Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Compositional clues to sources and sinks of terrestrial organic matter transported to the Eurasian Arctic shelf Emma Sofia Karlsson Doctoral thesis Environmental Science and Analytical Chemistry (ACES) Stockholm University Stockholm, 2015 1 © Emma S. Karlsson, Stockholm University 2015 Cover photo by Emma Karlsson after photos by Alexander Charkin and Oleg Dudarev ISBN: 978-91-7649-195-9 pp.1-40 Printed by Holmbergs, Malmö, Sweden 2015 Distributed by the Department of Environmental Science and Analytical Chemistry (ACES) 2 Muostakh Island coastal ice complex 2008, photo by Alexander Charkin 3 ABSTRACT The amount of organic carbon (OC) present in Siberian Arctic permafrost soils is estimated at twice the amount of carbon currently in the atmosphere. The shelf seas of the Arctic Ocean receive large amounts of this terrestrial OC from Eurasian Arctic rivers and from coastal erosion. Degradation of this landderived material in the sea would result in the production of dissolved carbon dioxide and may then add to the atmospheric carbon dioxide reservoir. Observations from the Siberian Arctic suggest that transfer of carbon from land to the marine environment is accelerating. However, it is not clear how much of the transported OC is degraded and oxidized, nor how much is removed from the active carbon cycle by burial in marine sediment. Using bulk geochemical parameters, total OC, 13C and 14C isotope composition, and specific molecular markers of plant wax lipids and lignin phenols, the abundance and composition of OC was determined in both dissolved and particulate carrier phases: the colloidal OC (COC; part of the dissolved OC), particulate OC (POC), and sedimentary OC (SOC). Statistical modelling was used to quantify the relative contribution of OC sources to these phases. Terrestrial OC is derived from the seasonally thawing top layer of permafrost soil (topsoil OC) and frozen OC derived from beneath the active layer eroded at the coast, commonly identified as yedoma ice complex deposit OC (yedoma ICD-OC). These carbon pools are transported differently in the aquatic conduits. Topsoil OC was found in young DOC and POC, in the river water, and the shelf water column, suggesting long-distance transport of this fraction. The yedoma ICDOC was found as old particulate OC that settles out rapidly to the underlying sediment and is laterally transported across the shelf, likely dispersed by bottom nepheloid layer transport or via ice rafting. These two modes of OC transport resulted in different degradation states of topsoil OC and yedoma ICDOC. Terrestrial CuO oxidation derived biomarkers indicated a highly degraded component in the COC. In contrast, the terrestrial component of the SOC was much less degraded. In line with earlier suggestions the mineral component in yedoma ICD functions as weight and surface protection of the associated OC, which led to burial in the sediment, and limited OC degradation. The degradability of the terrestrial OC in shelf sediment was also addressed in direct incubation studies. Molecular markers indicate marine OC (from primary production) was more readily degraded than terrestrial OC. Degradation was also faster in sediment from the East Siberian Sea, where the marine contribution was higher compared to the Laptev Sea. Although terrestrial carbon in the sediment was degraded slower, the terrestrial component also contributed to carbon dioxide formation in the incubations of marine sediment. These results contribute to our understanding of the marine fate of land-derived OC from the Siberian Arctic. The mobilization of topsoil OC is expected to grow in magnitude with climate warming and associated active layer deepening. This translocated topsoil OC component was found to be highly degraded, which suggests degradation during transport and a possible contribution to atmospheric carbon dioxide. Similarly, the yedoma ICD-OC (and or old mineral soil carbon) may become a stronger source with accelerated warming, but slow degradation may limit its impact on active carbon cycling in the Siberian Shelf Seas. 4 SVENSK SAMMANFATTNING Organiskt kol lagrat i permafrostjordar i Sibiriska Arktis uppskattas till två gånger den mängd kol som just nu finns i gasform i atmosfären. En stor del av detta kol transporteras från land till de intilliggande kusthaven via stora floder, och genom erosion av flodbankar och kuster. Nedbrytning av detta terrestra material skulle leda till bildning av koldioxid i havet, som i sin tur kan bidra till ökad mängd koldioxid i atmosfären. Enligt studier ökar dessutom mängden landmaterial som transporteras till kusthaven i Sibirien. Det är dock oklart hur mycket av materialet som bryts ned och oxideras till koldioxid, och hur mycket som istället lagras och begravs i marina sediment. Med hjälp av geokemiska parametrar - organisk kolhalt, kolisotopsammansättning 13C and 14C, och specifika biomarkörer i bladvaxer och ligninfenoler - undersöktes sammansättningen av organiskt kol i både lösta och partikulära bärarfaser: i kolloider (en del av det lösta organiska kolet), i partiklar och i sediment. Statistisk modellering användes för att kvantifiera de relativa bidragen av organiskt kol i dessa tre faser. Organiskt material från land kommer i huvudsak från det översta permafrostlagret som tinar varje sommarsäsong (ytjord) och ifrån fryst material med stor isvolym under det aktiva lagret som eroderas vid flodbankar och kuster och kallas yedoma. De här två varianterna av landbaserat kol transporteras olika i den akvatiska miljön. Ytjord återfanns i löst och partikulärt kol både i Lenafloden och i ytvattnet över kusthavet, vilket tyder på att ytjordsmaterial transporteras långa sträckor i ytvatten. Yedoma återfanns i den partikulära fasen i vattnet, men endast nära land, och dominerade istället ytsediment, vilket tyder på att yedoma också transporteras långra sträckor lateralt, men troligen via bentisk transport eller genom infrysning och förflyttning i havsis. De olika transportvägarna verkar resultera i olika stadier av nedbrytning. Ligninfenoler indikerade en högst nedbruten komponent i kolloidfraktionen, som dominerades av ytjordsmaterial. I kontrast visade samma markörer en mycket mindre nedbruten komponent i sediment, som domineras av yedoma. I linje med tidigare förslag verkar mineraler i yedoma fungera som vikt och skydd mot nedbrytning till det associerade kolmaterialet. Yedoma sjunker ut ur vattenkolonnen och begravs, och undergår sedan begränsad nedbrytning i sedimentet. Nedbrytbarhet av yedoma undersöktes också genom inkubationer av sediment. Molekylära markörer visade att materialet från marin produktion bröts ned i högre grad än terrestert material. Nedbrytningen var också snabbare i sediment från Östsibiriska havet där den marina fraktionen var högre än i sediment från Laptevhavet. Även om den terrestra komponenten bröts ned saktare bidrog terrestert material till bildandet av koldioxid under nedbrytningen i inkubationerna. De här resultaten bidrar till vår förståelse av vad som händer med landmaterial som transporterats till havet från det Sibiriska Arktis. Mobiliseringen av ytjord förväntas öka i omfattning i samband med klimatuppvärmning och fördjupad upptining av permafrost. Ytjordsmaterial var nedbrutet, vilket tyder på att det bryts ned under (och/eller innan) transport. Ökad frisättning av detta material kan leda till ökad halt koldioxid i atmosfären. På liknande sätt kan större mängder yedoma frisättas till havet med ett varmare klimat, men begränsad nedbrytning av yedoma i sediment kan begränsa dess eventuella omvandling till koldioxid i de Sibiriska kusthaven. 5 ABBREVIATIONS and DEFINITIONS carrier phases COC cryosphere CPI δ13C Δ14C DOC ESAS freeze-locked GC-MS HMW ICD LMW OC PF Pleistocene POC SOC topsoil OC thaw-remobilized TN yedoma The physical forms that facilitate the transport of OC (COC, and POC) Colloidal organic carbon part of dissolved organic carbon pool that is between 0.7 µm and 1 000 Dalton in size The part of Earth in which water is frozen, e.g. permafrost, glaciers Carbon preference index, tracer of terrestrial origin and degradation Conventionally defined stable carbon isotope ratio - 13C to 12C Conventionally defined radiocarbon isotope ratio - 14C to 12C Dissolved organic carbon East Siberian Arctic Shelf Inaccessible to cycling due to being frozen, generally refers to material in long-term permafrost deposits Gas chromatography – mass spectrometer High molecular weight Ice complex deposit, see yedoma Low molecular weight Organic carbon Permafrost Geological epoch, between 2.6 million to 11 700 years ago Particulate organic carbon Sedimentary organic carbon, a particulate phase that has settled out The relatively young surface layer of permafrost soils that is seasonally thawed, part of the active layer Exposed to active cycling, and has previously been freeze-locked Total nitrogen Organic-rich mineral-associated permafrost soil with high content of ice, often in form of ice wedges, also called ice complex deposit, of late Pleistocene to mid Holocene ages 6 CONTENTS ABSTRACT ..................................................................................................................................................... 4 SVENSK SAMMANFATTNING ....................................................................................................................... 5 ABBREVIATIONS and DEFINITIONS .............................................................................................................. 6 My CONTRIBUTION to the papers ............................................................................................................... 9 OBJECTIVES ................................................................................................................................................. 10 1. INTRODUCTION ...................................................................................................................................... 12 Cryosphere and freeze-locked carbon ................................................................................................... 12 Terrestrial matter in the aquatic environment ..................................................................................... 14 2. METHODS ............................................................................................................................................... 14 Study area ............................................................................................................................................... 14 Sample collection ................................................................................................................................... 15 Analytical work ....................................................................................................................................... 16 Bulk geochemical composition tools ................................................................................................... 16 Monte Carlo simulation in end-member mixing model analysis......................................................... 17 Molecular geochemical composition tools ......................................................................................... 18 Closed system incubations of sediment in microcosm ........................................................................ 19 3. RESULTS and DISCUSSION ...................................................................................................................... 22 Source apportioning of terrestrial OC ................................................................................................... 22 Transport and transportability .............................................................................................................. 25 Degradation state and degradability ..................................................................................................... 27 4. CONCLUSIONS ........................................................................................................................................ 30 5. PERSPECTIVES on FURTHER RESEARCH ................................................................................................. 31 Preservation of yedoma ICD-OC ............................................................................................................ 31 Role of oxygen ..................................................................................................................................... 31 Sorption of OC ..................................................................................................................................... 31 Heterogeneity and Hydrology ................................................................................................................ 32 Spatial and temporal coverage .............................................................................................................. 32 ACKNOWLEDGEMENTS .............................................................................................................................. 33 REFERENCES ................................................................................................................................................ 36 7 List of PAPERS Paper #1 Carbon isotopes and lipid biomarker investigation of sources, transport and degradation of terrestrial organic matter in the Buor-Khaya Bay, SE Laptev Sea. Biogeosciences, 8, 1865–1879, 2011, doi:10.5194/bg-8-1865-2011. © Author(s) 2011. CC Attribution 3.0 License. Published by Copernicus Publications on behalf of the European Geosciences Union. Karlsson, E.S., A. Charkin, O. Dudarev, I. Semiletov, J. E. Vonk, L. Sánchez-García, A. Andersson, and Ö. Gustafsson Paper #2 Contrasting regimes for organic matter degradation in the East Siberian Sea and the Laptev Sea assessed through microbial incubations and molecular markers. Published in Marine Chemistry, 170, 11-15, 2015, doi:10.1016/j.marchem.2014.12.005. 0304-4203/© 2014 Elsevier B.V. All rights reserved. Karlsson, E.S., V. Brüchert, T. Tesi, A. Charkin, O. Dudarev, I. Semiletov, and Ö. Gustafsson Paper #3 Different sources and degradation state of dissolved, particulate and sedimentary organic matter along the Eurasian Arctic coastal margin Karlsson, E.S., J. Gelting, T. Tesi, B. van Dongen, I. Semiletov, A. Charkin, O. Dudarev, Ö. Gustafsson (submitted manuscript) Paper #4 Contrasting sources of dissolved and particulate organic matter along 62N-72N in the Siberian Arctic Lena River Karlsson, E.S., B. van Dongen, I. Semiletov, A. Charkin, O. Dudarev, Ö. Gustafsson (manuscript) 8 My CONTRIBUTION to the papers I, Emma S. Karlsson, contributed to the papers included in this thesis through the following: Paper #1 Performed the laboratory work after the 2008 sampling campaign was planned and carried out by colleagues. Interpreted the data and wrote the paper in close collaboration with the co-authors. Paper #2 Had a major role in planning the experimental setup of incubations. Carried out the laboratory work. Interpretations were done in collaboration with the co-authors. Had the main role in writing the paper. Paper #3 Performed a substantial part of the laboratory work, but the 2008 sampling campaign was planned and carried out by colleagues. Interpretations of the data were done in collaboration with the coauthors. Took the leading role in writing the paper. Paper #4 The 2008 sampling campaign was planned and carried out by colleagues. Prepared the samples for analysis. Data interpretations were done in collaboration with the co-authors. Took the lead on writing the paper. 9 OBJECTIVES The overall aim of this thesis was to improve our understanding of the sources and sinks of terrestrial OC released from the Eurasian Arctic landscape to the adjacent shelf seas. Specifically the objectives of the thesis were to determine: I) The sources of OC: What is the bulk and molecular geochemical composition of the major carbon carrier phases, colloidal, particulate and sedimentary OC? This was investigated with plant wax lipid and CuO oxidation reaction product biomarkers. What are the relative contributions of the major terrestrial OC sources, topsoil OC and yedoma ICD-OC to these carriers in the recipients, the Lena River and Eurasian Arctic shelf? This was quantified through a dual carbon isotope, δ13C and Δ14C, end-member mixing model analysis. II) Transport modes: How far is topsoil OC and yedoma ICD-OC transported and is it dispersed over the shelf? This was investigated by comparing the geochemical source proxies in the different carrier phases of OC in the water column and in the underlying sediments in a spatial context. III) Degradability: What is the degradation state and degradability of topsoil OC and yedoma ICD-OC in the colloidal, particulate and sedimentary OC in different shelf sea areas and regimes? This was investigated through molecular degradation state proxies and relative degradation rates derived from closed system microcosm incubations. 10 Figure 1. Schematic representation of carbon cycling processes in the Arctic shelf seas. 11 1. INTRODUCTION Cryosphere and freeze-locked carbon Thaw release of long-frozen cryosphere carbon is one of the few processes potentially moving large amounts of carbon from land to the atmosphere, on decadal to centennial time scales, as a response to raised greenhouse gases (such as carbon dioxide and methane) concentrations. Such releases may represent the largest future transfer of OC from the biosphere to the atmosphere (Koven et al., 2011). More than a fifth of the world´s terrestrial area is underlain by permafrost. Most of this is situated in the high Arctic and northern Siberia (Brown et al., 1997, Heginbottom et al., 1984, 2002; Figure 2). The estimated amounts of organic carbon (OC) stored in these deposits are vast (~1100-1500 Pg; Tarnocai et al., 2009; Hugelius et al., 2014). The dominant portion of this carbon is stored in perennially frozen ground (~800 Pg), but a large share (~500 Pg) is repeatedly thawed and refrozen as part of the active layer dynamics during seasonal cycling (topsoil). At the same time, these thermally vulnerable regions are also experiencing global warming at an amplified strength (ACIA, 2004; Jefferies et al., 2014; Tingley and Huybers, 2013; IPCC, 2013) causing dramatic changes in the cryosphere landscape. Riverbanks and shorelines are subject to erosional processes which can be dramatically accelerated through parallel climate warming induced effects. A longer ice free season, increased wave fetch, more frequent storms and raised sea water temperatures (Barnhardt et al., 2014, Stroeve et al., 2014; IPCC, 2013) can act in concert to enhance erosion. Some coastlines – e.g. those dominated by ice rich permafrost – are particularly vulnerable to such processes (Lantuit et al., 2011; 2012). These coastlines exhibit very high erosional rates, five to seven times faster than ordinary permafrost coastlines. This translates to annual shoreline retreats between 0.5 and 10 m (Grigoriev et al., 1993, 2004; Rachold et al., 2004; Lantuit et al., 2011, 2012; Günther et al., 2013) and thus release of late Pleistocene carbon to the marine environment (e.g., Vonk et al., 2012; Sanchez-Garcia et al., 2014). The seasonal dynamics of the active layer in permafrost soils, the top part of the upper meter of a permafrost landscape, which thaws each summer, can release OC in episodic or abrupt events, for instance, through the formation of thermokarst lakes and ground subsidence. Thermokarst lakes are thaw-depression lakes that form from meltwater when underlying ice disappears. In areas of sporadic or isolated permafrost, deepening of the active layer also leads to lateral shrinkage. Areas of discontinuous permafrost are therefore more vulnerable to warming and thus anticipated to disappear first (Osterkamp et al., 1999). In a frozen soil environment, microbial communities are without water, and thawing therefore releases one of the physical constraints to microbial (or photochemical) degradation (Rothman et al., 2007). Previously freeze-locked permafrost OC is thus potentially exposed to decomposition and a possible subsequent conversion to greenhouse gases. The estimated feedback of permafrost OC to the atmosphere ranges between 7 and 500 Pg by the year 2100 (Schaefer et al., 2011; MacDougall et al., 2012; Schneider von Deimling et al., 2012). There is often a decoupling between permafrost thaw location and site of emission (Vonk and Gustafsson, 2013), and processing of old thaw-remobilized OC can take place at many niches in the 12 landscape, and during transport (Fig. 1). Lateral transport of thaw-remobilized old permafrost OC is highly dependent on the hydrological cycle, since water is the major vector for its transport. Although transport can be wind-driven, aeolian deposition of terrestrial OC is comparatively small over the Eurasian Arctic shelf (Shevchenko and Lizitsin, in Stein and MacDonald, 2004). The hydrological cycle has also been predicted and observed to undergo marked changes with climate warming, such as increased precipitation and runoff (Dickson et al., 2000; Peterson et al., 2002; Serreze et al., 2003). Peterson et al. (2002) for example report a 7% increase in average annual river discharge between 1936 and 1999 from six large Eurasian Arctic rivers. In a study of daily flow data from 111 northern rivers, Smith et al. (2007) suggest an increase of groundwater supply as part of the reason for observed increased river discharge rates reaching at least back to ~1985. They propose this to be a combined effect of the increase in declined seasonal ground freezing and precipitation. Permafrost thaw and remobilization processes are difficult to monitor off site, and remote sensing techniques often have poor coverage at high latitudes, too low resolution, or are simply insensitive to permafrost processes. Due to the high heterogeneity, local studies are also hard to upscale for describing processes on a pan-Arctic scale. Therefore many questions remain about the northern Siberian Arctic systems which, due to the harsh environments and the extensive areas, are inaccessible and understudied. Continuous PF (90-100%) Discontinuous PF (50-90%) Sporadic PF (10-50%) Isolated PF (<10%) No permafrost Yedoma Figure 2. Circum-Arctic permafrost coverage. Data from Brown et al, 1998. 13 Terrestrial matter in the aquatic environment One way to handle the large heterogeneities of the Eurasian Arctic landscape is to use an inverse approach: to study the fate of land derived material through the signal of terrestrial OC in the recipient aquatic environment. Terrestrial OC is transferred to the aquatic systems via streams, rivers, and erosion of coastlines directly to the seas. We view the streams and adjoining shelf waters as integrators of processes in their drainage basins, as well as potential reactors of the terrestrial OC. By studying terrestrial OC in the aquatic environment, we learn about land processes on a broader scale. The seas contain a mixture of terrestrial and marine OC, which creates a need for distinguishing between these pools when studying the terrestrial OC. Although the fate of terrestrial OC on the continental shelves has long been an active area of research, general discrepancies remain in closing the marine OC budgets. While marine OC is often considered more reactive than terrestrial OC, marine OC accounts for most of the sequestered OC in sediments (Burdige, 2005; Julies et al., 2010; Bianchi, 2011). In other words, there has been a missing sink of terrestrial OC in the marine compartment (Hedges et al., 1997). There are, however, growing indications of major portions of recalcitrant (unreactive and slow or unlikely to degrade) terrestrial OC being recycled upon entry in shelf waters (Keil et al., 1997; Aller and Blair, 2004). This thesis comprises studies of molecular degradation and source apportionment of terrestrial OC in the Siberian Arctic marine system through a range of geochemical tools. Radiocarbon age, for instance, is a vital tool for distinguishing between carbon from modern biological processes, and long freezelocked Pleistocene carbon. Throughout this thesis we generalize and refer to the old mineral-associated permafrost soil components from beneath the active layer as either yedoma ice complex deposits (ICD; Fig. 2) in the eastern Siberian Arctic, or deep permafrost OC in the western part of the study area. The aforementioned contemporary component of terrestrial OC is referred to as topsoil OC. Molecular proxies are used to trace the terrestrial matter and identify the provenance of various OC pools. By combining these tools, we seek to improve the current understanding of terrestrial OC as it passes over the land-ocean interface. This, we hope, will lead to a more detailed picture of mechanisms that act to mediate the potential transfer of OC from the biosphere to the atmosphere in the Arctic. 2. METHODS Study area The Eurasian Arctic continental shelf is the widest and shallowest of the world’s continental shelves (Stein and Macdonald, 2004). It reaches from the Kara Sea in the west at 50°E, past the Laptev Sea and the East Siberian Sea after which it meets the Pacific Ocean water near 165°E. Extending several hundred km offshore it has a water column depth of only 50 m on average (Jakobsson et al., 2008). The Eurasian Arctic shelf water receives fresh water and organic matter from several great Arctic rivers such as the Ob, Yenisey, Khatanga, Lena, Yana, Pechora, Indigirka and Kolyma Rivers. The Lena River is one of the largest rivers and its discharge is found far east as the major coastal currents and the Lena River discharge plume flow predominantly eastwards via the Dmitry Laptev Strait (Steele and Ermold, 2004; 14 Semiletov et al., 2005). River discharge mix with material from coastlines of the East Siberian Sea and Laptev Sea that to varying extents are comprised of yedoma ice complex deposits (ICD) extensively eroded during the open water season (Grigoriev et al., 1996; Lantuit et al., 2012; Günther et al., 2013). Sample collection The sample set used in this thesis was a collection of samples from the International Siberian Shelf Study. The main campaign was a 45 day expedition on the R/V Yakob Smirnitskyi in 2008 (ISSS-08; Semiletov and Gustafsson, 2009). There were three sub-campaigns to the ISSS-08. One conducted in the Buor-Khaya Bay on-board the R/V TB0012. One on-board the R/V Orlan in the Lena River delta area, and one on-board the passenger ship M/S Mechanic Kubilin along the main stem of the Lena River from Yakutsk to the Lena Delta. The samples used for the thesis work include mainly three phases of organic carbon and thus three types of samples: i) sediment, ii) particles from surface water and iii) colloidal matter from surface water. The sum of the carbon in the water column derived from unfiltered water, is referred to as the total OC (TOC) and is a combination of the dissolved, and particulate matter (Fig. 3). The colloidal pool is part of the dissolved pool, and sediment is the deposited particulate pool. Figure 3. Schematic relationship between different OC compartments. Total OC (TOC) consist of dissolved OC (DOC) and particulate OC (POC). DOC is divided into truly dissolved OC, and colloidal OC (COC). These pools are operationally defined by sampling methods. Sedimentary OC (SOC) is deposited POC found on lake- or sea-floors. Surface sediment samples were collected with a Van Veen grab sampler and used for paper #1, #2 and #3. The top three centimetres were subsampled with stainless steel spatulas from the collected grab samples and stored frozen in plastic containers. Samples were either freeze-dried for molecular analysis (paper #1 and #3), or used wet in incubation studies (paper #2). Particulate matter samples were retrieved through filtration of surface water (0-4 m depth) onto GF/F filters and used in paper #1, #3 and #4. GF/F is a type of glass fibre depth filter with a nominal cut-off of 0.7 µm. Sample sizes of up to 1000 L were GF/F filtered to allow molecular and isotopic analysis on the collected particles (high-volume GF/F filtration; Fig. 4). Particulate matter samples were freeze-dried before analyses (paper #1, #3 and #4). Filtrate from GF/F was filtered further onto membrane filters with a much smaller cut-off, at 1000 D. This method is referred to as cross flow filtration (CFF), tangential flow or ultrafiltration. The material collected is referred to as the colloidal pool or the high molecular weight dissolved pool. Through CFF you collect a concentrated fraction of material in a smaller volume of water, typically 0.5 to 1 L (called retentate) with concentration factors above 10. Colloidal samples were desalted (paper #3) and freezedried (paper #3 and #4) before analyses (Fig. 4). 15 Figure 4. Photographs of sample types in this thesis. Colloidal OC (COC) in polycarbonate bottles during two stages of freeze-drying, and particulate OC (POC) on GF/F filters in the stainless steel filter holder used for sampling. Analytical work The samples chosen for this thesis were analysed with bulk and molecular geochemical tools – organic carbon content, nitrogen content, stable carbon and radiocarbon isotopes, plant wax lipid and CuO oxidation product biomarkers. The theory and use of these analytical methods as well as their main limitations and value to this work are described in this section. An overview of the data collection and type of analysis done on which type of OC compartment is found in table 1. The biomarkers include a range of proxies which are given as an overview in table 2 and 3, where each proxy, its abbreviation(s), source(s), main function(s) and common use can be found. Bulk geochemical composition tools Bulk measures represent combined values from sometimes quite different sources, but often give a good indication of contributing sources or the dominant traits in mixed OC pools. Therefore, bulk measures were considered in both colloidal, particulate and sedimentary samples through this work. After freeze-drying and removal of inorganic carbon via acidification, samples were analysed for organic carbon content (TOC, or often only OC), total nitrogen content (TN) and stable carbon isotopes (δ13C) at either the Stable Isotope Facility at the University of California Davis (Davis, CA, USA) or at the Stable Isotope Laboratory (SIL) at the Department of Geological Sciences, Stockholm University (IGV, Stockholm, Sweden). Organic carbon content was used through this work to normalise biomarker proxies, since this is the fraction of the total carbon pool that many chemically hydrophobic markers such as lignin phenols and plant wax lipids partition into. (In contrast, acids, bases and metals, with stronger polarity are better 16 represented on the dry weight basis). The ratio between TOC and TN reflects the relationship between two of the three major constituents in living matter (phosphorus is the third, C:N:P; Redfield, 1959), differing among organism groups. Higher terrestrial plants exhibit C/N ratios over 20 while marine plankton commonly have C/N ratios of ~4-10 (Hedges et al., 1986; Meyers, 1994), although microbial degradation can sometimes preferentially remove nitrogen from an OC pool. Therefore the C/N ratio was used in conjunction with other proxies to distinguish terrestrial from marine sources. If samples exhibit low OC content, are very sandy or clayey, interpretations of an OC/TN ratio have to be done with regards to a potential inorganic nitrogen component in mind (Schubert and Calvert, 2001). The stable carbon isotope 13C comprises 1.1% of all C on Earth, but the proportion of the stable isotope 13 C incorporated into an organism compared to the 12C isotope varies and can provide information on different sources and processes. Fractionation, the preferential uptake or loss of 13C over 12C, occurs during many biological processes, such as carbon dioxide uptake by plants during photosynthesis, formation of marine carbonates, microbial degradation and photo-oxidation of organic matter. The ratio of 13C/12C is expressed as δ13C (and measured in ‰). Typical values of δ 13C with respect to the standard (Pee Dee Belemnite, PBD) were used in this thesis for source apportionment of different OC pools. A non-capitalized delta is used for δ 13C because it refers to a standard that does not change, in contrast to the capitalized delta in Δ14C, which is used because it refers to a changing value – atmospheric concentration at time of sampling (Coplen, 2011). Radiocarbon isotope Δ14C value is also a strong carrier of information, and was used in this work for identification of reservoir ages and transport times, as well as for distinguishing old and contemporary carbon pools. Samples for radiocarbon analysis were acidified for removal of inorganic carbon and sent to the National Ocean Sciences Accelerator Mass Spectrometry (NOSAMS) facility at Woods Hole Oceanographic Institution (WHOI, Woods Hole, MA, USA). Frequent swab tests are performed in the laboratory facilities were these samples were handled to prevent and avoid contamination. Age estimates are derived through quantifying the amount of the radiocarbon isotope 14C remaining in a sample compared to the concentration in the atmosphere at the time of sampling. Radiocarbon dating is limited to 60 000 years back since the estimate builds on the half-life of 14C (~5 000 years). Older material is said to be 14C-dead, and contemporary material is called modern if close to that of the atmospheric levels in 1950 (as defined by the oxalic acid standard). Samples in this thesis ranged from modern to 11 000 14C years. Monte Carlo simulation in end-member mixing model analysis In addition to qualitative source apportionment, statistical modelling was used to quantify the two main sources of terrestrial OC. The relative contributions of topsoil OC and yedoma ICD-OC, as distinguished from marine OC, were estimated using an end-member mixing model analysis (EMMA). This type of EMMA separates three sources based on two variables. In this work dual-carbon isotope values (δ13C and Δ14C) were used. Since all end-members represent a range of values, the input variability was accounted for by a random sampling strategy, a Monte Carlo simulation. The numerical background of the Monte Carlo EMMA simulation is provided in Andersson (2011). EMMA was applied for sediment OC (paper #1 and #3), particulate OC (paper #1 and #3) and for colloidal OC (paper #3). 17 Table 1. Type of analysis performed on colloidal OC (COC), particulate OC (POC) and sedimentary OC (SOC), from the sample collection in this thesis, divided into the papers they are used in. Organic carbon content (OC), plant wax lipid biomarkers (lipids), CuO oxidation reaction products (CuO products), stable carbon isotopes (δ¹³C) and radiocarbon isotopes (Δ¹⁴C). Paper #1 POC SOC OC 13 26 δ¹³C 14 23 Paper #2 SOC Incubations 11 CuO products 11 Paper #3 COC POC SOC OC 12 12 12 Paper #4 COC POC OC 10 10 Δ¹⁴C 6 8 lipids 12 22 δ¹³C 12 12 11 Δ¹⁴C 12 9 8 lipids 12 12 12 δ¹³C 10 10 Δ¹⁴C 10 10 CuO products 12 7 Molecular geochemical composition tools Physical and biological processes leave various chemical traces in the environment and through extraction of molecular components one can derive information on these processes. Biomarkers are thus measurable indicators of a biological state or condition. The presence of chlorophyll for example indicate that at some point in time (related to the chlorophyll degradation time), a photosynthetic organism has been present, or its remains have been transported there. Each biomarker carries constraints and is useful within certain conditions. The suitability of each biomarker molecule to provide information has to be taken into account for each sample type and environment they are applied to. Plant wax lipids and CuO oxidation reaction products, such as lignin phenols, are common biomarkers in marine geochemistry. I used both lipid and lignin markers as tracers of terrestrial matter, but also to assess relative degradation state. Both lipids and lignin represent parts of the bulk material, lipids commonly contribute 1% and lignin phenols 2-4% of the bulk OC in marine sediments, but their stability and specificity make them useful tools to understand terrestrial OC dynamics. The specific lipid and CuO oxidation product biomarkers are summarized in tables 2 and 3. The solvent extractable plant wax lipids were derived through soxhlet extraction or accelerated solvent extraction (ASE; 24 hours in dichloromethane and methanol, 2:1, v/v) resulting in a total lipid extract separated with regard to chemical polarities over Bond Elut and aluminium oxide columns in a workup protocol detailed in van Dongen et al., (2008). CuO oxidation reaction products were derived in alkaline CuO oxidation where samples were extracted under pressure and heating to 80°C in a microwave system following Goni and Montogomery (2000). For colloidal samples CuO oxidation was done with an 18 addition of glucose to avoid superoxidation (Louchcouarn et al., 2000). Measures were taken to avoid contamination, mainly of lipids from rubber and plastic materials, and laboratory work was performed in pre-combusted glass ware. Both plant wax lipids and CuO oxidation reaction products were analysed on GC-MS. High molecular weight (HMW) lipids are constituents of epicuticular waxes of higher plants (Eglinton, 1963) and are in principal absent from marine and freshwater plankton. During synthesis in higher plants the n-alkanes derive a predominance of odd chain lengths over even chain lengths in the range C24-C34 at a maxima around C27, C29 or C31 (Eglinton and Hamilton, 1967; Cranwell et al., 1973; Rieley et al., 1991). Therefore, a high abundance of HMW lipids, as well as the odd-over-even chain length pattern called the carbon preference index (CPI; Bray and Evans, 1961), is an indication of terrestrial matter in the aquatic environment (Farrington and Meyers, 1975). However, since the CPI decreases during degradation (Killops and Killops, 2005), it works also as a relative degradation state marker, in which higher indices reflect less degraded material (Table 3). Plant wax lipids were extracted for sedimentary OC (paper #1 and #3), particulate OC (paper #1, #3 and #4) and for colloidal OC (paper #3). Lignin is a phenolic biopolymer derived from vascular plants and not present in phytoplankton (Sarkanen and Ludwig, 1971; Stark et al., 1989), and was thus used for tracing terrestrial matter. The alkaline CuO oxidation also yield other products useful to trace sources and degradation state (Goni and Hedges, 1990a, b, 1995; Otto et al., 2006). Several of these products are specific to certain sources, such as marine plankton, angiosperm tissue, herbaceous tissue and Sphagnum (Hedges and Mann, 1979; Spencer et al., 2008; Lobbes et al., 2000; Table 3). The preferential loss of aldehydes compared to acids is the basis to also use lignin monomers to assess relative degradation state. The evaluation is done for the ubiquitous vanillyl and syringyl phenols through the ratio of aldehydes to acids (Ad/Al), in which lower indices reflect less degraded matter. Alkaline CuO oxidation was performed on sedimentary OC (paper #2 and #3) and on colloidal OC (paper #3). Closed system incubations of sediment in microcosm In addition to estimates of relative degradation state, laboratory based microcosm incubations were used to compare degradability in sediments from two contrasting areas. Microcosm is an artificial simplified system used to simulate certain properties of natural systems. These direct incubations were a simple closed system setup to measure the evolved carbon dioxide during anaerobic degradation of bulk sedimentary matter. Frozen sediment and artificial sea water were added to serum bottles that were sealed and flushed with N2-gas to remove oxygen. Three serum bottles were prepared for each of the four time points which were distributed over three weeks (initial value, 1 week, 2 weeks and 3 weeks). Part of the overlying headspace was withdrawn for injection on GC-FID to analyse the carbon dioxide concentration. A degradation rate was calculated from the carbon dioxide increase over the three weeks. OC pools include fast and slow OC pools that will degrade at different rates and on different time scales, and the calculated rate from this setup was derived from a fast pool degradation. The incubations were performed on sediment from the Buor-Khaya Bay in SE Laptev Sea and the Kolyma paleoriver canyon in the East Siberian Sea (paper #2). 19 Table 2. Source proxies derived from plant wax lipids and CuO oxidation reaction products. Abbreviation HMW n-alkanes LMW n-alkanes HMW/LMW n-alkanes Biomarkers Main source Reference High molecular weight (HMW) nalkanes, C20-C34 Low molecular weight (LMW) n-alkanes, C17-C19 High molecular weight n-alkanes over low molecular weight n-alkanes vascular plants - epicuticular plant waxes Eglinton and Hamilton, 1967; Simoneit 1997 marine phytoplankton, bacteria Schubert et al., 1997 V Vanillyl phenols C Cinnamyl phenols C/V S S/V P/V Pn/P Cinnamyl over vanillyl phenols Syringyl phenols Syringyl over vanillyl phenols p-Hydroxybenzenes over vanillyl phenols p-Hydroxyacetophenone over pHydroxybenzenes Cut Cutin acids FA Fatty acids from CuO oxidation DA Dicarboxylic acids from CuO oxidation B Benzoic acids from CuO oxidation terrestrial organic matter over marine organic matter (see HMW and LMW n-alkanes gymnosperm and angiosperm cell walls cell walls of non-woody vascular plants herbivore proxy - high ratio = woody material cell walls of angiosperms angiosperm proxy, a low ratio = gymnosperm (e.g. coniferous forest) typical of sphagnum, ratio > 6, else plankton ratio is high in mosses; particularly sphagnum vascular plants - epicuticular plant wax leaf tissue multiple sources, relatively abundant in phytoplankton, soil (bacteria) multiple sources, relatively abundant in phytoplankton, soil (bacteria) multiple sources, relatively abundant in phytoplankton, soil (bacteria) Eglinton and Hamilton, 1967; Schubert et al., 1997; Simoneit, 1997 Ertel and Hedges, 1984; Hedges et al., 1988 Ertel and Hedges, 1984; Hedges et al., 1988; Hartley, 1973 Goni et al., 2000; Stark et al., 1989; Hedges and Mann, 1979; Sarkanen and Ludwig, 1971 Ertel and Hedges, 1984; Hedges et al., 1988; Hartley, 1973 Goni et al., 2000; Stark et al., 1989; Hedges and Mann, 1979; Sarkanen and Ludwig, 1971 Williams et al., 1998; Amon et al., 2012 Williams et al., 1998; Amon et al., 2012 Riley and Kolattukudy, 1975; Goni and Hedges, 1990a,b Goni and Hedges, 1995; Otto et al., 2006; Wakeham et al., 1984 Goni and Hedges, 1995; Wakeham et al., 1984 Ertel and Hedges, 1984; Goni and Hedges, 1995; Otto and Simpson, 2006; Williams et al., 1998 20 Table 3. Relative degradation state proxies derived from plant wax lipids and CuO oxidation reaction products. Abbreviation Biomarkers n-alk CPI n-acid CPI HMW n-alkane carbon preference index (oddover-even-pattern), chain lengths C24-C34 HMW n-alkanoic acid carbon preference index (even-over-odd-pattern), chain lengths C20-C32 n-acids/ n-alks HMW n-alkanoic acids over HMW n-alkanes 3,5-Bd/V 3,5-Dihydroxybenzoic acid over vanillyl phenols Ad/Al Acid over aldehyde ratios, both valid for syringyl and vanillyl phenols, i.e. Sd/Sl and Vd/Vl Main source Reference leaf waxes of vascular plants leaf waxes of vascular plants leaf waxes of vascular plants vascular plants, 3,5Bd common in soil Bassel and Eglinton, 1983; Fahl and Stein, 1997 Bassel and Eglinton, 1983; Fahl and Stein, 1997 vascular plants Comment high index = less degraded high index = less degraded Brassel et al., 1984 high index = less degraded Prahl et al., 1994 low index = less degraded Opsahl and Benner, 1995, 1998; Amon et al., 2012 low index = less degraded 21 3. RESULTS and DISCUSSION The results of this thesis have been structured under three main themes. Sources, transport and degradation. These themes mirror the objectives of this work and the results are presented as a joint summary and discussion of the combined outcome from the included papers. Source apportioning of terrestrial OC Source apportionment was used in this work to identify terrestrial sources contributing to OC in the aquatic environment. The source identification was done in two ways for each of the carbon carrier phases. 1) Qualitative assessment and comparison of source proxies to distinguish marine from terrestrial matter. 2) Quantitative assessment of the relative contributions of the main terrestrial sources, topsoil OC and yedoma ICD-OC, in an end-member mixing model. Qualitative assessment. Several of the markers used to differentiate the terrestrial versus the marine OC pools indicated a heavy terrestrial dominance in OC over the Eurasian Arctic shelf. The δ13C values in the qualitative assessment of bulk SOC, POC and COC suggested an overall high contribution of terrestrial OC to the entire study area, indicated by relatively depleted δ13C values (mainly between -26 and -28 ‰). These δ13C values were in accordance with other bulk and terrestrial markers, namely relatively high OC/TN ratios, high concentrations of lignin phenols, high HMW lipid concentrations and high HMW/LMW n-alkane ratios. The high HMW/LMW n-alkane ratio also showed terrestrial OC consistently dominating over marine OC in both COC, POC and SOC (Fig. 5a). The terrestrial dominance was less pronounced in the very east, where Pacific-influenced water masses meets the East Siberian Sea; the latter having higher δ13C values. The marine signal towards the east, detected more clearly in a larger set of δ13C data (Vonk et al., 2012) is suggestive of a mixing trend between terrestrial OC with marine OC eastwards. However, vascular plant sources are varying east-westerly over the Laptev Sea and East Siberian Sea shelf region seen in a changing lignin phenol S/V ratio, and confirmed by a larger data set of CuO oxidation reaction products in SOC for the same area (Tesi et al., 2014). The terrestrial pool was further apportioned into sources derived from two different permafrost systems, the topsoil OC and the yedoma ICD-OC. Radiocarbon age was used to discriminate between them revealing regional east-westerly differences. The sediment in the eastern Siberian Arctic was much older (Δ14C ~-700‰) than in the western Kara Sea area (Δ14C ~-200‰). The distribution of older sediment ages in the eastern Eurasian Arctic margin support other reports of the same pattern (Guo et al., 2004; Gustafsson et al., 2011; Feng et al., 2013). The trend with increasing age toward the east was seen also in the colloidal pool, with ages varying from modern to 850 14C-years. The colloidal pool thus incorporates some component from older OC deposits despite being mostly modern. These regional radiocarbon differences likely reflect the high proportion of late Pleistocene yedoma ICD-OC in the eastern watersheds (Δ14C ~-940 ‰; Fig. 2), and the slightly younger age of deep peat deposits in the western Eurasian Arctic (Δ14C ~-530‰; Macdonald et al., 2006). 22 Figure 5. Source markers in colloidal OC (COC; white), particulate OC (POC; green) and sedimentary OC (SOC; black) on the Eurasian Arctic shelf from July/August 2008. A). Radiocarbon (Δ14C) over stable carbon (δ13C) isotope values. Also plotted are the end-member values for different sources derived from literature data. Detailed description of end-members can be found in paper #3 methods section. Briefly: The marine plankton end-member is split into two by geographical area, thus east and west of 165°E. Old permafrost OC is also split in two by geographical area, where yedoma ice complex deposit OC (ICDOC) exist in the eastern Siberian Arctic, but as deep peat permafrost soil (Deep-PF OC) in the western watersheds, i.e. for Kara Sea samples. Topsoil OC is split into the carrier phases dissolved OC (DOC) and particulate OC (POC and SOC). 23 Figure 6. Concentration of lignin phenols over radiocarbon isotope values (Δ14C) in colloidal OC The lignin vascular plant derived markers in the colloidal pool were however derived from a younger source, seen in a lignin concentration highly associated with young bulk age (Fig 6). The bulk geochemical markers together demonstrate a terrestrial OC dominated aquatic carbon cycle. The terrestrial dominance is manifested throughout the three studied carrier phases and reflects a high contribution from varied vascular plant sources and a marine influence in the east Quantitative assessment. In the end-member mixing model, topsoil OC and yedoma ICD-OC were found in different proportions in the water column compared to the underlying sediment (paper #1, #3, and #4). The relative contribution of topsoil OC dominated both the colloidal (66%) and particulate phase (62%) throughout the study area, with only minimal contribution from yedoma ICD-OC and other old permafrost (9% and 13% respectively; paper #3; Fig. 7). These relative contributions reflect the buoyant nature of the OC suspended in the water column, and the low density of humic aggregates and vegetation debris that are derived from upper soil horizons and surface soils part of the topsoil OC source. In contrast to the colloidal and particulate phases in the water column, the underlying sediment contained a smaller contribution from topsoil OC (23%), and a dominant ICD-OC contribution (54%; Fig. 7). The dominant yedoma ICD-OC fraction in sediment likely reflect the nature of its mineral component, which gives sinking properties to the OC rather than ability to stay suspended, explaining the high proportion of the yedoma ICD-OC at the seafloor. These results highlight partitioning of sources as highly dependent on physical properties, and imply that source contribution in the different carrier phases are related to its vertical and lateral transport patterns. Figure 7. Fractional contributions of topsoil OC (grey), yedoma ICD-OC plus deep PF-OC (red) and marine OC (blue) to colloidal OC (COC), particulate OC (POC) and sedimentary OC (SOC) on the Eurasian Arctic shelf from July/August 2008. 24 Transport and transportability Dispersion patterns of the carbon carrier phases were used to derive how different sources have been transported over the study area. Transport of topsoil OC and yedoma ICD-OC are closely coupled to their physical properties, and were found to exhibit different spatial distributions throughout the data set, thus both in the Lena River and the Eurasian Arctic continental shelf. Figure 8. Surface plots of Δ14C (‰) in surface water particulate OC and surface sediment OC in Buor-Khaya Bay, Laptev Sea from July/August 2008. Coastal erosion hotspot Muostakh Island denoted via text box. The topsoil OC was found as young OC in all surface water samples, but was of lesser relative abundance in the underlying sediment (paper #1, and #3; compare panel a and b in Fig. 8). The presence and dominance of topsoil OC in surface waters far out on the shelf suggest that topsoil OC is more readily advected than yedoma ICD-OC. In parallel, the topsoil OC only contributes a small fraction to surface sediment OC (topsoil OC is young and less depleted in Δ14C), which indicates that topsoil OC mostly stays suspended in the water column. Therefore, its fate is affected by lateral water movement and wind patterns. The difference in topsoil OC contribution between water column and sediment (Fig. 8) could be a sign of decomposition, or could reflect an input of yedoma ICD-OC to the seafloor from lateral transport. The latter is more likely since there was a strong vertical export of yedoma ICD-OC from the surface water, as seen in high but rapidly decreasing concentrations around coastal erosion hotspots (paper #1 and #3; Fig. 8). A vertical transport of yedoma ICD-OC is thus rapid and must happen within a few kilometres from the source. This is particularly evident in Buor-Khaya Bay where the fraction yedoma ICD-OC was strongly and negatively correlated with water depth (paper #1; Fig. 9a). The high near-shore concentrations of yedoma ICD in POC suggest that yedoma ICD-OC partitions into POC, but sinks out rapidly after entering the water (paper #1, and #3). The yedoma ICD is a mineral-carbon matrix with high density that rapidly sinks with the associated OC out of the water column. This means that very little yedoma ICD-OC are transported long distances over the Eurasian Arctic shelf by advection in surface waters. 25 Figure 9. The fraction OC contributed by topsoil OC (green triangles), yedoma ICD-OC (black squares) and marine plankton OC (blue circles) plotted over water column depth. A) Particulate OC (POC) and B) sedimentary OC (SOC) in Buor-Khaya Bay from July/August 2008. 26 Despite the presumed rapid settling of yedoma ICD-OC from water column to sediment, the stations far offshore (~500km) also have large shares of yedoma ICD-OC in SOC (paper #1 and #3; Fig. 9b). These large, relatively even and dominant contributions of yedoma ICD-OC to SOC also on stations far offshore (paper #3) suggest that ICD-OC is laterally transported and dispersed near the seafloor. Benthic or nepheloid layer transport (near-bottom transport; Fig. 1) has previously been reported to be a strong carrier of seafloor material over the Eurasian Arctic shelf (Are et al., 2002; Dudarev et al., 2006; Charkin et al., 2011) and is one of two possible carriers of yedoma ICD-OC. Such a near-bottom transport mechanism could potentially disperse a mineral-heavy fraction like the yedoma ICD-OC over the seafloor without it being spread to the overlying particulate or suspended OC pools. Near-bottom transport is potentially high on the Eurasian Arctic shelf due to turbidity and shallow water column depths, but there are other lateral transport mechanisms. A possible transport pathway of yedoma ICD-OC is the incorporation in sea ice (called ice rafting; Fig. 1). Ice rafting, suggested to partly control sediment distribution (Reimnitz et al., 1998; Viscosi-Shirley et al., 2003), particularly in ice export from the Laptev and East Siberian Sea (Nürnberg et al., 1994; Eicken et al., 1997, 2000), could equally result in a wide dispersal of yedoma ICD-OC as seen in this data from the shelf area. Degradation state and degradability Two different approaches were used to assess degradation of OC, one through the use of molecular biomarker proxies and the other by closed system incubations of sediment. Degradation is often closely linked to transport pathways and properties of the source material; therefore the connections between sources, transport and degradation are also discussed here. The topsoil OC and ICD-OC clearly have different behaviours, regarding both transport and degradation, once released to the shelf waters of the Eurasian Arctic. The two pools exhibit differences in degradation state closely connected to their respective tendencies for long-range transport. Plant wax lipid markers show a clear decoupling of degradation states between surface water OC (POC and COC) and the underlying SOC. The terrestrial OC component is more degraded in POC than in the SOC (paper #1, and #3; Fig. 10). Similarly, the COC is more degraded than the SOC. Degradation state markers from CuO oxidation (3,5-Bd/V, Sd/Sl and Vd/Vl; Table 2) indicate a highly degraded state of the terrestrial COC fraction (paper #3). Lateral transport of COC, following the surface water movement is likely faster than the near-bottom lateral transport of SOC. A shorter time frame which allows less time for degradation, and yet terrestrial COC is more degraded than terrestrial SOC. However, the COC is suspended in the water column and thus greater exposure, potentially to both aerobic-microbial and photo-oxidative processes, during lateral waterborne transport, which could explain the higher degradation state. The SOC displayed a less degraded terrestrial signal relative to both POC and COC (paper #1, and #3). However, around coastal erosion hotspots POC was also less degraded (Fig. 10a and c). The POC around erosion hotspots were at the same time old, as seen in the radiocarbon data (Fig. 8a), suggesting that the yedoma ICD-OC is less degraded. This low degradation state, and the widespread dispersal of yedoma ICD-OC in this data set, suggests that yedoma ICD-OC is degrading less and instead rather accumulating on the seafloor of the Eurasian Arctic shelf. 27 Figure 10. Surface distribution plots of surface water particulate OC and surface sediment OC from BuorKhaya Bay, SE Laptev Sea in July/August 2008. A) and B) high molecular weight (HMW) n-alkanoic acids over HMW n-alkanes. B) and C) HMW n-alkane carbon preference index (CPI). An explanation for the rapid sinking and preferential preservation of yedoma ICD-OC is the mineral matrix supplying protection against degradation. The mineral-carbon association could be a physical protection that prevents enzymatic cleavages of the adsorbed OC, as seen for OC on continental margins by many (e.g. Mayer et al., 1994; Keil and Mayer, 2013). In addition, burial in shelf sediment could cause less efficient degradation of OC (e.g. Hartnett et al., 1998), potentially delaying or halting degradation of yedoma ICD-OC. Degradation was directly measured through the formation of carbon dioxide, in sediment dominated by yedoma ICD-OC in closed system incubations. Comparison of carbon dioxide formation rates in conjunction with the molecular markers and CuO oxidation reaction products revealed two contrasting degradation regimes. Degradation rates correlated well with molecular markers of marine OC in the East Siberian Sea and of terrestrial OC markers in the Buor-Khaya Bay area. Furthermore, the degradation rate was six times faster for the East Siberian Sea samples than for the Buor-Khaya Bay samples (Fig. 11). Marine OC, generally considered to be more labile than terrestrial OC, was more abundant in the East Siberian Sea than in the Buor-Khaya Bay. However, the difference between the marine OC fractions was not proportional to the difference in degradation rate, i.e. there was a higher carbon dioxide formation 28 rate than predicted from the marine OC fraction in the East Siberian Sea samples. In addition, the two shelf systems had similar geochemical signatures of terrestrial OC, suggesting that differences in terrestrial OC composition are not the cause of the higher degradation rates in the East Siberian Sea. The differing degradation behaviour of sediment from the two areas may be a manifestation of many mechanisms: selective protection could favour a certain OC pool, co-metabolism (priming) could cause loss of more refractory terrestrial OC when marine OC is readily metabolized, and/or microbial community efficiencies could differ. Figure 11. Results from direct incubations of surface sediment OC (SOC) from Buor-Khaya Bay, SE Laptev Sea and Kolyma Paleoriver Canyon in East Siberian Sea in July/August 2008. A) Typical patterns of evolved CO2 (μmol CO2 g OC−1) illustrated by station #51 in LS-Buor-Khaya Bay (anaerobic, blue triangles), and station #36 on the East Siberian Sea Kolyma Paleoriver Canyon (aerobic, white squares and anaerobic, blue squares). B) Evolved CO2 at four time points (T0–T3 in consecutive order; shown by bar graphs where lightest gray is T0, darkest gray T3) during anaerobic incubation of surface sediments from Laptev Sea-Buor-Khaya Bay and East Siberian Sea-Kolyma Paleoriver canyon on the primary axis. Scatter plot (blue diamonds) showing the anaerobic degradation rate (μmol CO2 g OC−1 day−1) of the same incubations on the secondary axis. 29 When taking the results in this thesis together it is clear that the two terrestrial OC sources, topsoil OC and yedoma ICD-OC, have different fates regarding both transport and degradation over the Eurasian Arctic shelf system. Topsoil OC appears to be degrading more during waterborne transport in COC or POC, during river transport, or in the surface waters of the Eurasian Arctic shelf, and to some extent also transported over long distances suspended in surface water. In stark contrast, yedoma ICD-OC settled out in the near vicinity of its source release site, and maintained a terrestrial component (preferentially preserved) throughout transport over the East Siberian shelf sea floor. 4. CONCLUSIONS From the results in this thesis, a set of conclusions can be drawn regarding sources, transport and degradation of the terrestrial matter transported to the Eurasian Arctic shelf. The bulk and molecular geochemical composition of all the carbon carrier phases indicate the marine environment of the Eurasian Arctic shelf is dominated by terrestrial OC. This terrestrial dominance is manifested throughout the three studied carrier phases and reflects a high contribution from varied vascular plant sources, and a marine influence where the East Siberian Sea meets Pacific influenced waters near 165°E. This terrestrial dominance suggest that other processes in these marine environments, (e.g. primary production, gas exchanges between sediment-oceanatmosphere, inorganic carbon dynamics, pH) are influenced by the degradation and translocation of large amounts of terrestrial OC. The two main terrestrial sources throughout this data set, topsoil OC and ICD-OC, are transported differently over the Eurasian Arctic shelf system. The effect of transport of topsoil OC and yedoma ICD-OC is seen in their distributions in the three carrier phases. The end-member mixing model suggests that topsoil OC and yedoma ICD-OC are present in different proportions in the water column, as opposed to the underlying sediment. These proportions indicate that yedoma ICD-OC is accumulating on the shelf in a relatively undegraded state suggesting that the Eurasian Arctic shelf is a sink of yedoma ICD-OC. The observation that topsoil OC contributes mainly to the particulate and colloidal pools suggests that topsoil OC is transported out on the shelf in the surface water. The fact that topsoil OC is derived from low-density humic surface horizons of the active layer, indicates that it is easily suspended in the water column. Long range transport in surface water and exposure to UV radiation and microbial degradation may explain the highly degraded nature of the topsoil OC. This could imply that there is emission of carbon dioxide during the transport of topsoil OC in the rivers and during transport to the sea. It is not clear however whether the degradation is taking place during the transport, or close to (or at the) source site. 30 The distinct export out of the water column, and the dominance of yedoma ICD-OC in surface sediment far out on the shelf suggests that yedoma ICD-OC is laterally transported and dispersed over the shelf, likely through near-bottom transport or via ice rafting. The rapid export of yedoma ICD-OC from the water column to the underlying sediment may explain the relatively undegraded state of yedoma ICD-OC. The mineral matrix could provide protection to the associated OC, either through its high density causing burial in shelf sediment, or through the sorption causing physical hindrances to enzymatic cleavage and degradation. The undegraded state of yedoma ICD-OC, and the slower degradation of sediment that correlated to terrestrial proxies in incubations, suggest that yedoma ICD-OC may not be a strong source of atmospheric carbon dioxide on the Eurasian Arctic shelf. The yedoma ICD-OC may instead be contained, redistributed and/or exported off the shelf. The rapid formation of carbon dioxide in the incubation experiments suggests that sediment dominated by yedoma ICD-OC is degradable. In addition, there was a higher carbon dioxide formation rate than anticipated from the marine OC fraction in the East Siberian Sea compared to in Laptev Sea, proposing that the two areas have different degradation regime. Potential explanations to the different degradation regimes could involve selective preservation, microbial community composition or co-metabolism. In the case of co-metabolism increased primary production could enhance degradation of terrestrial ICD-OC in Eurasian Arctic shelf sediment. 5. PERSPECTIVES on FURTHER RESEARCH In this section I present some observations and suggestions for future work that would continue or strengthen the results obtained in this thesis. Preservation of yedoma ICD-OC Role of oxygen Oxygen exposure is known to affect degradation processes of organic matter in continental margin sediments (Hartnett et al., 1998; Keil et al., 2004; Burdige et al., 2007), and may play a particular role in control of yedoma ICD-OC degradation. The high density mineral-association of yedoma ICD causes it to sink out of oxygenated water into anoxic sediment layers where degradation is less efficient. This export and subsequent burial may (temporarily) halt or delay degradation of yedoma ICD-OC. While oxygen penetration depths may have been studied elsewhere, it would also be a natural step in this research area to investigate oxygen penetration depths in Eurasian Arctic shelf sediments, and relate this to penetration depths of mineral-associated yedoma ICD. This could for example be done through oxygen microelectrode measurements in combination with age profiles in sediment. Understanding the role of oxygen is relevant for quantifying long term storage processes, e.g. how long the shelf may be a sink of yedoma ICD-OC. Sorption of OC Mineral-association have been suggested to provide protection to OC in marine sediments (Keil et al., 1994b, Hedges et al., 1997, Mayer, 2004). However, the actual mechanisms behind mineral-associated protection in yedoma ICD-OC are yet to be detailed. Work could be directed specifically towards testing 31 sorption properties in organo-mineral complexes in yedoma ICD-OC, both on the Eurasian Arctic continental margins, or as sorption mechanisms in the yedoma ICD itself. Investigations of particular sorption phases for yedoma ICD-OC could also be done, as in Kaiser and Guggenberger, (2000) where sorption of DOM in soils related to aluminium and iron oxyhydroxides were investigated. Perhaps surface areas of manganese, that have been seen to provide large active surface areas in sediment, can play a role too. Additionally, different size fractions of the particulate (and sedimentary) pool could perhaps show different sorption properties and loading of topsoil OC and yedoma ICD-OC connected to surface area. Sorption processes may thus speed up or slow down degradation of yedoma ICD-OC accumulated on the Eurasian Arctic shelf. Heterogeneity and Hydrology The yedoma ICD itself is quite heterogeneous. Subdividing yedoma ICD into several subpools has, for example, been done to assess differences in reactivity (Dutta et al., 2006; Zimov et al., 2006). Yedoma ICD that comprises ice-rich permafrost of varying OC content is intersected with large ice wedges (Schirrmeister et al., 2002). Different amounts of ice included in incubations have been observed to induce different rates of degradation (Vonk, J. personal communication). The dry environment of frozen soil is not favourable to processing of organic material, and the onset of melt does not necessarily result in a wet environment, despite the conversion of ice to water. This means that since microbial communities are dependent on suitable hydrological conditions, thawing does not necessarily induce microbial activity. These effects of heterogeneity and hydrology in permafrost soils suggest that sampling strategies of yedoma ICD-OC likely play an important role in trying to quantify degradation processes, and sampling procedures should therefore be carefully designed. Spatial and temporal coverage This thesis has shown how composition, source, and transport of terrestrial organic material can affect the organic carbon composition of the Eurasian Arctic shelf environment, but there are many remaining questions regarding its long-term fate. For instance, the shelf system appears to function as a sink for yedoma ICD-OC, but it is unclear whether yedoma ICD-OC is buried on the shelf, degraded on a longer time scale, or exported off the shelf to the slope and central Arctic basin. Additional research to address these questions with studies of better areal coverage and temporal resolution is needed. Longer-term monitoring, seasonal studies, details of spring flood signatures - especially of particulate OC, winter carbon cycling - both coupled to inland ice and sea ice dynamics, as well as interannual comparisons would all give a better understanding of the fate of terrestrial matter transported to the Eurasian Arctic shelf. 32 ACKNOWLEDGEMENTS I want to start with thanking my family Tobias och Isak! Ni är de klokaste små människor jag känner. Tack för att ni funnits vid min sida under alla doktorandåren. Tack för att ni skällt på mig när jag varit odräglig, för att ni skrattat med mig när livet varit tungt, och för att ni stått ut med riktigt halvdassig mat, fler dammtussar i hörnen och färre lekstunder när tiden varit knapp. Jag är så stolt över er! Vill också säga tack till min övriga familj. Mamma, du ställer upp massor, och vad vi har skrattat, men jag hör på dig när du är orolig. Linus min älskade bror, och kära Frida. Pappa, som alltid blir rörd och är stolt över mig. Birthe vid hans sida. Tänker ofta på min morfar Bengt Gunnar Ellström som min stora saknad. Tänker att han gett mig inspiration, nyfikenhet och lugn. Han var forskare i försvaret i Halmstad. Vi var hans ögonstenar, och han log alltid med ögonen. Sörjer min farmor som var mig kär, och gick bort väldigt nyligen. Hon var min allra mest frekventa barnvakt, och den jag rymde till om det behövdes. Jag lovar att fortsätta sjunga dina sånger och ställa fram Gröna marmeladkulor. Brett. What can I say, the only one who dared visit with some groceries when I was down with the swine flu. You are me very dear, and what are you? My discussion partner, my computer nerd, my life philosopher, my greatest annoyance, my most time consuming but enjoyable friend. Thank you for taking part. The Arctic Carbon group at my department. Julia, what words to use? I am not sure. I shared a lot with you. Queen of gossip, queen of joy, queen of action! I totally love you. Marc – hang in there. You are great. Henry! <3 The wise and witty - let us share more time. August! Your laughter rings in my ears. Keep at it. Teach me play violin? I will continue sharing bird nerd stuff. Anton – office mate to trust! Martin my dear shift mate – we did a really good job, thanks! Lisa, Tommaso, Barbara, Nina, Pete. The SWERUS Arctic ocean expedition preparations and onboard time. Ups and downs right? What an adventure. Thanks for sharing it with me. All the old ITM current ACES faces. Frida, Cissi, Karin N, Hanna G, Anna S, ni har alla varit känslomässiga bollplank åt mig. Tack! Det har hjälpt. Yngve, tack för dansant sällskap, men också för all lab-hjälp. Alla glada ansikten i korridoren. Behöver flera rader med namn. Tack. Jorien och Laura – solvent extractable lipids right? Thanks for teaching me! Also for chats and bits during my first year, when life was tough. Johan Gelting. Min tidigare rumskompanjon. Du var ovanligt artig och trevlig att ha i samma rum. Tack för alla pratstunder om viktiga ting både utanför och i jobbet. Michael McLachlan. You said to me: there is no need to be this sad. Words that I carry with me. Thanks for listening all those times. Volker – min biträdande handledare. Tack mest av allt för att du tog dig tid för min skull trots ansökningar, undervisning och annat här på sluttampen. Det var mycket värt! Only responsible for the science right? ;) hahaha. 33 Örjan – vad skriver man om dig? Tack för att du kallat mig rising star, tack för att du sagt att jag gått från klarhet till klarhet. Du har varit tuff att jobba med, men jag har lärt mig en hel del. Både om sociala ting och om hur man gör forskning. Du har lämnat efter dig stora och små minnen som jag tar med mig. Jag har sett – du jobbar oerhört intensivt, du är familjefar till tre, och du räcker inte alltid till. Är trots alla turer imponerad av dig – för jag tror att det är ensamt på toppen, men du gör också ett oerhört bra jobb. Kör på positiva linjen, den behövs. I also want to thank our Russian collaborators. Spatsiba! It has been a pleasure to learn and be inspired, much by your field work competence. Oleg! Yikes, how many times have you been at longer expeditions? Over 30 - I am impressed. Also - thanks for helping me with “water catastrophy” at the SWERUS expedition. Sasha Charkin. My reliable team mate for submersible pump action. Kseniya for Russian poetry! Igor – also the night shift ;), and thanks for pointing out Cape Emma and Cape Sofia on Bennet island. Malin S och Karin på kommunikationssidan. Tack så hjärtligt för att ni visar att ni uppskattat/r min förmåga att kommunicera vetenskap. Visst är det otroligt kul. Who did I forget? Probably a lot of people. People that meant something for me through my life. I am not listing you all here. I do however want to thank all my friends outside of work! Now that I have my life back I will become visible again. Haha. All welcome to overdose me with texts, emails, chats, late evenings, long walks, day trips and other. - Itaí min Itaí -. Katarina och Håkan. Cecilia Lindh. Petra och Niklas. Maria och Stefan. Therese. Susanna. Bodil. Ulrica. Johan M. Pähr H. Marc Mathison. Gabriella-papp. Jakangänget. Maarit my dear red rog. (No doors in Finland still?) Next to last: Raymond Pierrehumbert. Ray you inspire me – people are rarely this humble and still dare to act with power upon the will to make change to this world. I want to - hope to, and will somehow also do so, to the best of my ability. Golden stars and rays of sun to you! (Everyone – listen to this man). Lastly I want to thank myself. Great job Emma. You really achieved a lot in these years. I am proud of you. Vänligen /hon-som-får-saker-gjort, isbjörnsmamman, polarfararen, pemmis, super-Emma. 34 The road not taken Two roads diverged in a yellow wood, And sorry I could not travel both And be one traveler, long I stood And looked down one as far as I could To where it bent in the undergrowth; Then took the other, as just as fair And having perhaps the better claim, Because it was grassy and wanted wear; Though as for that the passing there Had worn them really about the same, And both that morning equally lay In leaves no step had trodden black. Oh, I kept the first for another day! Yet knowing how way leads on to way, I doubted if I should ever come back. I shall be telling this with a sigh Somewhere ages and ages hence: Two roads diverged in a wood, and I — I took the one less traveled by, And that has made all the difference. 35 Muostakh Island coastal ice complex in winter 2007, photo Alexander Charkin REFERENCES ACIA, 2004. Impacts of a Warming Arctic: Impact Assessment. Cambridge University Press, Cambridge, UK. Aller, R.C. and Blair, N.E., 2004. Early diagenetic remineralization of sedimentary organic C in the Gulf of Papua deltaic complex (Papua New Guinea): Net loss of terrestrial C and diagenetic fractionation of C isotopes. Geochimica Et Cosmochimica Acta, 68(8): 18151825. Amon, R.M.W. et al., 2012. Dissolved organic matter sources in large Arctic rivers. Geochimica Et Cosmochimica Acta, 94: 217-237. Andersson, A., 2011. A systematic examination of a random sampling strategy for source apportionment calculations. Science of the Total Environment, 412: 232-238. Are, F. et al., 2002. Comparative Shoreface Evolution along the Laptev Sea Coast. Polarforschung, 70: 135-150. Barnhart, K.R., Overeem, I. and Anderson, R.S., 2014. The effect of changing sea ice on the vulnerability of Arctic coasts. The Cryosphere Discuss., 8(3): 2277-2329. Bianchi, T.S., 2011. The role of terrestrially derived organic carbon in the coastal ocean: A changing paradigm and the priming effect. Proceedings of the National Academy of Sciences of the United States of America, 108(49): 19473-19481. 36 Bray, E.E. and Evans, E.D., 1961. Distribution of normal-paraffins as a clue to recognition of source beds. Geochimica Et Cosmochimica Acta, 22(1): 2-15. Brown, J., O.J. Ferrians, Jr., J.A. Heginbottom, and E.S. Melnikov. 1998, revised February 2001. Circum-arctic map of permafrost and ground ice conditions. Boulder, CO: National Snow and Ice Data Center. Digital media. Burdige, D.J., 2007. Preservation of organic matter in marine sediments: Controls, mechanisms, and an imbalance in sediment organic carbon budgets? Chemical Reviews, 107(2): 467485. Charkin, A.N. et al., 2011. Seasonal and interannual variability of sedimentation and organic matter distribution in the Buor-Khaya Gulf: the primary recipient of input from Lena River and coastal erosion in the southeast Laptev Sea. Biogeosciences, 8(9): 2581-2594. Cranwell, P.A., 1973. Branched-chain and cyclopropanoid acids in a recent sediment. Chemical Geology, 11(4): 307-313. Dickson, R.R. et al., 2000. The Arctic Ocean response to the North Atlantic oscillation. Journal of Climate, 13(15): 2671-2696. Dudarev, O.V., Semiletov, I.P., Charkin, A.N. and Botsul, A.I., 2006. Deposition settings on the continental shelf of the East Siberian Sea. Doklady Earth Sciences, 409(6): 1000-1005. Dutta, K., Schuur, E.A.G., Neff, J.C. and Zimov, S.A., 2006. Potential carbon release from permafrost soils of Northeastern Siberia. Global Change Biology, 12(12): 2336-2351. Eglinton, G. and Hamilton, R.J., 1967. Leaf epicuticular waxes. Science, 156(3780): 1322-1335. Eicken, H. et al., 2000. A key source area and constraints an entrainment for basin-scale sediment transport by Arctic sea ice. Geophysical Research Letters, 27(13): 1919-1922. Eicken, H. et al., 1997. Sea-ice processes in the Laptev Sea and their importance for sediment export. Continental Shelf Research, 17(2): 205-233. Ertel, J.R. and Hedges, J.I., 1984. The lignin component of humic substances – distribution among soil and sedimentary humic, fulvic, and base-insoluble fractions. Geochimica Et Cosmochimica Acta, 48(10): 2065-2074. Farrington, J.W. and Meyers, P.A., 1975. Hydrocarbons in the marine environment. In: G. Eglinton (Editor), Environmental Chemistry. The Chemical Society, London, pp. 109136. Feng, X. et al., 2013. Differential mobilization of terrestrial carbon pools in Eurasian Arctic river basins. Proceedings of the National Academy of Sciences of the United States of America, 110(35): 14168-14173. Goni, M.A. and Hedges, J.I., 1990a. Cutin-derived CuO reaction-products from purified cuticle and tree leaves. Geochimica Et Cosmochimica Acta, 54(11): 3065-3072. Goni, M.A. and Hedges, J.I., 1990b. Potential applications of cutin-derived CuO reactionproducts for discriminating vascular plant sources in natural environments. Geochimica Et Cosmochimica Acta, 54(11): 3073-3081. Goni, M.A. and Hedges, J.I., 1995. Sources and reactivities of marine-derived organic-matter in coastal sediments as determined by alkaline CuO oxidation. Geochimica Et Cosmochimica Acta, 59(14): 2965-2981. Goni, M.A. and Montgomery, S., 2000. Alkaline CuO oxidation with a microwave digestion system: Lignin analyses of geochemical samples. Analytical Chemistry, 72(14): 31163121. Grigoriev, M.N., 1993. Cryomorphogenesis in the Lena Delta. Yakutsk: Permafrost Institute Press (in Russian). 37 Grigoriev, M.N. et al., 1996. Geology, seismicity and cryogenic processes in the arctic areas of Western Yakutia. Yakut Scientific Centre SD RAS, Yakutsk, 84. Grigoriev, M.N., Rachold, V., Schirrmeister, L. and Hubberten, H.-W., 2004. Organic carbon input to the Arctic Seas through coastal erosion. Günther, F., Overduin, P.P., Sandakov, A.V., Grosse, G. and Grigoriev, M.N., 2013. Short- and long-term thermo-erosion of ice-rich permafrost coasts in the Laptev Sea region. Biogeosciences, 10(6): 4297-4318. Guo, L.D. et al., 2004. Characterization of Siberian Arctic coastal sediments: Implications for terrestrial organic carbon export. Global Biogeochemical Cycles, 18(1). Gustafsson, O., van Dongen, B.E., Vonk, J.E., Dudarev, O.V. and Semiletov, I.P., 2011. Widespread release of old carbon across the Siberian Arctic echoed by its large rivers. Biogeosciences, 8(6): 1737-1743. Hartley, R.D., 1973. Lignin carbohydrate linkages in plant cell-walls. 1. Carbohydrate esters of ferulic acid as components of cell-walls of lolium-multiflorum. Phytochemistry, 12(3): 661-665. Hartnett, H.E., Keil, R.G., Hedges, J.I. and Devol, A.H., 1998. Influence of oxygen exposure time on organic carbon preservation in continental margin sediments. Nature, 391(6667): 572-574. Hedges, J.I., Clark, W.A. and Cowie, G.L., 1988. Organic matter sources to the water column and surficial sediments of a marine bay. Limnology and Oceanography, 33(5): 11161136. Hedges, J.I. et al., 1986. Compositions and fluxes of particulate organic material in the Amazon River. Limnology and Oceanography, 31(4): 717-738. Hedges, J.I., Keil, R.G. and Benner, R., 1997. What happens to terrestrial organic matter in the ocean? Organic Geochemistry, 27(5-6): 195-212. Hedges, J.I. and Mann, D.C., 1979. Characterization of plant-tissues by their lignin oxidationproducts. Geochimica Et Cosmochimica Acta, 43(11): 1803-1807. Heginbottom, J.A., 1984. The mapping of permafrost. Canadian Geographer-Geographe Canadien, 28(1): 78-83. Heginbottom, J.A., 2002. Permafrost mapping: a review. Progress in Physical Geography, 26(4): 623-642. Jeffries, M.O., Richter-Menge, J. and Overland, J.E., 2014. Arctic Report Card 2014. In: M.O. Jeffries, J. Richter-Menge and J.E. Overland (Editors). Julies, E.M., Fuchs, B.M., Arnosti, C. and Bruchert, V., 2010. Organic Carbon Degradation in Anoxic Organic-Rich Shelf Sediments: Biogeochemical Rates and Microbial Abundance. Geomicrobiology Journal, 27(4): 303-314. Kaiser, K. and Guggenberger, G., 2000. The role of DOM sorption to mineral surfaces in the preservation of organic matter in soils. Organic Geochemistry, 31(7-8): 711-725. Keil, R.G., Dickens, A.F., Arnarson, T., Nunn, B.L. and Devol, A.H., 2004. What is the oxygen exposure time of laterally transported organic matter along the Washington margin? Marine Chemistry, 92(1-4): 157-165. Keil, R.G., Mayer, L.M., Quay, P.D., Richey, J.E. and Hedges, J.I., 1997. Loss of organic matter from riverine particles in deltas. Geochimica Et Cosmochimica Acta, 61(7): 1507-1511. Keil, R.G., Montlucon, D.B., Prahl, F.G. and Hedges, J.I., 1994. Sorptive preservation of labile organic-matter in marine sediment. Nature, 370(6490): 549-552. Killops, S. and Killops, V., 2005. Introduction to Organic Geochemistry. Blackwell. 38 Koven, C.D. et al., 2011. Permafrost carbon-climate feedbacks accelerate global warming. Proceedings of the National Academy of Sciences of the United States of America, 108(36): 14769-14774. Lantuit, H. et al., 2011. Coastal erosion dynamics on the permafrost-dominated Bykovsky Peninsula, north Siberia, 1951-2006. Polar Research, 30. Lantuit, H. et al., 2012. The Arctic Coastal Dynamics Database: A New Classification Scheme and Statistics on Arctic Permafrost Coastlines. Estuaries and Coasts, 35(2): 383-400. Lobbes, J.M., Fitznar, H.P. and Kattner, G., 2000. Biogeochemical characteristics of dissolved and particulate organic matter in Russian rivers entering the Arctic Ocean. Geochimica Et Cosmochimica Acta, 64(17): 2973-2983. Louchouarn, P., Opsahl, S. and Benner, R., 2000. Isolation and quantification of dissolved lignin from natural waters using solid-phase extraction and GC/MS. Analytical Chemistry, 72(13): 2780-2787. MacDonald, G.M. et al., 2006. Rapid early development of circumarctic peatlands and atmospheric CH4 and CO2 variations. Science, 314(5797): 285-288. MacDougall, A.H., Avis, C.A. and Weaver, A.J., 2012. Significant contribution to climate warming from the permafrost carbon feedback. Nature Geoscience, 5(10): 719-721. Mayer, L.M., 1994a. Relationships between mineral surfaces and organic-carbon concentration in soils and sediments. Chemical Geology, 114(3-4): 347-363. Mayer, L.M., 1994b. Surface-area control on organic-carbon accumulation in continental-shelf sediments. Geochimica Et Cosmochimica Acta, 58(4): 1271-1284. Mayer, L.M., 2004. The inertness of being organic. Marine Chemistry, 92(1-4): 135-140. Meyers, P.A., 1994. Preservation of elemental and isotopic source identification of sedimentary organic matter. Chemical Geology, 114(3-4): 289-302. Nürnberg, D. et al., 1994. Sediments in arctic sea-ice - implications for entrainment, transport and release. Marine Geology, 119(3-4): 185-214. Osterkamp, T.E. and Romanovsky, V.E., 1999. Evidence for warming and thawing of discontinuous permafrost in Alaska. Permafrost and Periglacial Processes, 10(1): 17-37. Otto, A. and Simpson, M.J., 2006. Evaluation of CuO oxidation parameters for determining the source and stage of lignin degradation in soil. Biogeochemistry, 80(2): 121-142. Peterson, B.J. et al., 2002. Increasing river discharge to the Arctic Ocean. Science, 298(5601): 2171-2173. Rachold, V. et al., 2004. Modern terrigenous organic carbon input to the Arctic Ocean. In: R.a.M. Stein, R. W. (Editor), The Organic Carbon Cycle in the Arctic Ocean. Springer, New York, pp. pp. 33–56. Reimnitz, E., McCormick, M., Bischof, J. and Darby, D.A., 1998. Comparing sea-ice sediment load with Beaufort Sea shelf deposits: Is entrainment selective? Journal of Sedimentary Research, 68(5): 777-787. Rieley, G., Collier, R.J., Jones, D.M. and Eglinton, G., 1991. The biogeochemistry of Ellesmere lake, UK. 1. Source correlation of leaf wax inputs to the sedimentary lipid record. Organic Geochemistry, 17(6): 901-912. Riley, R.G. and Kolattukudy, P.E., 1975. Evidence for covalently attached para-coumaric acid and ferulic acid in cutins and suberins. Plant Physiology, 56(5): 650-654. Rothman, D.H. and Forney, D.C., 2007. Physical model for the decay and preservation of marine organic carbon. Science, 316(5829): 1325-1328. Sanchez-Garcia, L. et al., 2014. Characterisation of Three Regimes of Collapsing Arctic Ice 39 Complex Deposits on the SE Laptev Sea Coast using Biomarkers and Dual Carbon Isotopes. Permafrost and Periglacial Processes, 25(3): 172-183. Sarkanen, K.V. and Ludwig, C.H., 1971. Lignins: occurrence, formation, structure and reactions. Wiley-Interscience. Schaefer, K., Zhang, T.J., Bruhwiler, L. and Barrett, A.P., 2011. Amount and timing of permafrost carbon release in response to climate warming. Tellus Series B-Chemical and Physical Meteorology, 63(2): 165-180. Schirrmeister, L. et al., 2002. Paleoenvironmental and paleoclimatic records from permafrost deposits in the Arctic region of Northern Siberia. Quaternary International, 89: 97-118. Schubert, C.J. and Calvert, S.E., 2001. Nitrogen and carbon isotopic composition of marine and terrestrial organic matter in Arctic Ocean sediments: implications for nutrient utilization and organic matter composition. Deep-Sea Research Part I-Oceanographic Research Papers, 48(3): 789-810. Schubert, C.J. and Stein, R., 1997. Lipid distribution in surface sediments from the eastern central Arctic Ocean. Marine Geology, 138(1-2): 11-25. Semiletov, I. et al., 2005. The East Siberian Sea as a transition zone between Pacific-derived waters and Arctic shelf waters. Geophysical Research Letters, 32(10). Semiletov, I.P., and Gustafsson, Ö. 2009. East Siberian Shelf Study Alleviates Scarcity of Observations. Eos, Transactions American Geophysical Union, 90(17): 145-146. Serreze, M.C. et al., 2003. Large-scale hydro-climatology of the terrestrial Arctic drainage system. Journal of Geophysical Research-Atmospheres, 108(D2). Simoneit, B.R.T., 1977. Black Sea, a sink for terrigenous lipids. Deep-Sea Research, 24(9): 813830. Smith, L.C., Pavelsky, T.M., MacDonald, G.M., Shiklomanov, A.I. and Lammers, R.B., 2007. Rising minimum daily flows in northern Eurasian rivers: A growing influence of groundwater in the high-latitude hydrologic cycle. Journal of Geophysical ResearchBiogeosciences, 112(G4). Spencer, R.G.M., Aiken, G.R., Wickland, K.P., Striegl, R.G. and Hernes, P.J., 2008. Seasonal and spatial variability in dissolved organic matter quantity and composition from the Yukon River basin, Alaska. Global Biogeochemical Cycles, 22(4). Stark, R.E., Zlotnikmazori, T., Ferrantello, L.M. and Garbow, J.R., 1989. Molecular-structure and dynamics of intact plant polyesters – solid state NMR studies. Acs Symposium Series, 399: 214-229. Steele, M. and Ermold, W., 2004. Salinity trends on the Siberian shelves. Geophysical Research Letters, 31(24). Stein, R. and Macdonald, 2004. The Organic Carbon Cycle in the Arctic Ocean. Springer, Berlin, Germany. Stocker, T.F. et al., 2013. IPCC, 2013: climate change 2013: the physical science basis. Contribution of working group I to the fifth assessment report of the intergovernmental panel on climate change. Stroeve, J.C., Markus, T., Boisvert, L., Miller, J. and Barrett, A., 2014. Changes in Arctic melt season and implications for sea ice loss. Geophysical Research Letters, 41(4): 1216-1225. Tarnocai, C. et al., 2009. Soil organic carbon pools in the northern circumpolar permafrost region. Global Biogeochemical Cycles, 23. Tesi, T. et al., 2014. Composition and fate of terrigenous organic matter along the Arctic landocean continuum in East Siberia: insights from biomarkers and carbon isotopes. 40 Geochimica et Cosmochimica Acta (133) 235-256. Tingley, M.P. and Huybers, P., 2013. Recent temperature extremes at high northern latitudes unprecedented in the past 600 years. Nature, 496(7444): 201-204. van Dongen, B.E., Semiletov, I., Weijers, J.W.H. and Gustafsson, O.R., 2008. Contrasting lipid biomarker composition of terrestrial organic matter exported from across the Eurasian Arctic by the five great Russian Arctic rivers. Global Biogeochemical Cycles, 22(1). Viscosi-Shirley, C., Pisias, N. and Mammone, K., 2003. Sediment source strength, transport pathways and accumulation patterns on the Siberian-Arctic's Chukchi and Laptev shelves. Continental Shelf Research, 23(11-13): 1201-1225. von Deimling, T.S. et al., 2012. Estimating the near-surface permafrost-carbon feedback on global warming. Biogeosciences, 9(2): 649-665. Vonk, J.E. and Gustafsson, O., 2013. Permafrost-carbon complexities. Nature Geoscience, 6(9): 675-676. Vonk, J.E. et al., 2012. Activation of old carbon by erosion of coastal and subsea permafrost in Arctic Siberia. Nature, 489(7414): 137-140. Wakeham, S.G., Lee, C., Farrington, J.W. and Gagosian, R.B., 1984. Biogeochemistry of particulate organic matter in the oceans – results from sediment trap experiments. DeepSea Research Part a-Oceanographic Research Papers, 31(5): 509-528. Williams, C.J., Yavitt, J.B., Wieder, R.K. and Cleavitt, N.L., 1998. Cupric oxide oxidation products of northern peat and peat-forming plants. Canadian Journal of Botany-Revue Canadienne De Botanique, 76(1): 51-62. Zimov, S.A. et al., 2006. Permafrost carbon: Stock and decomposability of a globally significant carbon pool. Geophysical Research Letters, 33(20). Bolshoy Lyakhovskiy Island, photo Alexander Charkin 41