels and

201 large amounts of debris in the Lena River. The average discharge in the end of July at the

202 gauging station Tabaga (close to Yakutsk) is generally around 10000 m3s-1, but at the flood in

203 2012 the discharge rose to almost 40000 m3s-1, which is almost as high as the 2012 spring flood

204 [Tananaev 2016].

205 At each sample station, surface water was collected in 10 L low density polyethylene (LDPE)

206 bottles by grab sampling. Water temperature, pH and conductivity were measured in situ with an

207 YSI 556 multi probe system, with accuracies at ±0.15°C, ±0.2 pH units and ±0.5% respectively.

208 These variables were also measured along a transect from sample station 73 (Figure 1b) to the

209 eastern shoreline of the Lena main channel in 2013 to investigate mixing patterns within the

210 river. Positions were recorded with a hand-held GPS receiver.

211 Water samples were filtered through 25 mm pre-combusted (at 450°C for 2 hours) glass fiber

212 filters (Whatman® GF/F) with 0.7 µm pore size using a portable peristaltic pump. The mass of

213 water pumped through the filters was recorded. The GF/F-filters were then folded into packets of

214 aluminum foil and stored in freezers until analysis of particulate carbon and nitrogen. Filtered

215 water was stored in dark high density polyethylene bottles for analyses of DOC and dissolved

216 nitrogen. All bottles, vials and filter equipment used during field work were pre-washed with

10 217 10% HNO3 and distilled water (MilliQ 0.1µm filter). The filtered water samples collected in July

218 2012 were preserved with phosphoric acid (87%) and stored cool, while samples collected in

219 June 2013 were stored in freezers until analyses. Plant samples were dried and packed in LDPE

220 bags.

221 2.3 Analytical Methods

222 Concentrations of DOC were determined within two months by high-temperature catalytic

223 oxidation (Shimadzu TOC-VCPH) at the Department of Environmental Science and Analytical

224 Chemistry, Stockholm University (SU). Two internal controls were used; phenanthroline and

225 humic lake water with known carbon concentrations (18-19 mg/l). The total uncertainty interval

226 based on laboratory inter-comparisons was less than ±10%.

227 Dissolved water samples were freeze-dried at -110°C and then weighed and put into tin (2012) or

228 silver capsules (2013). A small aliquot of 2 M HCl (PA grade) was added to remove any

229 inorganic carbon in the samples and thereafter the samples were left in an oven at 60° C for 1

230 hour to allow evaporation of HCl. GF/F filters containing the POC and particulate organic

231 nitrogen (assuming that inorganic N was low and negligible) were placed in desiccators together

232 with a beaker with HCl (35%) overnight to remove inorganic carbon. The filters were then dried

233 at 60°C and packed in silver capsules. The dried plant material was also placed into silver

234 capsules. These prepared samples were measured with an isotope ratio mass spectrometer

235 (IRMS) (Finnigan® Delta V advantage) at the stable isotope laboratory at SU. The mass

236 spectrometer was connected to a CarloErba® NC2500 elemental analyzer through a ConfloIV

237 open split interface.

11 238 The organic carbon and total nitrogen mass received from the IRMS measurements were used to

239 calculate POC concentrations and C/N atomic ratios. The concentrations of POC were calculated

240 from the mass of particulate carbon on the filters and the weighted water pumped through the

241 GF/F filters. For the C/N ratios it was assumed that the total N was organic nitrogen. The total

242 accuracy and precision of the C/N atomic ratios and POC concentrations were lower than 3%.

243 Isotope values of DOC, POC and plant material were reported in per mil (‰), δ13C, relative Pee

244 Dee Belemnite (PDB):

푅 245 (1) 훿 = ( 푆푎푚푝푙푒 − 1) ∙ 1000 푅푆푡푑

13 12 246 where RSample and RStd are the C/ C ratios in the sample and in the standard respectively. The

247 standards used for carbon and nitrogen calibrations in the IRMS were internal standards with

248 known isotope compositions relative to PDB and AIR. The accuracy and precision of δ13C

249 measurements were less than ±0.2 ‰.

250 2.4 Export calculations

251 A first preliminary estimate of the export and area specific export of carbon for the period June to

252 July from different source areas were calculated from concentrations of DOC and POC from the

253 main channels of Upper Lena, Aldan River, Viliui River and Lower Lena, and long term monthly

254 discharge data from the hydrologic stations Lena at Tabaga (A), Aldan at Verkhoyanski’ Perevos

255 (B), Viliui at Hatyrik-Homo (C) and Lena at Kiusiur (D) (Figure 2 and Table 1). The discharge

256 data used in the calculations were downloaded from the Russian Hydro-meterological Services -

257 R-ArcticNet (http://www.r-arcticnet.sr.unh.edu/v4.0/index.html), ArcticRIMS

12 258 (http://rims.unh.edu/data.shtml) and Arctic Great Rivers Observatory (NSF-1107774)

259 (http://arctic greaterivers.org/data.html), which only consist of discharge data from the Lena main

260 channel and a few of the sampled tributaries The hydrologic station in Kiusiur was taken to

261 represent the outflow of the Lena basin and the mean OC concentrations from the two

262 northernmost sample stations 45 and 48 (Table S1), were used to estimate the total organic

263 carbon (TOC) export and area specific export in June and July to the Arctic Ocean. The average

264 TOC concentration for these two samples (990 µM) was similar to the annual mean value of the

265 Arctic-GRO data (970 µM) collected between 2003 and 2013 at Zhigansk. In export calculations

266 from sub-basins, mean OC concentrations from samples collected in the main channels of the

267 Aldan River (station 8,13 and 38 from July 2012 and June 2013) and the Viliui River (station 59,

268 60, 62 and 66 from June 2013) were used. In the Upper Lena, the mean OC concentrations from

269 stations 28, 30 and 31 collected in July 2012 were used (Table S1). Samples collected after the

270 flood in the Upper Lena were excluded in the export estimations, since the flood in 2012 was

271 substantially larger than average discharge for July [Tananaev 2016] and we do not have data to

272 assess how exceptional floods of this magnitude are in this area in July. The remaining TOC

273 export and area specific export were assigned to the northern region of the basin, here called the

274 Lower Lena (Figure 2).

275 POC concentrations commonly increase with depth in rivers [McClelland et al., 2016]. In this

276 study only surface water was sampled and the TOC concentrations and export could therefore

277 have been underestimated. What the differences are between the surface and deeper waters in the

278 rivers is not clear and so are the uncertainties related to that. Other uncertainties associated with

279 the export calculations were sampling in different regions during different months and years,

280 which means that most of the stations were only sampled once. A Monto-Carlo analysis was

13 281 therefore applied to estimate the uncertainties in the TOC export from the export equation in

282 Table 1. Random numbers (200,000 normally distributed) for TOC concentrations and discharge

283 values were generated assuming a total uncertainty of ± 20% for both TOC concentrations and

284 measured discharge. The error estimates are reported in Table 1.

285 2.5 Slope calculation

286 The catchment specific mean slope for each tributary and stations 32, 13 and 62 (representing the

287 upper regions of the Lena, Aldan and Viliui catchments) (Figure 1b) was estimated with the

288 Slope tool in the Spatial Analyst toolbox in ArcGIS 10, based on the digital elevation model

289 (DEM) GTOPO30 (https://lta.cr.usgs.gov/GTOPO30) with horizontal grid spacing of 30 arc

290 second.

291 2.6 Grouping of data

292 Spatial patterns of OC were studied by dividing the samples into groups based on geographic

293 region (Figure 1a, Table S1): (1) the main channel of the Lena River, (2) the main channel of the

294 Aldan River, (3) the main channel of the Viliui River, (4) tributaries in the mountain range of

295 Verkhoyansk (VER), (5) tributaries in the lowland area of Central Siberian Plateau (CSP) and (6)

296 tributaries in the Lena-Amginsky inter-river area (LAIRA). The Aldan and Viliui Rivers were

297 treated as separate groups due to their large catchment sizes and runoff to the Lena.

14 298 3 Results

299 3.1 Organic carbon concentrations

300 The total range in DOC concentrations for all sample stations was 90 to 3400 μM, with a median

301 concentration of 900 μM (Table 2). We observed spatial variations in OC concentrations, where

302 tributaries with headwaters in VER had much lower median DOC concentrations, 260 μM,

303 compared to the rest of the study area (Table 2, Figure 3a). Conversely, tributaries with

304 headwaters in the CSP and the main channel of the Viliui River had substantially higher median

305 concentrations, 1200 μM and 1000 μM, respectively (Table 2, Figure 3a). We found a significant

306 negative correlation between mean catchment slope and DOC concentrations with the catchments

307 with higher mean slope having the lowest DOC concentrations (Figure 4a).

308 In samples collected in the Lena main channel, concentrations of DOC varied between 590 and

309 1300 μM, with a median concentration of 970 μM (Table 2, Figure S1a). Samples collected

310 between the outflow of the Toulba River (station 33) and Aldan River, i.e. in the upstream part of

311 the Lena River, had a narrower range (840 to 1000 μM) compared to the samples collected

312 further downstream (590 to 1300 μM) (Figure S1a). The large variations in DOC concentrations

313 downstream from the confluence of Aldan River showed lower concentrations on the eastern

314 side, the Verkhoyansk side, compared to those collected on the western side, the Central Siberian

315 Plateau side (Figure S2a). The same pattern was also seen in conductivity, with higher

316 conductivity on the CSP side (Figure S2b). In addition, observations from the transect across the

317 Lena River showed a gradient in conductivity, temperature and pH in the main channel (Figure

318 S3). Values for all three variables were higher on the western side than on the eastern side of the

319 river.

15 320 The concentrations of POC were consistently much lower than those of DOC in the entire basin

321 (Table 2, Figure 3a), and were on average 5% of the TOC. The median POC concentration for all

322 samples stations was 38 μM with a total range of 6 to 640 μM (Table 2). The POC concentrations

323 in the main channel of Lena River varied between 10 and 640 μM, with a median value of 42 μM

324 (Table 2, Figure S1b). This large variation was caused by two outliers (sample 32 and 34)

325 collected after the flooding in the southern part of Lena River at the end of the expedition in July

326 2012. During the flood, POC concentrations increased with approximately 500 μM in the main

327 channel of Lena River (Figure S1b). There was a significant difference in POC concentrations

328 among groups, with a similar pattern as for DOC concentrations (Figure 3a), and a similar

329 correlation between POC and slope (Figure 4b).

330 3.2 Transport and Export of TOC

331 The export of TOC in June and July from the Lena basin was estimated to be 3.54 Tg TOC,

332 where 3.46 Tg was DOC (98%) and 0.08 Tg POC (2%) (Table 1, Figure 2). The upstream areas

333 from the hydrologic stations Lena at Tabaga (Upper Lena), Aldan at Verkhoyanski’ Perevos and

334 Viliui at Hatyrik-Homo were estimated to contribute 1.35, 0.84 and 0.34 Tg TOC for the period,

335 respectively. The remaining contribution from the lower northern region of the Lena River basin

336 was 1.01 Tg TOC. The area specific export rate in June and July in the Lena River basin, at

337 Kiusiur, was estimated to be 1.5 g m-2. The rate from the Upper Lena was the same as in Kiusiur,

338 while the rate from the Aldan River was lower, (1.2 g m-2) and the Lower Lena was higher (2.6 g

339 m-2). The area specific export rate in Viliui River was considerably lower for the period

340 compared to all other source regions (0.8 g m-2).

16 341 3.3 δ13C values

13 13 342 The δ C values for DOC, δ CDOC, had a median value of -27.2‰ with a range of -28.4‰ to -

13 343 25.7‰ (Table 2, Figure 3b). There was a significant difference in δ CDOC between sampling

13 344 years, with lower δ CDOC in the samples from June 2012 compared to samples from July 2013

13 345 (Figure 5a). There were also significant geographical differences in δ CDOC among groups

346 (Figure 3b). A comparison for each pair of groups showed that the Viliui River and the tributaries

13 347 with headwater in CSP had lower δ CDOC compared to other regions and main channels (Figure

13 348 3b). A negative correlation was found between δ CDOC and latitude (p<0.0001 ρ= -0.66), and

13 349 positive correlation between δ CDOC and longitude (p<0.0001, ρ =0.52). The combined data

13 350 shows that samples in the north and west of the study had the lowest δ CDOC.

13 13 13 351 The range in δ C values for POC, δ CPOC, (-36.1‰ to -24.3‰) was larger than for δ CDOC

352 while the median value was lower (-29.2‰) (Table 2, Figure 3b). There were significant

13 353 differences in δ CPOC among geographical areas and the comparison for each pair showed that

13 354 the Viliui River, CSP, and LAIRA tributaries had significantly lower δ CPOC compared to other

13 355 groups (Figure 3b). The δ CPOC were lower in tributaries with low mean catchment slope

356 compared to catchment with higher mean slope (Figure 4c). Additionally, a significant

13 357 correlation was found between δ CPOC and water temperature (Figure S4). Water temperatures

358 were also correlated with mean catchment slope (p<0.0003 ρ =- 0.52). Tributaries with lower

13 359 mean catchment slope thus generally had lower δ CPOC and higher water temperatures, while

13 360 tributaries originating in mountainous areas had higher δ CPOC and lower water temperatures.

13 13 361 The δ C values in plant samples, δ Cplant, of Carex, Betula, Larix and Salix varied between -

362 29.9‰ and -25.4‰, with a median value of -29.6‰.

17 363 3.4 C/N atomic ratios

364 The median C/N ratios of the dissolved fraction of organic matter, C/NDOM, was 28 and the range

365 was large (4 to 41) with significantly lower values in samples collected in July 2012 compared to

366 samples collected in June 2013 (Table 2, Figure 5a). The same pattern was shown in the main

367 channel of the Lena, with lower C/NDOM in the upstream area which was sampled in 2012 and

368 higher C/NDOM in the downstream area that was sampled in 2013 (Figure S1e). There was also a

369 significant difference among all groups, with higher median values in the Viliui (34) and CSP

370 (29) compared to the other groups (19-21) (Table 1, Figure 3c). Where C/NDOM were high, there

13 371 was in general lower values of δ CDOC (p<0.0001, ρ=-0.58). There was positive correlation

372 between C/NDOM and latitude (p<0.0001 ρ= 0.59), and negative correlation between C/NDOM and

373 longitude (p<0.0001, ρ = -0.55). This means that the further North and West the samples were

374 collected in the Lena basin, the higher was the C/NDOM, while samples collected in the southern

375 and eastern regions had lower C/NDOM.

376 The range in C/N ratios in the particulate organic matter, C/NPOM, was not as large as the range in

377 C/NDOM. Observed values were from 5 to 17 with a median of 10 (Table 2 and Figure 3c).

378 Median C/NPOM were largest in the main channels of the Lena River and the Aldan River. There

379 were no significant differences among geographical areas.

380 C/N ratios for plant samples, C/Nplant, ranged between 26 and 42, with a median of 36 (Table 2,

381 Figure 6).

18 382 4 Discussion

383 4.1 Organic carbon concentrations

384 DOC concentrations in the main channel of the Lena River (590-1300 µM) were similar to

385 previous observations in the Lena River (170-1230 µM) and comparable to the range in other

386 Arctic Rivers [Amon et al., 2012; Cauwet and Sidorov, 1996; Cooper et al., 2008; Kuzmin et al.,

387 2009; Lara et al., 1998; Le Fouest et al., 2013; Lobbes et al., 2000]. However, DOC data from

388 the tributaries had an even larger range (Figure 3a). The range in POC concentrations was larger

389 (6-640 μM) compared to other studies in the Lena River (2.5-430 μM), but similar to the range in

390 other Arctic Rivers [Cauwet and Sidorov, 1996; Guo et al., 2007; Kuzmin et al., 2009; Le Fouest

391 et al., 2013; Lobbes et al., 2000; McClelland et al., 2016]. The large range in POC was primarily

392 due to the flood event in the Upper Lena in 2012. The flood increased POC concentration by

393 approximately 500% with a simultaneous increase in discharge (Figure S1b). At Tabaga,

394 discharge increased to about 40000 m3s-1, which is almost 30000 m3s-1 more than the average

395 flow in the end of July and of the same magnitude as spring floods [Tananaev 2016]. At Kiusiur

396 discharge increased from 30,000 to 46,000 m3 s-1 as observed 17 days later, which was the

397 approximate in-river water travel time at this period (Figure S1). Notably, there was no

398 concurrent change in DOC concentrations in the main channel of Lena River following the flood.

399 The observed differences in OC concentrations, conductivity, water temperatures and pH across

400 the Lena main channel suggest incomplete mixing in the main channel of the Lower Lena.

401 Downstream the outflow of Aldan River, the Lena River main channel was wide and had a

402 meandering deep channel with patches of islands in the channel that could act as barriers

403 restricting mixing of the water. The incomplete mixing of Lena River might also have been a

19 404 result of density differences between water from the CSP and the VER tributaries due to different

405 sediment loads.

406 A correlation between mean slope and both DOC and POC concentrations, seen in the Lena

407 tributaries (Figure 4a, 4b), has previously been reported by Mulholland [1997], who observed a

408 strong correlation between average channel slope and DOC concentration in data from 33 North

409 American catchments, along a climatic gradient from Puerto Rico to Alaska including streams

410 with stream orders from 1 to 7. Recently a study by Jasechko et al. [2016], comparing global

411 watersheds that included permafrost dominated regions, showed that the stream water in

412 mountainous catchments in steep terrain generally have a higher proportion of older water

413 (>2.3±0.8 month old) compared to lower relief landscapes. As the proportion of young versus old

414 water is a proxy for shifting pathways in stream flow and travel times, it suggests that areas with

415 steeper relief have deeper vertical infiltration and are more influenced by ground water [Jasechko

416 et al., 2016; McGuire et al., 2005]. In permafrost areas, the permafrost could restrict deeper

417 percolation by acting as semi-impermeable aquitards. When hydrologic flow paths are forced to

418 surface layers of soils due to restrictions in vertical drainage, the interaction with organic rich

419 surface soil layers lead to higher DOC concentrations [Laudon et al., 2011; Laudon et al., 2012;

420 Mulholland, 1997]. As the distribution of permafrost in steep terrain vary depending on aspect,

421 elevation and regional hydroclimate, a small-scale mosaic of permafrost and permafrost-free

422 areas could shift the hydrology on local scale [Woo et al., 2008] and thus the OC concentrations.

423 Our results are consistent with deeper flow pathways dominating mountainous catchment

424 drainage compared to that of catchments with lower relief. Deeper flow pathways will likely

425 transfer water through mineral soils allowing substantial adsorption and degradation of organic

426 matter and consequently lower stream DOC and POC concentrations [Kaiser and Kalbitz, 2012;

20 427 Klaminder et al., 2011]. In the VER region, headwaters are located in alpine areas but reach

428 lowland areas with wetland type soils before discharging into the Lena River. Wetland areas are

429 generally key contributors of DOC to aquatic systems due to the high content of SOM [Eckhardt

430 and Moore, 1990; Laudon et al., 2011; Mulholland, 1997]. However, the wetland areas did not

431 seem to have a major influence on DOC concentrations in the VER tributaries, which remain low

432 (Figure 3a). Similarly, tributaries to the Aldan River with headwaters primarily in mountainous

433 areas, which have steeper slopes, had lower DOC and POC concentrations compared to the other

434 stations in the Aldan catchment (Figure 3a). In addition, lower OC concentrations in rivers

435 draining mountainous areas could be due to less soil organic carbon in these regions and less

436 extensive tree cover at high altitudes [Chevychelov and Bosikov, 2010; Hugelius et al., 2014].

437 The Viliui River and its tributaries, which drain the flat lowland area of CSP, had high

438 concentrations of DOC and POC (Figure 3a). The CSP is mainly underlain by continuous

439 permafrost [Chevychelov and Bosikov, 2010], and may have hydrologic flow path restricted to

440 more organic rich superficial soil layers. Catchment with large riparian wetlands that in general

441 have low slopes and broad floodplains, commonly have high concentrations of DOC

442 [Mulholland, 2003]. A recent study by Boike et al. [2016] reported that the area close to the

443 Viliui River in generally became wetter and increased its lake distribution between 2002 and

444 2009, even though some parts had a shrinking lake distribution. An increase in wetlands may

445 result in even greater OC concentrations in CSP tributaries.

446 4.2 Export of OC

447 The calculated export of TOC in June and July from the Lena River to the Arctic Ocean, 3.5 Tg

448 TOC (Figure 2, Table 1), was more than half of the annual TOC export estimated by previous

21 449 studies (6.6 Tg TOC year-1) [Le Fouest et al., 2013]. DOC in June and July dominated the

450 organic carbon transport; TOC consisted of 98 % DOC, which is a common pattern for Arctic

451 and boreal rivers [Gordeev and Kravchishina, 2009; Le Fouest et al., 2013; Lobbes et al., 2000;

452 Meybeck, 1982].

453 The TOC export from the Upper Lena and the Aldan and Viliui Rivers, showed that these areas

454 contributed 34%, 27% and 10% of the TOC, respectively (Figure 2). The Lower Lena contributed

455 29% of the TOC transported from the Lena River to the Arctic Ocean. Since discharge varies by

456 several orders of magnitude (both spatially and temporally) while OC concentrations rarely vary

457 by more than a factor of 10 [Godsey et al., 2009], export estimates were probably dominated by

458 variations in discharge (Table 1), and the higher contribution to TOC from the Upper Lena

459 (Figure 2) could be explained by a higher mean precipitation in this region [Chevychelov and

460 Bosikov, 2010]. Compared to the other sub-basins the Upper Lena also has more sporadic and

461 discontinuous permafrost, which have been shown to contribute with higher fluxes and

462 concentrations of river OC [Frey and McClelland, 2009]. This area also has variable soil and

463 topographical characteristics, the mildest climate and the most productive forest in the Lena basin

464 [Kuznetsova et al., 2010]. These features may influence OC dynamics both through variable OC

465 sources and through its influence on water flow pathways, which have been shown to partly

466 control OC concentrations in streams and rivers [Mulholland, 2003; Winterdahl et al., 2011]. In

467 contrast, the export of TOC from the Viliui River was low (Figure 2), although DOC

468 concentrations were high (Figure 2a). The area specific export of TOC was significant lower

469 compared to other parts of the Lena basin (Table 1, Figure 2). A large dam in the upstream area

470 of the Viliui catchment reduces discharge at the mouth of the Viliui River during summer peak

22 471 flow by 20-25 % [Ye et al., 2003], which partly could explain the low export of TOC from that

472 area.

473 Previous studies of seasonal OC dynamics in the Lena main channel have shown that discharge

474 and OC concentrations peak during spring flood and reach annual lows during winter [Holmes et

475 al., 2012; Le Fouest et al., 2013; Raymond et al., 2007]. This behavior is typical for both DOC

476 and POC dynamics in Arctic and boreal rivers [Eckhardt and Moore, 1990; Holmes et al., 2012;

477 Le Fouest et al., 2013; Mulholland, 2003; Winterdahl et al., 2014]. Water samples collected

478 earlier in the summer could thus have higher concentrations compared to samples collected later

479 in the summer. Sampling during two different months and in two different years, the fact that

480 most sample stations were only sampled once and that the POC likely was underestimated were

481 uncertainties that likely have influenced our export estimates. Since spatial patterns in OC

482 transport within the Lena River basin currently are not well understood, our export calculations

483 could be considered as a first preliminary estimate of spatial OC export in the area. To reduce the

484 uncertainties in the carbon export calculations, due to the incomplete mixing in the Lena River,

485 the average concentration of the two most downstream sample stations in the Lena main channel

486 were used. One sample was collected at the eastern side of the main channel whereas the other

487 was collected on the western side.

488 4.3 δ13C values and C/N atomic ratios

489 The δ13C values and the C/N ratios of freshwater OC vary depending on sources and

490 biogeochemical pathways [Cleveland and Liptzin, 2007; Finlay and Kendall, 2008; McGroddy et

491 al., 2004; Sterner and Elser, 2002]. These parameters could therefore be used to identify sources

492 or processes controlling OC. The results in this study showed different patterns in δ13C and C/N

23 13 493 between DOC and POC (Figure 6). In general, there was a narrow range in δ CDOC, but a wide

13 494 range in C/NDOM. In contrast, there was high variation in δ CPOC and a low variation in C/NPOM.

495 These patterns indicated different dominant sources and water flow paths influencing the OC

13 496 composition. The DOC composition agreed with the range in both δ C values and C/N ratios for

497 the plant material collected in this study and SOM in the Tamara river catchment (central part of

498 Lena basin) reported by Zech et al. [2007] (Figure 6). The majority of the dissolved samples in

499 this study had C/NDOM higher than what is typical for autochthonous sources, thus excluding

500 these as main sources of DOC. This indicated that riverine DOC was primarily a mixture of

501 relatively fresh plant material and SOM. Previous studies in the main channel of the Lena River

502 and other Arctic Rivers based on organic biomarkers and 14C data suggest that the major source

503 of DOC in the spring freshet is primarily relatively fresh vegetation material [Amon et al., 2012;

504 Lara et al., 1998; Lobbes et al., 2000].

505 The sources of DOC could be treated as a mixture between different distinctive sources (Figure

506 5a). Since fresh plant material primarily is found at the soil surface and in surficial soil layers

507 while SOM is derived from deeper soil layers, these results are consistent with water flow paths

13 508 through fresh organic material in top soils in catchments with high C/NDOM and low δ CDOC,

13 509 while the OC in catchments with low C/NDOM and high δ CDOC values was influenced more by

510 SOM and deeper water flow paths (Figure 5a, b). Studies of Siberian soil profiles have shown

511 higher C/N in surficial soil layers compared to deeper soil layers [Gittel et al., 2014; Hugelius et

512 al., 2012; Menyailo and Hungate, 2006; Palmtag et al., 2015; Zech et al., 2007] and lower δ13C

513 values (-28 to -25‰) in surficial soil layers [Kaiser et al., 2007; Menyailo and Hungate, 2006;

514 Zech et al., 2007]. The processes causing isotope fractionation of SOM are still not fully

515 understood, but are believed to be associated with decomposition or diagenesis. For instance,

24 516 during decomposition of plant litter, microbial organisms favor the light isotope (12C) and

517 specific components of SOM resulting in increasing δ13C values [Boström et al., 2007;

518 Ehleringer et al., 2000]. In addition, the release of CO2 during aerobic consumption of SOM can

519 result in lower SOM C/N ratios. In general, the C/N is lower in SOM than in terrestrial plant

520 material [Cleveland and Liptzin, 2007; McGroddy et al., 2004], suggesting the importance of

521 organic matter microbial recycling in soils.

13 522 There were differences in C/NDOM and δ CDOC values (Figure 5a) among sampling months and

523 years. Water samples collected early in the summer (June 2013) in the northern parts of the Lena

524 basin, in for example the Viliui River and the northernmost tributaries of CSP and VER, showed

13 525 higher C/NDOM and lower δ CDOC (Figure 3b, c). This indicated a stronger influence of fresh

526 plant material, which could be due to surficial water flow paths caused by the shallow active

527 layer in the beginning of summer and sampling at higher latitudes in the 2013 samples (Figure 5a,

528 b). In contrast, samples collected at lower latitudes and later in the summer when temperatures

529 were higher (July 2012), and at lower latitudes in the tributaries of Aldan and Upper Lena,

13 530 showed opposite patterns (high δ CDOC and low C/NDOM) indicating more influence of SOM and

531 thus probably deeper dominant water flow paths due to deeper active layers (Figure 3b,c and 5).

532 The samples from the Upper Lena were collected late in the summer (July 2012), when air

533 temperatures were high and thus the active layers probably were thicker. This in turn allowed for

534 deeper water flow paths and an increased leakage of more degraded OC, probably primarily

535 derived from SOM to the rivers. The Upper Lena region has the mildest climate and more

536 widespread discontinuous permafrost distribution than in other parts of the Lena basin

537 [Chevychelov and Bosikov, 2010] that might impact water flow paths and OC fluxes [Frey and

538 McClelland, 2009]. Similarly, the samples collected from the Aldan outflow in early summer

25 13 539 (June 2013) had higher OC concentrations and C/NDOM and lower δ CDOC, indicating a larger

540 influence of fresh plant material as a source of DOC, whereas the influence of autochthonous

541 sources as a source of POC was lower than in July 2012. These results showed that the OC was

542 strongly dependent on sampling period and year.

543 The results presented here indicate that POC was derived from partly different sources than DOC

544 or was influenced by other processes either in the river water or in the soil during transport to the

13 545 river. The δ CPOC and C/NPOM varied between typical values of terrestrial SOM and values found

546 in freshwater autochthonous sources (Figure 6) [Farquhar et al., 1989; Finlay and Kendall,

547 2008; Osmond et al., 1981; Rundel et al., 1979; Smith and Epstein, 1971]. In general, the C/NPOM

548 was lower than C/Nplant, excluding fresh terrestrial vegetation as a main source of riverine POC.

549 The sources of POC were thus probably mixtures of SOM and a substantial fraction of

550 autochthonous sources (Figure 6). In some of the rivers in the LAIRA and CSP regions, where

551 results indicated autochthonous sources of POC, there were visible algae blooms during field

552 sampling.

553 The results of this study suggest that POC in the Lena River is influenced by aquatic primary

554 production. Previous studies in the Lena River found that aquatic primary production is low

555 [Sorokin and Sorokin, 1996] and that the POC mainly is derived from old and highly degraded

556 terrestrial plant material [Kuptsov and Lisitsin, 1996; Lobbes et al., 2000]. In contrast, recent

13 557 studies report δ CPOC and C/NPOM that suggests an influence of autochthonous sources, whereas

558 14C data indicates that the POC is old and thus of SOM origin [Karlsson, 2015; McClelland et al.,

559 2016]. These seemingly contradictory results could be due to recycling of respired old POC

560 within the aquatic food web.

26 13 561 δ CPOC was correlated to both mean catchment slope and water temperatures (Figure 4c and S3).

562 These results indicate that tributaries draining lowland areas, like the Viliui River and CSP

563 tributaries, in general had a larger influence from autochthonous sources and higher water

564 temperatures than tributaries originating in mountainous areas. High water temperatures induce

565 higher metabolic rates and thus aquatic primary production as well as heterotrophic activity

566 [Allen et al., 2005; Gillooly et al., 2001], which could partly explain the correlation between high

13 567 water temperatures and low δ CPOC. Also, the low gradient in the CSP region could have led to

568 longer watershed water residence times and in turn higher water temperatures. Boike et al. [2016]

569 show that surface land temperatures and lake area in this region have increased consistently

570 during the period 2002 to 2009. Larger lake coverage generally favors growth of aquatic

571 organisms, including primary producers. The large influence of SOM in both POC and DOC

572 isotope compositions and concentrations, as well as the lower water temperatures further

573 indicates deeper water flow paths in the rivers mainly draining mountainous areas.

574 4.4 Conclusions

575 This study has provided insights into river OC dynamics in the Lena River basin. Spatially

576 variable concentrations, export and major sources of OC were found within the basin. These were

577 probably driven by differences in water flow pathways, topography, climate and land cover. The

578 slope in elevation across each sub-catchment was a significant predictor of OC concentrations.

579 Water derived from mountainous areas had in general low concentrations of OC and cold water

580 temperatures, and the primary source of DOC and POC was dominated by SOM. In contrast,

581 lowland areas had high concentrations of OC, higher water temperatures and the DOC sources

582 were dominated by relatively fresh plant litter, while the primary POC sources were

27 583 autochthonous produced material. We also observed temporal differences in DOC sources with

584 tributaries sampled early in the summer having a larger influence of fresh plant material,

585 indicating that the dominant water flow paths were concentrated to surficial soil horizons. DOC

586 in tributaries sampled later in the summer had a larger influence of SOM, indicating deeper water

587 flow paths. In addition, the results showed that the spatial variations in export and area specific

588 export of OC were mainly due to variation in discharge, but with additional impacts by local

589 climate, variations in vegetation and differences in permafrost extent.

590 The results suggested that the Lena River basin could experience shifts in OC fluxes and

591 concentrations with continued permafrost degradation. A warming climate will likely lead to

592 changes in water flow paths that in turn would affect the export of OC from the Lena River basin.

593 Our results indicate that continued permafrost degradation leading to deeper water flow pathways

594 may induce lower concentrations of OC. Exports and fluxes of OC are, however, driven by

595 variations in discharge, and as this study indicates fluxes and concentration are spatially variable,

596 and so future shifts will ultimately be dependent on changes in hydrology and precipitation

597 patterns in the Lena River basin.

598 5 Acknowledgments

599 We thank the staff at the laboratories at department of Geosciences at the Museum of Natural

600 History, department of Environmental Science and Analytical Chemistry at Stockholm University

601 and the Stable Isotope Laboratory at Stockholm University for assistance. We also thank Peter

602 Raymond, Sebastian Sobek and Mattias Winterdahl for insightful comments on earlier version of

603 the manuscript. Finally, we want to thank all our Russian colleagues at the International Center

604 for BioGeoScience Education and Scientific Training (BEST) of the North-Eastern Federal

28 605 University in Yakutsk for all assistance in the field as well as the crew members at R/V

606 Geokryolog, R/V Akademik, and R/V Merelotoved. Biogeochemical data used in this study are

607 included as one table in an SI file. The discharge data used were measured by the Russian Hydro-

608 meterological Services and are freely available at R-ArcticNet (http://www.r-

609 arcticnet.sr.unh.edu/v4.0/index.html) and ArcticRIMS (http://rims.unh.edu/data.shtml). GIS data

610 were downloaded from (https://lta.cr.usgs.gov/GTOPO30). This project was supported by the

611 Swedish Research Council (VR 621-2010-3917) and the Swedish Polar Research Secretariat

612 (SIMO 2011-165 and 2012-213).

613 6 References

614 Allen, A. P., J. F. Gillooly, and J. H. Brown (2005), Linking the global carbon cycle to individual 615 metabolism, Functional Ecology, 19(2), 202-213, doi: 10.1111/j.1365-2435.2005.00952.x.

616 Amon, R. M. W., et al. (2012), Dissolved organic matter sources in large Arctic rivers, Geochimica et 617 Cosmochimica Acta, 94(0), 217-237, doi: 10.1016/j.gca.2012.07.015.

618 Boike, J., T. Grau, B. Heim, F. Günther, M. Langer, S. Muster, I. Gouttevin, and S. Lange (2016), 619 Satellite-derived changes in the permafrost landscape of central Yakutia, 2000–2011: Wetting, drying, and 620 fires, Global and Planetary Change, 139, 116-127, doi: 10.1016/j.gloplacha.2016.01.001.

621 Boström, B., D. Comstedt, and A. Ekblad (2007), Isotope fractionation and 13C enrichment in soil profiles 622 during the decomposition of soil organic matter, Oecologia, 153(1), 89-98, doi: 10.1007/s00442-007- 623 0700-8.

624 Cauwet, G., and I. Sidorov (1996), The biogeochemistry of Lena River: organic carbon and nutrients 625 distribution, Marine Chemistry, 53(3–4), 211-227, doi: 10.1016/0304-4203(95)00090-9.

626 Chevychelov, A. P., and N. P. Bosikov (2010), Natural Conditions, in The Far North, edited by E. I. 627 Troeva, A. P. Isaev, M. M. Cherosov and N. S. Karpov, pp. 1-23, Springer Netherlands.

628 Cleveland, C., and D. Liptzin (2007), C:N:P stoichiometry in soil: is there a “Redfield ratio” for the 629 microbial biomass?, Biogeochemistry, 85(3), 235-252, doi: 10.1007/s10533-007-9132-0.

630 Collins, M., R. Knutti, J. Arblaster, J.-L. Dufresne, T. Fichefet, P. Friedlingstein, X. Gao, W.J. Gutowski, 631 T. Johns, G., and M. S. Krinner, C. Tebaldi, A.J. Weaver, M. Wehner (2013), Long-term Climate Change: 632 Projections, Commitments and Irreversibility, in Climate Change 2013: The Physical Science Basis. 633 Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on 634 Climate Change, edited by T. F. Stocker, D. Qin, G.-K. and M. T. Plattner, S.K. Allen, J. Boschung, A.

29 635 Nauels, Y. Xia, V. Bex and P.M. Midgley, pp. 1029-1136, Cambridge University Press, United Kingdom 636 and New York, NY, USA.

637 Cooper, L. W., J. W. McClelland, R. M. Holmes, P. A. Raymond, J. J. Gibson, C. K. Guay, and B. J. 638 Peterson (2008), Flow-weighted values of runoff tracers (δ18O, DOC, Ba, alkalinity) from the six largest 639 Arctic rivers, Geophysical Research Letters, 35(18), L18606, doi: 10.1029/2008gl035007.

640 Costard, F., E. Gautier, D. Brunstein, J. Hammadi, A. Fedorov, D. Yang, and L. Dupeyrat (2007), Impact 641 of the global warming on the fluvial thermal erosion over the Lena River in Central Siberia, Geophysical 642 Research Letter, 34(14), L14501, doi: 10.1029/2007gl030212.

643 del Giorgio, P. A., and M. L. Pace (2008), Relative independence of organic carbon transport and 644 processing in a large temperate river: The Hudson River as both pipe and reactor, Limnology and 645 Oceanography, 53(1), 185-197, doi: 10.4319/lo.2008.53.1.0185.

646 Drake, T. W., K. P. Wickland, R. G. M. Spencer, D. M. McKnight, and R. G. Striegl (2015), Ancient low– 647 molecular-weight organic acids in permafrost fuel rapid carbon dioxide production upon thaw, 648 Proceedings of the National Academy of Sciences, 112(45), 13946-13951, doi: 10.1073/pnas.1511705112.

649 Dudarev, O. V., I. P. Semiletov, and A. N. Charkin (2006), Particulate material composition in the Lena 650 River-Laptev Sea system: Scales of heterogeneities, Doklady Earth Sciences, 411(2), 1445-1451, doi: 651 10.1134/s1028334x0609025x.

652 Eckhardt, B. W., and T. R. Moore (1990), Controls on Dissolved Organic Carbon Concentrations in 653 Streams, Southern Québec, Canadian Journal of Fisheries and Aquatic Sciences, 47(8), 1537-1544, doi: 654 10.1139/f90-173.

655 Ehleringer, J. R., N. Buchmann, and L. B. Flanagan (2000), Carbon isotope ratios in belowground carbon 656 cycle processes, Ecological Applications, 10(2), 412-422, doi: 10.1890/1051- 657 0761(2000)010[0412:ciribc]2.0.co;2.

658 Elser, J. J., et al. (2000), Nutritional constraints in terrestrial and freshwater food webs, Nature, 408(6812), 659 578-580.

660 Farquhar, G. D., J. R. Ehleringer, and K. T. Hubick (1989), Carbon isotope discrimination and 661 photosynthesis, Annual Review of Plant Physiology and Plant Molecular Biology, 40, 503-537, doi: 662 10.1146/annurev.arplant.40.1.503.

663 Fedorov, A. N., R. N. Ivanova, H. Park, T. Hiyama, and Y. Iijima (2014), Recent air temperature changes 664 in the permafrost landscapes of northeastern Eurasia, Polar Science, 8(2), 114-128, doi: 665 10.1016/j.polar.2014.02.001.

666 Finlay, J. C., and C. Kendall (2008), Stable Isotope Tracing of Temporal and Spatial Variability in 667 Organic Matter Sources to Freshwater Ecosystems, in Stable Isotopes in Ecology and Environmental 668 Science, edited, pp. 283-333, Blackwell Publishing Ltd.

669 Finlay, J. C., M. E. Power, and G. Cabana (1999), Effects of water velocity on algal carbon isotope ratios: 670 Implications for river food web studies, Limnolology and Oceanography, 44(5), 1198-1203.

671 Frey, K. E., and L. C. Smith (2005), Amplified carbon release from vast West Siberian peatlands by 2100, 672 Geophysical Research Letters, 32(9), L09401, doi: 10.1029/2004gl022025.

30 673 Frey, K. E., and J. W. McClelland (2009), Impacts of permafrost degradation on arctic river 674 biogeochemistry, Hydrological Processes, 23(1), 169-182, doi: 10.1002/hyp.7196.

675 Galy, V., B. Peucker-Ehrenbrink, and T. Eglinton (2015), Global carbon export from the terrestrial 676 biosphere controlled by erosion, Nature, 521(7551), 204-207, doi: 10.1038/nature14400 677 http://www.nature.com/nature/journal/v521/n7551/abs/nature14400.html#supplementary-information.

678 Gillooly, J. F., J. H. Brown, G. B. West, V. M. Savage, and E. L. Charnov (2001), Effects of Size and 679 Temperature on Metabolic Rate, Science, 293(5538), 2248-2251, doi: 10.1126/science.1061967.

680 Gittel, A., et al. (2014), Distinct microbial communities associated with buried soils in the Siberian tundra, 681 The ISME Journal, 8(4), 841-853, doi: 10.1038/ismej.2013.219.

682 Godsey, S. E., J. W. Kirchner, and D. W. Clow (2009), Concentration–discharge relationships reflect 683 chemostatic characteristics of US catchments, Hydrological Processes, 23(13), 1844-1864, doi: 684 10.1002/hyp.7315.

685 Gordeev, V., and M. Kravchishina (2009), River flux of dissolved organic carbon (DOC) and particulate 686 organic carbon (POC) to the Arctic Ocean: what are the consequences of the global changes?, in Influence 687 of Climate Change on the Changing Arctic and Sub-Arctic Conditions, edited by J. J. Nihoul and A. 688 Kostianoy, pp. 145-160, Springer Netherlands.

689 Guo, L., C.-L. Ping, and R. W. Macdonald (2007), Mobilization pathways of organic carbon from 690 permafrost to arctic rivers in a changing climate, Geophysical Research Letters, 34(13), L13603, doi: 691 10.1029/2007gl030689.

692 Holmes, R., et al. (2012), Seasonal and Annual Fluxes of Nutrients and Organic Matter from Large Rivers 693 to the Arctic Ocean and Surrounding Seas, Estuaries and Coasts, 35(2), 369-382, doi: 10.1007/s12237- 694 011-9386-6.

695 Hugelius, G., J. Routh, P. Kuhry, and P. Crill (2012), Mapping the degree of decomposition and thaw 696 remobilization potential of soil organic matter in discontinuous permafrost terrain, Journal of Geophysical 697 Research: Biogeosciences, 117(G2), G02030, doi: 10.1029/2011JG001873.

698 Hugelius, G., et al. (2014), Estimated stocks of circumpolar permafrost carbon with quantified uncertainty 699 ranges and identified data gaps, Biogeosciences, 11(23), 6573-6593, doi: 10.5194/bg-11-6573-2014.

700 Jansson, M., L. Persson, A. M. De Roos, R. I. Jones, and L. J. Tranvik (2007), Terrestrial carbon and 701 intraspecific size-variation shape lake ecosystems, Trends in Ecology & Evolution, 22(6), 316-322, doi: 702 10.1016/j.tree.2007.02.015.

703 Jasechko, S., J. W. Kirchner, J. M. Welker, and J. J. McDonnell (2016), Substantial proportion of global 704 streamflow less than three months old, Nature Geoscience, advance online publication, doi: 705 10.1038/ngeo2636 706 http://www.nature.com/ngeo/journal/vaop/ncurrent/abs/ngeo2636.html#supplementary-information.

707 Kaiser, C., H. Meyer, C. Biasi, O. Rusalimova, P. Barsukov, and A. Richter (2007), Conservation of soil 708 organic matter through cryoturbation in arctic soils in Siberia, Journal of Geophysical Research: 709 Biogeosciences, 112(G2), G02017, doi: 10.1029/2006JG000258.

31 710 Kaiser, K., and K. Kalbitz (2012), Cycling downwards – dissolved organic matter in soils, Soil Biology 711 and Biochemistry, 52(0), 29-32, doi: 10.1016/j.soilbio.2012.04.002.

712 Karlsson, E. S. (2015), Compositional clues to sources and sinks of terrestrial organic matter transported 713 to the Eurasian Arctic shelf, PhD thesis, Stockholm University, Sweden.

714 Keller, K., J. D. Blum, and G. W. Kling (2010), Stream geochemistry as an indicator of increasing 715 permafrost thaw depth in an arctic watershed, Chemical Geology, 273(1–2), 76-81, doi: 716 10.1016/j.chemgeo.2010.02.013.

717 Kerner, M., H. Hohenberg, S. Ertl, M. Reckermann, and A. Spitzy (2003), Self-organization of dissolved 718 organic matter to micelle-like microparticles in river water, Nature, 422(6928), 150-154, doi: 719 http://www.nature.com/nature/journal/v422/n6928/suppinfo/nature01469_S1.html.

720 Klaminder, J., H. Grip, C. M. Mörth, and H. Laudon (2011), Carbon mineralization and pyrite oxidation in 721 groundwater: Importance for silicate weathering in boreal forest soils and stream base-flow chemistry, 722 Applied Geochemistry, 26(3), 319-325, doi: 10.1016/j.apgeochem.2010.12.005.

723 Kuptsov, V. M., and A. P. Lisitsin (1996), Radiocarbon of quaternary along shore and bottom deposits of 724 the Lena and the Laptev Sea sediments, Marine Chemistry, 53(3–4), 301-311, doi: 10.1016/0304- 725 4203(95)00096-8.

726 Kuzmin, M., E. Tarasova, V. Bychinskii, E. Karabanov, A. Mamontov, and E. Mamontova (2009), 727 Hydrochemical regime components of Lena water, Water Resources, 36(4), 418-430, doi: 728 10.1134/s0097807809040058.

729 Kuznetsova, L. V., et al. (2010), Flora of Yakutia: Composition and Ecological Structure, in The Far 730 North, edited by E. I. Troeva, A. P. Isaev, M. M. Cherosov and N. S. Karpov, pp. 24-140, Springer 731 Netherlands.

732 Lam, B., A. Baer, M. Alaee, B. Lefebvre, A. Moser, A. Williams, and A. J. Simpson (2007), Major 733 structural components in freshwater dissolved organic matter, Environmental Science & Technology, 734 41(24), 8240-8247, doi: 10.1021/es0713072.

735 Lara, R. J., V. Rachold, G. Kattner, H. W. Hubberten, G. Guggenberger, A. Skoog, and D. N. Thomas 736 (1998), Dissolved organic matter and nutrients in the Lena River, Siberian Arctic: Characteristics and 737 distribution, Marine Chemistry, 59(3–4), 301-309, doi: 10.1016/S0304-4203(97)00076-5.

738 Laudon, H., M. Berggren, A. Ågren, I. Buffam, K. Bishop, T. Grabs, M. Jansson, and S. Köhler (2011), 739 Patterns and Dynamics of Dissolved Organic Carbon (DOC) in Boreal Streams: The Role of Processes, 740 Connectivity, and Scaling, Ecosystems, 14(6), 880-893, doi: 10.1007/s10021-011-9452-8.

741 Laudon, H., J. Buttle, S. K. Carey, J. McDonnell, K. McGuire, J. Seibert, J. Shanley, C. Soulsby, and D. 742 Tetzlaff (2012), Cross-regional prediction of long-term trajectory of stream water DOC response to 743 climate change, Geophysical Research Letters, 39(18), L18404, doi: 10.1029/2012gl053033.

744 Le Fouest, V., M. Babin, and J. É. Tremblay (2013), The fate of riverine nutrients on Arctic shelves, 745 Biogeosciences, 10(6), 3661-3677, doi: 10.5194/bg-10-3661-2013.

32 746 Lobbes, J. M., H. P. Fitznar, and G. Kattner (2000), Biogeochemical characteristics of dissolved and 747 particulate organic matter in Russian rivers entering the Arctic Ocean, Geochimica et Cosmochimica Acta, 748 64(17), 2973-2983, doi: 10.1016/S0016-7037(00)00409-9.

749 Mann, P. J., T. I. Eglinton, C. P. McIntyre, N. Zimov, A. Davydova, J. E. Vonk, R. M. Holmes, and R. G. 750 M. Spencer (2015), Utilization of ancient permafrost carbon in headwaters of Arctic fluvial networks, 751 Nature Communications, 6, 7856, doi: 10.1038/ncomms8856.

752 McClelland, J. W., et al. (2016), Particulate organic carbon and nitrogen export from major Arctic rivers, 753 Global Biogeochemical Cycles, doi: 10.1002/2015GB005351.

754 McGroddy, M. E., T. Daufresne, and L. O. Hedin (2004), Scaling of C:N:P stoichiometry in forests 755 worldwide: implications of terrestrial redfield-type ratios, Ecology, 85(9), 2390-2401, doi: 10.1890/03- 756 0351.

757 McGuire, K. J., J. J. McDonnell, M. Weiler, C. Kendall, B. L. McGlynn, J. M. Welker, and J. Seibert 758 (2005), The role of topography on catchment-scale water residence time, Water Resources Research, 759 41(5), W05002, doi: 10.1029/2004wr003657.

760 Menyailo, O. V., and B. A. Hungate (2006), Carbon and nitrogen stable isotopes in forest soils of Siberia, 761 Doklady Earth Science, 409(5), 747-749, doi: 10.1134/s1028334x06050151.

762 Meybeck, M. (1982), Carbon, nitrogen, and phosphorus transport by world rivers, American Journal of 763 Science, 282(4), 401-450, doi: 10.2475/ajs.282.4.401

764 Meybeck, M. (1993), C, N, P and S in Rivers: From Sources to Global Inputs, in Interactions of C, N, P 765 and S Biogeochemical Cycles and Global Change, edited by R. Wollast, F. Mackenzie and L. Chou, pp. 766 163-193, Springer Berlin Heidelberg.

767 Mulholland, P. J. (1997), Dissolved organic matter concentration and flux in streams, Journal of the North 768 American Benthological Society, 16(1), 131-141, doi: 10.2307/1468246.

769 Mulholland, P. J. (2003), Large-scale patterns in dissolved organic carbon concentration, flux and sources, 770 in Aquatic ecosystems interactivity of dissolved organic matter, edited by S. E. G. Findlay and R. L. 771 Sinsabaugh, pp. 139-159, Elsevier Science (USA).

772 Neff, J. C., and D. U. Hooper (2002), Vegetation and climate controls on potential CO2, DOC and DON 773 production in northern latitude soils, Global Change Biology, 8(9), 872-884, doi: 10.1046/j.1365- 774 2486.2002.00517.x.

775 O'Leary, M. H. (1988), Carbon Isotopes in Photosynthesis, BioScience, 38(5), 328-336, doi: 776 10.2307/1310735.

777 Osmond, C. B., N. Valaane, S. M. Haslam, P. Uotila, and Z. Roksandic (1981), Comparisons of δ13C 778 values in leaves of aquatic macrophytes from different habitats in Britain and Finland; some implications 779 for photosynthetic processes in aquatic plants, Oecologia, 50(1), 117-124, doi: 10.1007/bf00378804.

780 Palmtag, J., G. Hugelius, N. Lashchinskiy, M. P. Tamstorf, A. Richter, B. Elberling, and P. Kuhry (2015), 781 Storage, Landscape Distribution, and Burial History of Soil Organic Matter in Contrasting Areas of 782 Continuous Permafrost, Arctic, Antarctic, and Alpine Research, 47(1), 71-88, doi: 10.1657/aaar0014-027.

33 783 Peterson, B. J., R. M. Holmes, J. W. McClelland, C. J. Vörösmarty, R. B. Lammers, A. I. Shiklomanov, I. 784 A. Shiklomanov, and S. Rahmstorf (2002), Increasing River Discharge to the Arctic Ocean, Science, 785 298(5601), 2171-2173, doi: 10.1126/science.1077445.

786 Pipko, I. I., S. P. Pugach, O. V. Dudarev, A. N. Charkin, and I. P. Semiletov (2010), Carbonate parameters 787 of the Lena River: Characteristics and distribution, Geochemistry International, 48(11), 1131-1137, doi: 788 10.1134/s0016702910110078.

789 Rachold, V., A. Alabyan, H. W. Hubberten, V. N. Korotaev, and A. A. Zaitsev (1996), Sediment transport 790 to the Laptev Sea–hydrology and geochemistry of the Lena River, Polar Research, 15(2), 183-196, doi: 791 10.1111/j.1751-8369.1996.tb00468.x.

792 Raymond, P. A., J. W. McClelland, R. M. Holmes, A. V. Zhulidov, K. Mull, B. J. Peterson, R. G. Striegl, 793 G. R. Aiken, and T. Y. Gurtovaya (2007), Flux and age of dissolved organic carbon exported to the Arctic 794 Ocean: A carbon isotopic study of the five largest arctic rivers, Global Biogeochemical Cycles, 21(4), 795 GB4011, doi: 10.1029/2007gb002934.

796 Ridge, I. (2002), Plants, 345 pp., Oxford University Press, New York, United States.

797 Romanovsky, V. E., T. S. Sazonova, V. T. Balobaev, N. I. Shender, and D. O. Sergueev (2007), Past and 798 recent changes in air and permafrost temperatures in eastern Siberia, Global and Planetary Change, 56(3– 799 4), 399-413, doi: 10.1016/j.gloplacha.2006.07.022.

800 Romanovsky, V. E., et al. (2010), Thermal state of permafrost in Russia, Permafrost and Periglacial 801 Processes, 21(2), 136-155, doi: 10.1002/ppp.683.

802 Rundel, P., W. Stichler, R. Zander, and H. Ziegler (1979), Carbon and hydrogen isotope ratios of 803 bryophytes from arid and humid regions, Oecologia, 44(1), 91-94, doi: 10.1007/bf00346404.

804 Sardans, J., A. Rivas-Ubach, and J. Peñuelas (2012), The elemental stoichiometry of aquatic and terrestrial 805 ecosystems and its relationships with organismic lifestyle and ecosystem structure and function: a review 806 and perspectives, Biogeochemistry, 111(1-3), 1-39, doi: 10.1007/s10533-011-9640-9.

807 Semiletov, I. P., I. I. Pipko, N. E. Shakhova, O. V. Dudarev, S. P. Pugach, A. N. Charkin, C. P. McRoy, 808 D. Kosmach, and Ö. Gustafsson (2011), Carbon transport by the Lena River from its headwaters to the 809 Arctic Ocean, with emphasis on fluvial input of terrestrial particulate organic carbon vs. carbon transport 810 by coastal erosion, Biogeosciences, 8(9), 2407-2426, doi: 10.5194/bg-8-2407-2011.

811 Shiklomanov, A. I., and R. B. Lammers (2009), Record Russian river discharge in 2007 and the limits of 812 analysis, Environmental Research Letters, 4(4), 045015, doi:10.1088/1748-9326/4/4/045015.

813 Smith, B. N., and S. Epstein (1971), Two Categories of 13C/12C Ratios for Higher Plants, Plant 814 Physiology, 47(3), 380-384, doi: 10.1104/pp.47.3.380.

815 Smith, L. C., and T. M. Pavelsky (2008), Estimation of river discharge, propagation speed, and hydraulic 816 geometry from space: Lena River, Siberia, Water Resources Research, 44(3), W03427, doi: 817 10.1029/2007wr006133.

818 Sorokin, Y. I., and P. Y. Sorokin (1996), Plankton and Primary Production in the Lena River Estuary and 819 in the South-eastern Laptev Sea, Estuarine, Coastal and Shelf Science, 43(4), 399-418, doi: 820 10.1006/ecss.1996.0078.

34 821 Sterner, R. W., and J. J. Elser (2002), Ecological Stoichiometry: The biology of elements from molecules 822 to the biosphere, 439 pp., Princeton Univeristy Press, Princeton, New Jersey.

823 Stolbovoi, V., and I. McCallum (2002), CD-ROM "Land Resources of Russia" edited, International 824 Institute for Applied Systems Analysis and the Russian Academy of Science, Laxenburg, Austria.

825 Striegl, R. G., G. R. Aiken, M. M. Dornblaser, P. A. Raymond, and K. P. Wickland (2005), A decrease in 826 discharge-normalized DOC export by the Yukon River during summer through autumn, Geophysical 827 Research Letters, 32(21), L21413, doi: 10.1029/2005gl024413.

828 Tananaev, N. I. (2016), Hydrological and sedimentary controls over fluvial thermal erosion, the Lena 829 River, central Yakutia, Geomorphology, 253, 524-533, doi: 10.1016/j.geomorph.2015.11.009.

830 Tipping, E. (2002), Cation binding by humic substances, Cambridge University Press.

831 Watson, V., H. Kooi, and V. Bense (2013), Potential controls on cold-season river flow behavior in 832 subarctic river basins of Siberia, Journal of Hydrology, 489(0), 214-226, doi: 833 10.1016/j.jhydrol.2013.03.011.

834 Winterdahl, M., M. Erlandsson, M. N. Futter, G. A. Weyhenmeyer, and K. Bishop (2014), Intra-annual 835 variability of organic carbon concentrations in running waters: drivers along a climatic gradient, Global 836 Biogeochemical Cycles, 2013GB004770, doi: 10.1002/2013gb004770.

837 Winterdahl, M., M. Futter, S. Köhler, H. Laudon, J. Seibert, and K. Bishop (2011), Riparian soil 838 temperature modification of the relationship between flow and dissolved organic carbon concentration in a 839 boreal stream, Water Resourses Research, 47(8), W08532, doi: 10.1029/2010wr010235.

840 Vonk, J. E., B. E. van Dongen, and Ö. Gustafsson (2010), Selective preservation of old organic carbon 841 fluvially released from sub-Arctic soils, Geophysical Research Letters, 37(11), n/a-n/a, doi: 842 10.1029/2010GL042909.

843 Woo, M.-K., D. L. Kane, S. K. Carey, and D. Yang (2008), Progress in permafrost hydrology in the new 844 millennium, Permafrost and Periglacial Processes, 19(2), 237-254, doi: 10.1002/ppp.613.

845 Vuorio, K., M. Meili, and J. Sarvala (2006), Taxon-specific variation in the stable isotopic signatures 846 (δ13C and δ15N) of lake phytoplankton, Freshwater Biology, 51(5), 807-822, doi: 10.1111/j.1365- 847 2427.2006.01529.x.

848 Ye, B., D. Yang, and D. L. Kane (2003), Changes in Lena River streamflow hydrology: Human impacts 849 versus natural variations, Water Resources Research, 39(7), 1200, doi: 10.1029/2003wr001991.

850 Ye, B., D. Yang, Z. Zhang, and D. L. Kane (2009), Variation of hydrological regime with permafrost 851 coverage over Lena Basin in Siberia, Journal of Geophysical Research: Atmospheres, 114(D7), D07102, 852 doi: 10.1029/2008jd010537.

853 Zech, M., R. Zech, and B. Glaser (2007), A 240,000-year stable carbon loess-like palaeosol sequence in 854 and nitrogen isotope record from a the Tumara Valley, Northeast Siberia, Chemical Geology, 242(3-4), 855 307-318, doi: 10.1016/j.chemgeo.2007.04.002.

35 856 Zhang, T., et al. (2005), Spatial and temporal variability in active layer thickness over the Russian Arctic 857 drainage basin, Journal of Geophysical Research: Atmospheres, 110(D16), D16101, doi: 858 10.1029/2004jd005642. 859

36 860 7 Tables

861 Table 1. Carbon export and area specific export (ASE) over the period June and July from areas upstream of hydrologic 862 stations in the tributaries Aldan River and Viliui River as well as in the upstream and downstream Lena main channel. 863 Stations are listed from upstream to downstream.

Upper Lena Aldan Viliui Lena at Kiusiur Lower Lena

Lena at Aldan at Verkhoyanski’ Viliui at Remaining Name hydrologic station Lena at Kiusiur Tabaga Perevos Hatyrik-Homo c area Hydrologic station code a 3042/6147 3229/6236 3329/6266 3821/6342

Latitude (DD) 61.83 63.32 63.95 70.68

Longitude (DD) 129.6 132.02 124.83 127.39

Data period discharge 1936-2011 1942-1999 1936-1998 1936-2015

Drainage area (km2) 897000 696000 452000 2430000 385000 Drainage area (%) 37 29 19 100 15 Mean Q (m3 s-1) b 19355 14921 4950 56259 17033 Mean Q (%) b 34 27 9 100 30 Specific Q (l s-1 km-2) b 22.6 21.4 11.0 23.2 44.2 -1 d ETOC (Tg period ) 1.35±0.4 0.84±0.2 0.34±0.09 3.54±0.5 1.01±0.3 ETOC (%) 38.1 23.8 9.6 100 28.5 -1 d EDOC (Tg period ) 1.20 0.79 0.33 3.46 1.14 -1 d EPOC (Tg period ) 0.15 0.05 0.01 0.08 -0.13 -2 -1 e ASETOC (g m period ) 1.5 1.2 0.8 1.5 2.6 a Station ArcticRIMS code/R-ArcticNetID b Data based on discharge (Q) downloaded from the webpages of ArcticRIMS, R-ArcticNETand Arctic Great Rivers Observatory (NSF-1107774). c Re-analysed Q at this site by sources. d E=Q x C, where E represent the export of OC, Q is the water discharge and C is the concentrations OC. The data is based on mean monthly Q in June and July for above mentioned data periods and concentration for the months June 2012 and July 2013. The uncertainty in the export is estimated from a Monte Carlo simulation given as range, see methods section. e -1 ASETOC = ETOC A , where the ASETOC is the area specific export of TOC, i.e. the export per unit area (A).

864

865

37 866 Table 2. Distribution of stream water chemistry divided into groups and plant sample data

13 13 Quantiles TAir TH2O pH Cond DOC δ CDOM C/NDOM POC δ CPOM C/NPOM (°C) (°C) (μS/cmC) (μM) (‰) (atomic) (μM) (‰) (atomic) Water samples

All water samples Median 23 17.9 7.2 112 900 -27.2 28 38 -29.2 10 Q1 21 15.9 7 87 540 -27.5 18 22 -32.6 9 Q2 25 19.3 7.5 159 1100 -26.6 34 70 -27.1 11 Min 13 8.6 6.3 42 90 -28.4 4 6 -36.1 5 Max 31 23.8 9.4 346 3400 -25.7 41 640 -24.3 17

1. Lena main channel Median 24 16.5 7.2 101 970 -27.1 35 42 -27.1 11 Q1 22 15.8 6.9 85 840 -27.4 20 20 -30.9 9 Q2 26 18.2 7.5 140 1000 -26.6 38 77 -26.0 12 Min 13 12.0 6.5 69 590 -27.8 11 10 -34.0 7 Max 31 19.8 7.8 174 1300 -26.2 41 640 -25.3 15

2. Aldan main channel Median 20 18.4 7.3 92 730 -26.8 21 33 -27.6 10 Q1 19 17.9 7.1 86 700 -27.1 17 24 -29.0 7 Q2 24 19.1 7.5 99 790 -26.4 30 64 -26.3 14 Min 18 10.9 6.8 86 670 -27.2 17 19 -30.0 5 Max 25 19.2 7.7 105 1100 -26.3 40 94 -25.3 17

3. Viliui main channel Median 24 19.4 7.2 111 1000 -27.9 34 44 -33.4 10 Q1 23 18.6 7.2 110 1000 -28.2 34 33 -34.3 9 Q2 25 20.0 7.4 126 1100 -27.7 35 77 -32.0 11 Min 21 17.9 7.2 110 1000 -28.4 34 31 -34.6 8 Max 25 20.0 7.4 140 1200 -27.6 35 84 -31.1 12

4. Verkhoyansk Median 23 16.3 7.1 111 260 -27.1 19 17 -27.6 10 tributaries (VER) Q1 20 15.1 6.9 71 170 -27.5 15 9 -29.3 9 Q2 25 18.3 7.4 172 450 -26.5 29 29 -27.0 11 Min 16 11.2 6.3 49 90 -27.6 4 6 -32.9 8 Max 28 22.3 7.8 244 1900 -25.7 38 116 -24.3 13

5. Central Siberian Median 23 19.4 7.4 147 1200 -27.6 29 61 -32.5 10 Plateau tributaries Q1 22 18.2 7.1 57 930 -28.2 25 45 -34.6 7 (CSP) Q2 25 22.4 8.4 179 1800 -27.2 31 141 -30.8 11 Min 17 16.2 6.7 42 780 -28.4 19 40 -36.1 5 Max 30 23.8 9.4 346 2100 -26.6 39 215 -27.1 12

6. Lena-Amginsky Median 20 19.0 7.4 299 1100 -26.6 19 58 -33.1 8 inter-river area Q1 19 12.5 7.2 133 580 -26.8 16 39 -33.9 6 tributaries (LAIRA) Q2 23 21.6 8.2 310 1800 -26.4 28 100 -29.6 9 Min 18 8.6 6.5 106 570 -27.5 13 38 -34.0 6 Max 24 20.7 8.2 336 3400 -25.8 33 108 -28.5 9

Plant samples Plant material Median -29.6 36 Betula -25.4 42 Carex -29.6 26 Larch -29.9 36 Salix 1 -29.8 39 Salix 2 -27.0 33 867

38 868 8 Figures

869

39 870 Figure 1. a) Map of the Lena River basin and sample routes during 2012 (red) and 2013 (green). Number 1-6 indicate

871 separation of collected samples in main channels (1-3) and tributaries (4-6). b) Black dots show sample stations along the

872 main channels and red triangles show sample stations in tributaries. Colored areas indicate the catchment area of sampled

873 tributaries. c) Lena River discharge hydrograph at the hydrologic gauging station at Kiuisur during 2012 and 2013. Grey

874 areas show the sampling periods in 2012 and 2013.The black vertical line indicate the first sign of the flood close to the city

875 of Sinsk in the Upper Lena and the dashed line show highest discharge peak from the flood at Kiuisur 17 days later.

876

877

878 Figure 2. Export and area specific export (ASE) of total organic carbon (TOC) for the period June and July (p) from the

879 Upper Lena, the Aldan River, the Viliui River, and the Lower Lena based on carbon concentrations and discharge data

880 from the hydrologic gauging stations at Lena at Tabaga (A), Aldan at Verkhoyanski’ Perevos (B), Viliui at Hatyrik-Homo

881 (C) and Lena at Kiuisur (D). Black arrows indicate the fractional export from each sub-basin of the total export from the

882 Lena River outflow at station Kiuisur (D).

40 883

41 884

13 13 Figure 3. Maps showing a) concentrations of DOC and POC, b) δ CDOC and δ CPOC values and c) C/NDOM and C/NPOM atomic ratios at the outflow of sampled tributaries and the most upstream sample stations of the Lena River, 13 13 Aldan River and Viliui River. Boxplots show the distribution of a) DOC and POC b) δ CDOC and δ CPOC, and c) C/NDOM and C/NPOM divided into source areas where 1-3 are the main channels of the Lena River, Aldan River and Viliui River and 4-6 are groups of tributaries originating in Verkhoyansk (VER), Central Siberian Plateau (CSP) and the Lena-Amginsky inter-river area (LAIRA). Dashed lines in boxplots indicate median values of all samples.

42 885

13 886 Figure 4. Mean catchment slope against a) DOC concentrations and b) POC concentrations and c) δ CPOC in tributaries.

887 Colors indicate the different source areas.

43 888

889

13 890 Figure 5. a) Correlation between δ CDOC and C/NDOM, where red and black circles indicate the significant differences in

13 2 2 891 δ CDOC (p<0.0001, χ =49.1, n=77) and C/NDOM (p<0.0001, χ =39.7, n=77) between samples collected early in the summer

892 (June 2013) and later in the summer (July 2012). Color range indicates gradient in primary organic matter source with

893 relative importance of fresh plant material and SOM. b) Schematic showing the influence of shifting active layer on water

894 flow pathways.

44 895

896 Figure 6. Plot of C/N and δ13C in DOC and POC. The brown ellipse shows the range in C/N and δ13C of soil samples

897 collected in a 15 m soil profile in the catchment Tumara by Zech et al. [2007]. The ranges in C/N ratios and δ13C values in

898 autochthonous material and terrestrial plants are from Cleveland and Liptzin [2007], Finlay and Kendall [2008]; Finlay et

899 al. [1999]; O'Leary [1988], Sterner and Elser [2002] and Vuorio et al. [2006].

45

1

2

3

4

Tabel S 1 Data divided into six groups. The first three groups are the main channels of Lena, Aldan and Viliui Rivers and the last three groups are tributaries with headwaters in the Verkhoyansk, Central Siberian Plateau and Lena-Amginsky inter-river area. All samples are sorted from upstream to downstream direction.

13 13 Sample name Sample no River name Date Latitude Longitude T Air T H2O pH Cond DOC δ CDOC C/NDOM POC δ CPOC C/NPOM M D, Y (dec deg) (dec deg) (°C) (°C) (μS/cmC) (μM) (‰) (atomic) (μM) (‰) (atomic) Group 1. Lena/Лена main channel LR2012-32 32 Lena upstream outflow of July 27, 2012 60.6092 124.1885 13 15.7 7.9 174 1000 -26.2 27 640 -25.4 10 Tuolba/Tуолба (LR2012-33)

LR2012-31 31 Lena upstream Kytyl- July 26, 2012 60.8722 125.6335 21 16.9 7.2 146 920 -26.3 28 Diura/Кытыл-Дюра LR2012-34 34 Lena upstream Kytyl- July 27, 2012 60.8736 125.6438 22 15.8 7.4 159 990 -26.2 19 560 -25.6 9 Diura/Кытыл-Дюра LR2012-30 30 Lena at Sinsk/Синск. outflow of July 25, 2012 61.1079 126.8905 25 16.9 7.7 125 1000 -26.5 14 120 -26.8 12 Siniaia/Синяя (LR2012- 29.35.36) LR2012-28 28 Lena at July 24, 2012 61.2637 128.7397 23 16.2 7.5 80 1000 -26.7 20 120 -27.1 10 Bulgunniakhtakh/Булгунняхтах. upstream outflow of Buotama/Буоtама (LR2012-27)

LR2012-04 4 Lena downstream July 14, 2012 62.1577 129.9080 31 18.7 6.6 81 980 -27.1 20 46 -27.8 13 Yakutsk/Якутск LR2012-01 1 Lena downstream July 12, 2012 62.2626 130.0172 23 16.7 6.8 90 920 -27.2 28 55 -25.3 7 Tupagino/Тупагино LR2012-25 25 Lena upstream Khatas/Хатас July 22, 2012 62.6492 129.9074 25 19.6 6.9 90 840 -26.3 11

LR2012-03 3 Lena at Stolby/Столбы July 13, 2012 63.0194 129.6847 27 17.2 6.5 84 950 -26.7 14 81 -26.1 13 LR2013-78 78 Lena at Kharyialakh/Харыялах June 28, 2013 63.1327 129.6232 26 19.8 7.6 139 590 -27.4 36 17 -32.6 11

LR2012-02 2 Lena upstream outflow of July 12, 2012 63.3879 129.5393 25 16.7 6.6 69 970 -27.0 11 64 -27.0 15 Aldan/Алдан LR2013-75 75 Lena upstream outflow of June 27, 2013 63.5282 128.8600 24 16.3 7.6 92 670 -27.4 38 10 -30.0 9 Belianka/Белянка (LR2013-74)

LR2013-73 73 Lena between outflow of June 27, 2013 63.4904 128.8031 23 18.1 7.3 142 670 -27.4 37 19 -34.0 11 Belianka/Белянка (LR2013-74) and Kengkeme/Кенгкеме (LR2013-72)

LR2013-71 71 Lena upstream Sangar/Сангар June 27, 2013 63.8859 127.5421 16.1 7.4 85 720 -27.3 37 31 -28.7 9 13 13 Sample name Sample no River name Date Latitude Longitude T Air T H2O pH Cond DOC δ CDOC C/NDOM POC δ CPOC C/NPOM M D, Y (dec deg) (dec deg) (°C) (°C) (μS/cmC) (μM) (‰) (atomic) (μM) (‰) (atomic) LR2013-39 39 Lena upstream outflow of June 12, 2013 64.2183 126.8644 25 12.0 7.6 101 970 -27.2 39 38 -25.9 8 Viliui/Вилюи LR2013-41 41 Lena downstream outflow of June 13, 2013 64.3945 126.3659 28 18.9 7.0 140 1300 -27.8 33 66 -31.7 11 Viliui/Вилюи LR2013-57 57 Lena downstream outflow of June 21, 2013 65.0511 124.8055 22 14.9 7.0 83 900 -27.3 41 13 -25.8 10 Linde/Линде (LR2013-55) and Dianyshka/Дянышка (LR2013- 56

LR2013-54 54 Lena channel west side - No June 19, 2013 65.9486 123.9140 18 19.5 7.1 128 1200 -27.6 37 51 -31.6 11 name (close to North Polar circle) LR2013-42 42 Lena channel west side - Demon June 14, 2013 66.3439 123.6759 18 15.1 7.3 139 1200 -27.5 39 31 -31.2 11 stream LR2013-48 48 Lena upstream outflow of June 16, 2013 67.8737 123.0923 24 16.1 7.1 121 1100 -27.4 39 11 -29.2 11 Muna/Муна (LR2013-50) LR2013-49 49 Lena channel downstream June 16, 2013 67.9291 123.0021 28 16.1 7.0 130 1100 -27.0 38 24 -27.0 9 outflow of Muna/Муна (LR2013- 50) LR2013-45 45 Lena downstream June 15, 2013 68.7433 123.9966 21 14.2 7.1 91 840 -27.0 38 33 -26.2 14 Dzhardzhan/Джарджан Total mean 23 16.7 7.2 112 945 -27.0 29 100 -28.3 11 Total median 24 16.5 7.2 101 970 -27.1 35 42 -27.1 11 13 13 Sample name Sample no River name Date Latitude Longitude T Air T H2O pH Cond DOC δ CDOC C/NDOM POC δ CPOC C/NPOM M D, Y (dec deg) (dec deg) (°C) (°C) (μS/cmC) (μM) (‰) (atomic) (μM) (‰) (atomic) Group 2. Aldan/Алдан main channel LR2012-13 13 Aldan upstream of outflow July 19, 2012 62.6383 134.9219 18 19.1 7.7 86 700 -26.4 17 34 -28.1 10 Amga/Амга (LR2012-12) LR2012-11 11 Aldan upstream Megino- July 18, 2012 62.7106 134.6952 22 18.5 7.3 99 670 -26.3 17 Aldan/Мегино-Алдан. downsteam of outflow Tompo/Томпо (LR2012-10)

LR2012-09 9 Aldan upstream research site July 18, 2012 62.9229 134.1728 18.4 7.5 94 720 -26.8 17 Mammals Task LR2012-08 8 Aldan downstream outflow of July 17, 2012 63.2231 133.2462 25 17.9 7.1 92 730 -26.8 30 29 -27.4 17 Baraiy/Барайы (LR2012-07)

LR2012-05 5 Aldan downstream outflow of July 16, 2012 63.3495 131.6771 20 18.3 7.1 91 730 -27.1 21 33 -27.6 11 Baibakan/Байбакан (LR2012- 20) LR2012-22 22 Aldan at outflow July 21, 2012 63.4382 129.6667 19.2 6.8 105 790 -26.5 30 19 -30.0 9 LR2013-38 38 Aldan upstream outflow of June 12, 2013 63.4339 129.6398 19 10.9 7.4 86 1100 -27.2 40 94 -25.3 5 Aldan Total mean 21 17.5 7.3 93 780 -26.7 25 42 -27.7 10 Total median 20 18.4 7.3 92 730 -26.8 21 33 -27.6 10

Group 3. Viliui/Вилюи main channel LR2013-62 62 Viliui at Viliuisk/Bилюйск June 23, 2013 63.7582 121.5983 25 20.0 7.4 111 1000 -28.0 34 35 -34.0 8 LR2013-60 60 Viliui downstream Kysyl- June 23, 2013 63.9089 123.1457 24 19.4 7.2 110 1000 -27.8 34 31 -33.4 9 Syr/Кысыл-Cыр LR2013-59 59 Viliui downstream Khatyryk- June 22, 2013 63.8700 125.1667 24 20.0 7.2 110 1000 -27.9 34 44 -34.6 10 Khomo/Хатырык-Xомо

LR2013-66 66 Viliui 42 km from outflow June 24, 2013 64.0530 126.0691 24 19.4 7.4 112 1000 -28.4 35 71 -32.9 10 LR2013-40 40 Viliui by outflow June 13, 2013 64.3280 126.3720 21 17.9 7.2 140 1200 -27.6 35 84 -31.1 12 Total mean 24 19.3 7.3 117 1000 -28.0 35 53 -33.2 10 Total median 24 19.4 7.2 111 1000 -27.9 34 44 -33.4 10 13 13 Sample name Sample no River name Date Latitude Longitude T Air T H2O pH Cond DOC δ CDOC C/NDOM POC δ CPOC C/NPOM M D, Y (dec deg) (dec deg) (°C) (°C) (μS/cmC) (μM) (‰) (atomic) (μM) (‰) (atomic) Group 4. Verkhoyansk tributaries LR2012-10 10 Tompo/Томпо July 18, 2012 62.7085 134.7212 22 16.4 7.0 148 220 -26.1 14 LR2012-16 16 De pinne/Де пинне. July 20, 2012 63.1033 134.0384 23 14.6 6.3 103 330 -26.5 15 40 -26.2 10 Uiana/Уяна. Paygy/Паыгы LR2012-07 7 Baraiy/Барайы/Baraiy July 17, 2012 63.2044 133.2340 25 15.6 7.0 168 130 -26.5 13 7 -27.0 11 LR2012-18 18 No name - upstream Urasa July 20, 2012 63.3677 133.2736 24 19.9 6.6 118 1900 -27.3 28 29 -32.9 8 iuriage/Ураса Юряге (LR2012- 19) LR2012-19 19 Urasa/iuriage/Ураса Юряге July 20, 2012 63.3859 133.1369 26 14.8 6.6 50 280 -26.6 17 9 -27.5 10 LR2012-06 6 Tukulan/Тукулан July 16, 2012 63.3209 131.9309 20 16.4 7.4 206 110 -26.8 18 9 -28.3 10 LR2012-20 20 Baibakan/Байбакан July 20, 2012 63.3558 131.7532 18.0 6.8 60 230 -26.5 4 LR2012-21 21 Kele/Kеле July 21, 2012 63.3440 130.3791 22 18.1 7.0 244 170 -25.7 6 12 -26.0 8 LR2012-23 23 Tumara/Тумара July 21, 2012 63.4615 129.5693 23 22.3 7.1 170 410 -26.3 19 LR2013-77 77 Tumara/Тумара June 28, 2013 63.4681 129.5942 23 15.3 7.6 206 170 -27.5 18 17 -27.1 11 LR2012-24 24 Batamai/Батамай July 21, 2012 63.5202 129.3998 18.7 6.8 50 230 -26.4 13 22 -28.7 9 LR2013-76 76 Batamai/Батамай June 28, 2013 63.5221 129.3963 18 14.8 7.5 49 240 -27.4 29 6 -29.3 11 LR2013-74 74 Belianka/Белянка June 27, 2013 63.5203 128.8318 24 17.0 7.8 93 160 -27.3 17 8 -27.6 8 LR2013-70 70 Chochuma/Чочума June 26, 2013 64.0191 127.3502 19.3 7.4 88 500 -27.5 32 16 -30.2 10 LR2013-67 67 Liunkiubei/Люнкюбей June 26, 2013 64.1635 126.9666 25 16.2 7.2 73 590 -27.6 35 17 -27.5 10 LR2013-58 58 Liapiske/Ляписке June 22, 2013 64.6002 125.7179 16 15.5 7.1 89 850 -27.6 38 10 -27.6 10 LR2013-56 56 Dianyshka/Дянышка June 21, 2013 65.0055 124.9442 23 14.1 7.3 178 400 -26.9 29 24 -31.5 10 LR2013-53 53 Undiuliung/Ундюлюнг June 18, 2013 66.2329 124.1614 27 18.1 7.2 159 230 -26.8 19 23 -27.7 10 LR2013-51 51 Sobolokh Maian/Соболох Маян June 17, 2013 67.2521 123.4084 20 15.8 7.4 140 360 -27.6 27 29 -26.6 13

LR2013-47 47 Menkere/Менкере June 16, 2013 68.0216 123.4151 28 11.2 7.1 104 440 -27.3 28 120 -24.3 12 LR2013-46 46 Natara/Наtара June 15, 2013 68.3886 123.9738 19 15.7 7.1 64 470 -27.6 30 33 -29.9 10 LR2013-44 44 Dzhardzhan/Джарджан June 15, 2013 68.7325 124.0597 20 18.8 7.2 182 90 -27.5 23 33 -27.2 9 Total mean 23 16.7 7.1 125 390 -27.0 21 24 -28.1 10 Total median 23 16.3 7.1 111 260 -27.1 19 17 -27.6 10 13 13 Sample name Sample no River name Date Latitude Longitude T Air T H2O pH Cond DOC δ CDOC C/NDOM POC δ CPOC C/NPOM M D, Y (dec deg) (dec deg) (°C) (°C) (μS/cmC) (μM) (‰) (atomic) (μM) (‰) (atomic) Group 5. Central Siberian Plateau tributaries LR2012-29 29 Siniaia/Синяя July 25, 2012 61.1461 126.8620 25 23.4 8.6 163 2100 -26.9 24 150 -31.0 5 LR2012-36 36 Siniaia/Синяя July 28, 2012 61.1680 126.8677 23 23.8 9.4 171 2100 -26.6 19 140 -31.1 5 LR2012-35 35 Siniaia/Синяя July 28, 2012 61.1651 126.9110 17 22.3 9.0 170 2100 -26.6 22 140 -30.0 6 LR2013-72 72 Kengkeme/Кенгкеме June 27, 2013 63.4678 128.7899 23 19.1 7.4 201 790 -27.9 33 42 -34.8 11 LR2013-69 69 Berge-Tiugene/Берге-Тюгене June 26, 2013 63.9742 127.0290 22 18.8 7.5 346 870 -28.3 29 50 -35.5 8

LR2013-68 68 Lungkha/Лунгха June 26, 2013 64.1086 126.7406 22 19.6 7.5 288 950 -28.4 25 75 -36.1 9 LR2013-61 61 Tiung/Tюнг June 23, 2013 63.7804 121.5229 25 22.9 7.6 118 950 -27.8 29 52 -34.2 10 LR2013-63 63 Tangnary/Taнгнары June 24, 2013 64.0249 123.8852 24 19.2 7.0 50 1200 -27.5 31 70 -32.8 11 LR2013-64 64 Bappagai/Баппагай June 24, 2013 64.0276 124.0923 21 17.7 8.4 159 1300 -28.2 29 46 -27.1 8 LR2013-65 65 Uoranga/Уоранга June 24, 2013 64.0268 124.3989 25 18.4 7.2 70 1700 -27.4 26 220 -32.1 12 LR2013-55 55 Linde/Линде June 21, 2013 64.9520 124.5963 22 20.6 6.7 58 1300 -28.2 31 70 -34.5 11 LR2013-43 43 Outflow at Zhigansk/Жиганск June 14, 2013 66.7711 123.3601 26 16.2 6.9 52 780 -27.3 29 40 -30.2 10

LR2013-52 52 Khoruogka/Хоруонгка June 17, 2013 67.2141 123.1354 20 21.2 7.3 42 970 -27.6 29 47 -33.0 10 LR2013-50 50 Muna/Муна June 16, 2013 67.8763 123.0364 30 16.9 7.1 134 1100 -27.4 39 41 -31.8 11 Total mean 23 20.0 7.7 144 1300 -27.6 28 85 -32.4 9 Total median 23 19.4 7.4 147 1200 -27.6 29 61 -32.5 10

Group 6. Lena-Amginsky inter-river area LR2012-33 33 Tuolba/Tуолба July 27, 2012 60.5937 124.2726 19 17.7 8.2 305 1200 -26.6 19 LR2012-37 37 Oddokun/Оддокун July 29, 2012 61.1946 128.2840 21 12.5 7.3 336 3400 -25.8 28 73 -28.5 7 LR2012-27 27 Buotama/Буоtама July 24, 2012 61.2510 128.7695 24 20.6 8.0 310 570 -26.6 19 43 -34.0 9 LR2012-26 26 Tamma/Tамма July 24, 2012 61.9039 129.8472 19.0 7.2 106 920 -27.5 17 38 -32.9 8 LR2012-12 12 Amga/Амга July 19, 2012 62.6147 134.9229 18 20.3 8.2 299 580 -26.7 13 LR2012-15 15 Mammals task July 19, 2012 62.9465 134.0084 19 8.6 6.5 145 1800 -26.4 33 LR2012-17 17 Tatta/Татта July 20, 2012 63.0203 133.4082 23 20.7 7.4 133 1100 -26.8 16 110 -33.3 6 Total mean 21 17.0 7.5 233 1400 -26.6 21 66 -32.2 8 Total median 20 19.0 7.4 299 1100 -26.6 19 58 -33.1 8 13 13 Sample name Sample no River name Date Latitude Longitude T Air T H2O pH Cond DOC δ CDOC C/NDOM POC δ CPOC C/NPOM M D, Y (dec deg) (dec deg) (°C) (°C) (μS/cmC) (μM) (‰) (atomic) (μM) (‰) (atomic) Plant samples Betula P1 July 12, 2012 63.3879 129.5393 -25.4 42 Carex P2 July 14, 2012 62.1577 129.9080 -29.6 26 Larch P3 July 15, 2012 62.68 129.8 -29.9 36 Salix (1) P4 July 14, 2012 62.1577 129.9080 -29.8 39 Salix (2) P5 July 14, 2012 62.1577 129.9080 -27.0 33 Total mean -28.3 35 Total median -29.6 36 a) All river names are translitterated according to ICAO Doc 9303 Machine readable travel documents, 2015