Accepted Manuscript

Compositional and biogeochemical variations of sediments across the terrestri-al-marine continuum of the Baker-Martínez fjord system (Chile, 48°S)

Lorena Rebolledo, Sebastien Bertrand, Carina B. Lange, Fabian J. Tapia,Eduardo Quiroga, Matthias Troch, Nelson Silva, Paola Cárdenas, Silvio Pantoja

PII: S0079-6611(18)30048-XDOI: https://doi.org/10.1016/j.pocean.2018.12.004Reference: PROOCE 2045

To appear in: Progress in Oceanography

Received Date: 23 March 2018Revised Date: 30 November 2018Accepted Date: 10 December 2018

Please cite this article as: Rebolledo, L., Bertrand, S., Lange, C.B., Tapia, F.J., Quiroga, E., Troch, M., Silva, N.,Cárdenas, P., Pantoja, S., Compositional and biogeochemical variations of sediments across the terrestrial-marinecontinuum of the Baker-Martínez fjord system (Chile, 48°S), Progress in Oceanography (2018), doi: https://doi.org/10.1016/j.pocean.2018.12.004

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customerswe are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, andreview of the resulting proof before it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

PiO special issue “Integrative Research in High-Latitude Austral Systems: Linking Ecosystems and Society “

Guest Managing Editor: José Luis Iriarte

Compositional and biogeochemical variations of sediments across the terrestrial-marine

continuum of the Baker-Martínez fjord system (Chile, 48°S)

Lorena Rebolledoa,b,1, Sebastien Bertrandc,1, Carina B. Langeb,d,e,1, Fabian J. Tapiad,e, Eduardo

Quirogae,f, Matthias Trochc, Nelson Silvaf, Paola Cárdenasb,e,g, Silvio Pantojad,e

a Instituto Antártico Chileno, INACH, Plaza Muñoz Gamero 1055, Punta Arenas, Chile

b Centro de Investigación Dinámica de Ecosistemas Marinos de Altas Latitudes (IDEAL), Universidad

Austral de Chile, Campus Isla Teja, Valdivia, Chile

c Renard Center of Marine Geology, Ghent University, Krijgslaan 281 S8, 9000, Gent, Belgium

d Departamento de Oceanografía, Universidad de Concepción, Barrio Universitario, Concepción,

Chile

e Centro de Investigación Oceanográfica COPAS Sur-Austral, Universidad de Concepción, Barrio

Universitario, Concepción, Chile

f Pontificia Universidad Católica de Valparaíso, Escuela de Ciencias del Mar, Avenida Altamirano

1480, Valparaíso, Chile

g Postgrado en Oceanografía, Facultad de Ciencias Naturales y Oceanográficas, Universidad de

Concepción, Barrio Universitario, Concepción, Chile

1 These authors contributed equally to this paper.

Corresponding author: sebastien.bertrand@ugent.be

2

ABSTRACT

The Baker-Martínez fjord system (Chile, 48⁰S) is a transitional environment between the terrestrial

ecosystems of Patagonia and the SE Pacific Ocean. This unique setting makes it an ideal system to

evaluate land-ocean gradients in sediment composition and in a range of biogeochemical

variables. Here, we studied the composition and organic geochemistry of surface sediments

deposited in the Baker-Martínez fjord system to assess spatial changes in sediment properties and

identify the best tracers of terrestrial input. We determined concentrations in biogenic opal,

organic carbon, calcium carbonate, and lithogenic particles, and we measured diatom abundance,

bulk elemental and isotopic organic geochemistry, and n-alkanes distribution, on a series of

surface sediment samples distributed along two proximal-to-distal transects (one in the Martínez

Channel and a second one in the Baker Channel). Results reveal clear proximal-to-distal trends in

the proportion of freshwater diatoms and in the fraction of organic carbon of terrestrial origin

calculated from the bulk organic δ13C measurements. Diatoms are particularly sensitive to

freshwater input in proximal environments, whereas bulk organic geochemistry better reflects

terrestrial input over long distances across the fjord system. Other terrestrial proxies frequently

used in the literature, such as the N/C ratio and the n-alkane-based terrestrial/aquatic ratio, did

not show clear variations with distance from the fjord head. Our results suggest that the isotopic

composition of bulk organic matter (δ13Corg) and the relative proportion of diatoms of freshwater

origin are the best-suited proxies to estimate past changes in terrestrial input from fjord sediment

archives, irrespective of the glacial status of the fjord and its watershed.

KEYWORDS: Chilean fjords, sediment composition, organic geochemistry, carbon stable isotopes,

n-alkanes, diatoms

3

1. Introduction

The Chilean fjords (41–55°S) constitute a dynamic estuarine system characterized by sharp

gradients in salinity, aquatic productivity, and suspended sediment concentrations (Aracena et al.,

2011; Sepúlveda et al., 2011; Silva et al. 2011; Vargas et al., 2011). Due to their location at the

interface between marine and terrestrial environments, they rapidly respond to changes in

continental and marine conditions alike (Sepúlveda et al., 2011; Bertrand et al., 2012b).

High freshwater and detrital inputs from Patagonian rivers and calving glaciers result in a

water column that is typically highly stratified with variable amounts of suspended solids (Silva et

al., 1998; González et al., 2013; Marín et al., 2013; Quiroga et al., 2016). During the productive

season, the fjords can also efficiently export carbon to the sediments, and therefore act as an

efficient CO2 sink (Sepúlveda et al., 2011; Torres et al., 2014; Smith et al., 2015). As a result,

Patagonian fjord sediments are increasingly used to reconstruct past changes in river discharge

(Bertrand et al., 2014; Rebolledo et al., 2015), glacier variability (Kilian and Lamy, 2012; Bertrand

et al., 2014, 2017), coastal ocean temperature (Sepúlveda et al., 2009; Caniupán et al., 2014), and

marine productivity (Rebolledo et al., 2008; Mayr et al., 2014; Aracena et al., 2015).

In recent years, the Baker-Martínez fjord system (48⁰S) has been the focus of several studies,

encompassing zooplankton community dynamics and along-fjord changes in estuarine circulation

(Meerhoff et al., 2013, 2014, 2015, 2018); internal tides related to river discharge pulses (Ross et

al., 2014); productivity changes in the water column (González et al., 2013); and benthic diversity

patterns and trophic structure (Quiroga et al., 2012, 2016). Seasonal changes in river outflow, for

example, have been shown to force the along-fjord displacement of salinity and turbidity fronts,

influencing plankton and benthic patterns (Meerhoff et al., 2014; Quiroga et al., 2016). Likewise,

the warming of subsurface waters along the Martínez Channel in summer and autumn has been

linked to intrusions of warmer water from the Gulf of Penas (Aiken, 2012; Moffat et al., 2018).

4

Semidiurnal internal tides in the Martínez Channel near the pycnocline were also shown to

correlate with high discharge pulses of the Baker River (Ross et al., 2014). In addition, turbidity in

Baker Fjord is increasingly affected by Glacial Lake Outburst Floods (GLOFs; Dussaillant et al., 2010;

Jacquet et al., 2017), which supply high volumes of freshwater with high suspended loads (Marín

et al., 2013). So far, very few studies have focused on the spatial patterns in the composition of

the sediments deposited along the Baker-Martínez fjord system.

In other Patagonian fjords, sediment carbon stable isotopes have provided reliable

information about the sources of organic matter (e.g., Sepúlveda et al., 2011; Silva et al., 2011;

Vargas et al., 2011; Lafon et al., 2014; Mayr et al., 2014), and long-chain n-alkanes have been used

as proxies for the provenance of terrigenous inputs (Pinto and Bonert, 2005; Sepúlveda et al.,

2011; Placencia et al., 2017). Additionally, biogenic opal and marine diatoms preserved in fjord

sediments have proven to be reliable proxies of changes in siliceous phytoplankton export

production (e.g., Rebolledo et al., 2005, 2015; Sepúlveda et al., 2005; Aracena et al., 2011), and

increases in the contents of freshwater diatoms in sediments have been related to periods of high

precipitation and river discharge (Rebolledo et al., 2015).

With this in mind, the main goal of this study was to evaluate the spatial distribution of

sedimentological and biogeochemical proxies, as well as diatom abundance and species

composition, of surface sediments from the Baker-Martinez fjord system to test their application

for paleoclimate and paleoenvironmental reconstructions in a region that appears to be

particularly sensitive to climate change. Since the Baker-Martinez fjord system is one of the few

Patagonian systems directly open to the Pacific Ocean, it has the advantage of covering the entire

continuum of terrestrial to marine conditions, making our results applicable to investigate

sediments deposited under different environmental conditions in the past.

5

2. Study area

The Baker-Martinez fjord system is a ~150 km long, W-E oriented basin that connects the

glaciated Patagonian Andes to the east with the SE Pacific Ocean to the west through the Gulf of

Penas (Fig. 1). It is located between the Northern and Southern Patagonian Icefields (NPI and SPI)

and it comprises two main channels: Martínez to the north and Baker to the south (Fig. 1). The

fjord system has a complex bathymetry with shallow depths (<50 m) at the head near the town of

Tortel and deep basins of more than 1000 m in the middle of Baker Channel (Fig. 2, top panel).

This system formed during the last deglaciation when the ice masses extended to the Pacific

Ocean and it is thought to have reached its present-day configuration after the separation of the

Patagonian Ice Sheet into the NPI and SPI approximately 8000 years ago (Glasser et al., 2016).

The Baker-Martinez fjord system is almost entirely fed by two large rivers: Baker (mean

discharge ~1100 m3/s) and Pascua (~750 m3/s), which originate from the large lakes General

Carrera and O’Higgins, respectively (Fig. 1). The discharge of Baker and Pascua rivers is highest in

austral summer (December–March), i.e., during the snowmelt season, and lowest in winter. This

hydrological regime may be defined as nival with glacial contributions (e.g., Lara et al., 2015).

Additionally, the channels receive freshwater from Steffen Fjord, which is fed by meltwaters from

Steffen Glacier (the third largest glacier of the NPI; Willis et al., 2012) via Huemules River, and

from the calving Jorge Montt Glacier (the northernmost SPI glacier) (Fig. 1). The latter has

experienced a rapid retreat during the last decade, with a rate of almost a kilometer per year

(Rivera et al., 2012).

Circulation studies based on ADCP deployments and CTD measurements have shown that the

circulation in the area corresponds to a typical two-layer estuarine system, with maximum surface

velocities that can reach up to 20 cm s-1 in summer (Cáceres, 2007; Ross et al., 2014, 2015). Tides

are mixed, mainly semidiurnal with amplitudes of 1 m (Ross et al., 2015), and wind velocities are

6

controlled year-round by the strength of the southern westerly winds (SWW). Modeling studies

have shown that regional wind forcing is a key factor explaining the circulation and hydrography of

the Gulf of Penas and Baker Channel (Aiken, 2012), including the modulation of warm water

intrusions into proglacial fjord Jorge Montt (Moffat, 2014; Moffat et al., 2018).

Although data on primary productivity in this region are scarce, they reveal low values of 91

mg C m-2 d-1 near Caleta Tortel (Aracena et al., 2011) and 650 mg C m-2 d-1 in Steffen Fjord (Jacob et

al., 2014). Pico- and nano-plankton dominates at inner sites, whereas a dominance of chain-

forming diatom genera (i.e., Thalassiosira, Chaetoceros, Skeletonema) is observed in outer, more

oceanic sections of the fjord. During spring, high loads of suspended sediments (42−105 mg/L in

the Martínez Channel; Marín et al., 2013) dramatically increase the surface attenuation of incident

light, which may limit primary productivity (e.g., González et al., 2013; Quiroga et al., 2016).

The region’s climate is maritime hyperhumid, with highest precipitation between May and

August (late fall and winter), and a mean air temperature of 8.9°C (Hijmans et al., 2005).

Vegetation cover in the Baker and Pascua river basins consists of evergreen and deciduous native

forests, shrublands, and peatlands (Luebert and Pliscoff, 2006; Rodríguez et al., 2008). Barren land

is located above the treeline. All prominent species use the C3 photosynthetic pathway (Powell

and Still, 2009). In the vicinity of Caleta Tortel, located next to the Baker River mouth, the

vegetation is dominated by Nothofagus, shrubs, and ferns. Anthropogenic disturbances consist of

logging by rural landowners, overgrazing by livestock, and road construction, which are promoting

biological invasions in the forests of the Baker River basin (Vidal et al., 2011).

3. Material and methods

3.1 Hydrographic data collection

7

Sixteen vertical CTD profiles were acquired in the Baker-Martinez fjord system (Fig. 2) during

the COPAS Sur-Austral 2014 campaign (October 2014), using a Seabird SBE model 25 CTD with a

sampling frequency of 8 Hz and average descent rate of 1 m s-1. The CTD profiles included

temperature, salinity, oxygen concentration, turbidity and fluorescence, down to a maximum of

447 m (station 14; Fig. 2). The same stations were revisited in October 2015 (spring), and January

and March 2016 (summer) to capture temporal variability in the spatial structure of hydrographic

variables during the productive season. The October 2015 survey was conducted using a different

CTD (RBR XR-620), with a sampling rate of 6 Hz. During post-processing, all CTD data were depth-

binned to obtain profiles with a vertical resolution of 1 m (i.e. 1 dbar).

To characterize temporal variability in salinity along the Martínez Channel, we used a set of

sub-surface salinity records gathered over one year (October 2012 through November 2013) at 5

points along the Martinez Channel. At each station, conductivity-temperature loggers (HOBO U24,

Onset Computer, USA) were deployed at depths that ranged between 1 and 2 m below mean

lower low water (MLLW). Salinity time series were obtained from conductivity data based on

calibration factors provided by the manufacturer. For each time step in the salinity record along

the Martinez Channel, linear interpolation was used to determine the along-fjord position of the

10 and 20 salinity contours. These values were later used to characterize the seasonal variability in

the position and strength of the surface salinity front.

3.2 Sampling

During the October 2014 campaign, water samples were collected at 11 stations with Niskin

bottles to measure chlorophyll-a (Chl-a) concentrations down to 30, 40, or 50 m depth, depending

on location (Table S1). Duplicate samples were frozen at -20° C prior to analysis by fluorometry

(Parsons et al., 1984). Chl-a values were integrated over the upper 30 m of the water column.

8

Vertical net tows (25 µm mesh) were used to collect phytoplankton samples from 0–15 m for

qualitative analysis at stations CO14-E3, E4, E5, E6, E8, and E13 (Fig. 1; Table S1). Observation of

phytoplankton samples was done at 400X magnification using a Zeiss Axio Scope A1 microscope

with phase contrast.

Surface sediment samples were collected during several campaigns carried out in the Baker-

Martinez fjord system and in Jorge Montt Fjord between 2008 and 2014: CIMAR 14 Fiordos

(October-November 2008; CF14), Jorge Montt 2013 (March and November 2013; Gu13), Martínez

Channel 2014 (March and August 2014; Qu14; Quiroga et al., 2016), COPAS Sur-Austral 2014

(October 2014; CO14), and CIMAR 20 Fiordos (November 2014; CF20) (Fig. 1; Table S2). A total of

53 surface sediment samples were obtained using box and Rumhor corers (0–1 cm) and grab

samplers. For each sampling location, we measured the distance from the nearest river or calving

glacier front using Google Earth (paths running through the center of the fjords). The spatial

distribution of the sedimentological and geochemical variables measured in the surface sediments

(Table S2) was visualized on maps made using Ocean Data View 4.7.10 (Schlitzer, 2017) and Global

Mapper 16.

In addition, living vegetation was collected near Caleta Tortel during the COPAS Sur-Austral

2014 campaign with the main goal of defining the terrestrial end-member of geochemical

parameters to be compared with the signal preserved in the surface sediments. This terrestrial

end-member was additionally constrained by a bulk river sediment sample (RS11-05) collected 8

km upstream of the outflow of Baker River.

3.3 Sediment composition

Fjord sediments are generally composed of variable mixtures of lithogenic particles, organic

matter, biogenic opal, and carbonates (Bertrand et al., 2012b). The relative contents of these four

9

sediment components were quantified according to the following methods.

Biogenic opal was estimated following the alkaline extraction procedure described by

Mortlock and Froelich (1989) and modified by Müller and Schneider (1993), which consists in

extracting 50 mg of sediment with NaOH at 85°C for 5 hours. Extraction and analysis by

molybdate-blue spectrophotometry were conducted at the laboratories of Marine Organic

Geochemistry and Paleoceanography, University of Concepción, Chile. Values are expressed as

biogenic opal after multiplication of the Si (%) values by 2.4 (Mortlock and Froelich, 1989).

Analytical precision was ± 0.5%.

The weight percentage of total inorganic carbon (TIC) was determined using an UIC CM5017

coulometer equipped with a CM5330 acidification module at Ghent University, Belgium. For each

sample, approximately 50 mg of sediment was precisely weighed into a glass flask and treated

with 5 ml 2N H3PO4 to release CO2. This method assumes that 100% of the measured CO2 is

derived from the dissolution of calcium carbonate. Calcium carbonate concentrations were

calculated as CaCO3 = TIC × 8.33. The detection limit was 0.03% CaCO3.

The bulk organic geochemistry of the sediments was measured in two different laboratories:

Bioanalytical laboratory, School of Biological Sciences, Washington State University, USA (most

samples), and Laboratory of Biogeochemistry and Applied Stable Isotopes (LABASI-PUC), Chile (the

11 samples from the CF14 cruise, and samples Qu14-E1Q and Qu14-E4Q). To determine organic

carbon (Corg) and total nitrogen (Ntot) contents, freeze-dried sediments were ground and acidified

(with 2N phosphoric acid at Bioanalytical laboratory and with HCl 2:1 v/v at LABASI) to remove

inorganic carbon. Corg and Ntot measurements were made on a Eurovector (Milan, Italy) elemental

analyzer at Bioanalytical laboratory, and on a Flash EA2000 elemental analyser at LABASI. Organic

matter (OM) was calculated from the Corg results as OM = Corg × 2.2 (Bertrand et al., 2012b).

10

The content of lithogenic particles (LP) was calculated as: LP = 100 % - CaCO3 - 2.2 × Corg -

biogenic opal, following Bertrand et al. (2012b).

3.4 Bulk organic geochemistry

Carbon (δ13Corg) and nitrogen (δ15N) stable isotopes were measured at the two laboratories

mentioned above. Most samples were measured at Bioanalytical laboratory using an elemental

analyzer coupled with an Isoprime isotope ratio mass spectrometer (IRMS), with a precision

±0.1‰ for δ13Corg and 0.3‰ for δ 15N. The running standard was a protein hydrolysate calibrated

against NIST standards. At LABASI, a Thermo Delta Advantage IRMS coupled with the Flash EA

2000 elemental analyser was used. Acetanilide, atropine, caffeine, glutamic acid, and high organic

sediment were used as internal standards. δ13Corg and δ15N were determined with a precision of

±0.18 and 0.2 ‰, respectively.

The vegetation samples collected near Caleta Tortel were analysed for Corg and Ntot

concentrations and stable isotopic composition at Bioanalytical laboratory, USA (Table S3). Prior to

analysis, samples were washed with distilled water, freeze-dried, and ground. About 1 mg of plant

tissue was put into a tin capsule and acidified with 2N phosphoric acid before analysis.

The <90 µm fraction of river sediment sample RS11-05 was isolated by dry sieving, and its bulk

organic geochemical composition was analyzed at the UCDavis Stable Isotope Facility after

acidification of the sample with 1N sulfurous acid. The precision, calculated by replicate analysis of

an internal standard, was 0.05‰ for δ13Corg and 0.08‰ for δ15N.

We used the δ13Corg of the sediment samples to estimate the relative contributions of

terrestrial and aquatic organic carbon (Meyers, 1997; Perdue and Koprivnjak, 2007; Sepúlveda et

al., 2011). The fraction of terrestrial Corg (fT) was calculated using a two end-member (linear)

mixing model (Thornton and McManus, 1994):

11

–

–

(1)

where δ13Corg sample, δ13Corg marine, and δ

13Corg terr are the isotopic values of the sample, terrestrial

end-member, and marine end-member, respectively. The end-member values used in this study

were determined on samples collected in fully marine and fully terrestrial conditions: -19.86‰ for

the marine end-member (site GeoB3323-4 on the continental margin off the coast of Northern

Chilean Patagonia; Hebbeln et al., 2000) and -27.72 ‰ for the terrestrial end-member (average of

five <125 µm sediment samples from rivers that discharge into the fjords; Bertrand et al., 2012b).

The latter value is supported by the δ13C composition of sediment sample RS11-05 <90 µm from

Baker River (-26.74‰). The proportion of marine organic carbon (ƒM) was calculated by difference

(100 % - ƒT). For three samples, the application of Eq. (1) provided ƒT values above 100, which were

then modeled to 100.

3.5 n-alkanes

Extraction of lipid biomarkers from sediments was carried out according to a modified Bligh

and Dyer (1959) protocol. Briefly, 0.3–2 g of freeze-dried samples (surface sediments and plant

tissues) were extracted by accelerated solvent extraction (Dionex ASE350) at high pressure (1500

psi) and high temperature (100 °C) with a mixture of dichloromethane and methanol (DCM:MeOH,

3:1 v/v). Column chromatography was applied to the total lipid extract using ca. 1g of silica gel as

stationary phase and successively 3/8 DV (Dead Volume) Hexane, 2 DV (DCM), 2DV DCM/EtoAC

(1:1) and 2DV DCM/MeOH (7:3) to elute the fractions containing hydrocarbons (F1), esters and

ketones (F2), sterols and alcohols (F3) and fatty acids (F4).

Samples for n-alkanes were analyzed on an Agilent 6890 GC series coupled to an Agilent 5972

mass spectrometer (MS), at the laboratory of Marine Organic Geochemistry, University of

12

Concepción, Chile. The chromatograph was equipped with a fused silica capillary column (30 m

length, i.d. = 0.25 mm HP5-MS, film thickness = 0.25 µm) and an automated injection system. The

GC oven temperature program was 80 °C to 130 °C at 20 °C min-1, and then to 310 °C at 4 °C min-1.

The MS was operated in electron impact mode (70 eV).

Long chain n-alkanes are important components of leaf waxes (Bianchi and Canuel, 2011). In

higher plants, n-alkanes show a predominance of molecules containing between 27 and 35 carbon

atoms, with a prevalence of odd- over even-numbered carbon chains (Eglinton and Hamilton,

1967). Aquatic sources such as algae and photosynthetic bacteria are dominated by C15, C17, and

C19. The shorter chain n-alkanes (C14−C25) are therefore mainly of algal phytoplankton origin (e.g.,

Eglinton and Hamilton 1967; Liu and Liu, 2016).

Three n-alkane-based indices were calculated as follows:

(1) The average chain length (ACL), i.e., the weight-averaged number of carbon atoms of

the higher plant C23-35 n-alkane, was calculated following Poynter et al. (1989):

ACL23-35 =

(2)

(2) The Carbon Preference Index (CPI) was calculated following Allan and Douglas (1977):

CPI =

(3)

(3) The terrigenous/aquatic ratio (TAR), which provides an estimate of the algal vs.

terrestrial organic matter, was calculated following Bourbonniere and Meyers (1996):

TAR =

(4)

ACL mostly reflects vegetation type (El Nemr et al., 2016), but it has also been shown to reflect

temperature, humidity, carbon assimilation pathway (C3 vs. C4), and the woody vs. non-woody

13

nature of the plants (Bush and McInerney, 2015). In paleo-studies, ACL is generally used as a

climate proxy, based on the observation that, on average, plants in forests form shorter chains

than grassland leaves (Cranwell, 1973). Likewise, in warmer climates, plants tend to produce

longer chains of n-alkanes (Jeng, 2006).

CPI and TAR have been used to estimate the source of OM in sediments. CPI values >5 are

indicative of n-alkanes derived from terrestrial sources, i.e., land-plants, which have a

predominance of odd-numbered carbon chains (Eglinton and Hamilton, 1963; Hedges and Prahl,

1993), whereas values 1.0–1.2 and < 1.0 reflect marine plankton (Corner, 1979) and petrogenic

inputs (Pendoley, 1992), respectively. However, it is worth noting that marine plankton does not

affect bulk CPI values much since plankton mostly produces short-chain n-alkanes (Rosell-Melé

and McClymont, 2007). In addition, CPI may reflect the state of degradation of organic matter

because the odd-even preference tends to decrease with diagenesis (Bianchi and Canuel, 2011).

TAR values reflect the ratio of vascular land plants over aquatic sources (Bourbonniere and

Meyers, 1996; Meyers, 1997). The absolute values may however be affected by selective

biodegradation (Peters et al., 2007).

3.6 Siliceous microorganisms

For siliceous microorganisms (mainly diatoms but also including sponge spicules,

silicoflagellates, and chrysophyte cysts), 0.5 g of freeze-dried sediment of each sample was

oxidized using a mix of 30% hydrogen peroxide, sodium pyrophosphate, and hydrochloric acid to

remove organic matter, clays, and carbonates, respectively. 0.2 ml of clean sample was deposited

onto a cover slip, air-dried, and mounted with Naphrax mounting medium (refraction index = 1.7).

Two permanent slides per sample were prepared in this fashion and observed under a Zeiss Axio

Scope A1 microscope with phase contrast at 400× and 1000×. Siliceous microorganisms were

14

counted on transects on microslides covering 1/2 or 1/5 of the slide, following counting

procedures of Schrader and Gersonde (1978). Concentrations are expressed as numbers per gram

dry weight (i.e., diatom valves g-1). Diatoms were determined to the lowest taxonomic level

following Cupp (1943), Rivera (1981), Hasle and Syvrtsen (1996), Sims (1996), and Witkowski et al.

(2000). Diatoms were grouped by ecological affinity into the following categories (based on Cupp

(1943), Round et al. (1990), Sims (1996), Witkowski et al. (2000), and Romero and Hebbeln (2003);

Table S4): high nutrient (HN), coastal planktonic (CP), warm water (WW), cold water (CW), marine

benthic (MB), and freshwater (FW). Laboratory and microscopy work was carried out at the

Paleoceanography laboratory, University of Concepción, Chile.

3.7 Statistical analysis

To identify the combination of variables that are most influential in terms of hydrographic

structure along the fjord, Principal Component Analysis (PCA) was applied to 0–20 m depth

averages of the four main variables (temperature, salinity, turbidity, and fluorescence) measured

during the CTD surveys of October 2014, October 2015, January 2016, and March 2016. A second

PCA (Spearman; XLStat 2016.01) was performed on the following 14 sediment variables: distance

from the nearest river mouth/calving glacier front, biogenic opal, Corg, CaCO3, lithogenic particles

(LP), N/C molar ratio, δ13Corg, δ15 N, total diatoms, total n-alkanes, ACL, CPI, TAR, and the

proportion of freshwater diatoms. Spearman PCA is preferred over Pearson PCA because

Spearman evaluates the monotonic, instead of linear, relationship between variables. The same

software was used to calculate Pearson correlation coefficients and p-values (two-tailed test of

significance) between variables. Correlations at p < 0.01 were considered significant.

4. Results

15

4.1 Water column

Surface water temperature in the Baker-Martinez fjord system in October 2014 ranged from

3.9 to 10.4°C, with a progressive warming towards the outer section of the channels (Fig. 2). The

warmest waters were found at ~50 m depth. The lowest salinity values (< 3 PSU) were observed

within the upper water column (0–5 m) at inner-fjord stations of the Baker-Martinez fjord system,

i.e., close to the mouths of Baker and Pascua rivers (Fig. 2). This low-salinity layer became thinner

towards the west, where it reached less than 3 m. Overall, salinity fluctuated widely between 1

and 33.4 PSU, with the highest values occurring in the deepest waters. Dissolved oxygen ranged

from 8.2 mL/L in surface waters to 3.2 mL/L at depth. No anoxic/hypoxic waters were detected

during the survey. Turbidity was higher in the upper water column, with the highest values

occurring in the colder surface water, especially near river mouths and calving glaciers. Turbidity

was lower at the outer fjord stations, where it fluctuated between 1.0 and 24.6 NTU. Surface

turbidity was 10x higher at inner and glacier sites than at more oceanic sites (Fig. 2). Fluorescence,

which is often used as a proxy for phytoplankton biomass, was essentially limited to the upper 20–

25 m of the water column, and showed a clear increase towards the outer part of the fjord system,

especially in Baker Channel (Fig. 2).

The PCA of the CTD data (Fig. 3) showed high positive covariance between fluorescence and

salinity, which covaried negatively with turbidity. Our analysis also indicated that fluorescence and

temperature were practically independent.

Concentrations of Chl-a integrated over the upper 30 m of the water column ranged between

8.3–92.5 mg Chl-a m-2 in the Martínez Channel and 11.5‒114.3 mg Chl-a m-2 in the Baker Channel

(Fig. 4a; Table S1). In both channels, the lowest values occurred in the most proximal

environments.

16

Twenty-five diatom taxa were recorded in the net tows of the six stations that were sampled.

Among them, chains of the genera Chaetoceros, Pseudo-nitzschia, Skeletonema, and Cerataulina

pelagica, and colonies of Thalassionema nitzschioides were commonly observed. Inner fjord

station E13 was the least diverse in terms of the numbers of diatom taxa (only 7 taxa were

observed; Table S5).

4.2 Bulk sediment composition

Throughout the Baker-Martinez fjord system, organic matter contents varied between 0.35

and 9.09 wt. %, biogenic opal contents ranged from 1.1 to 10.2 wt. %, and CaCO3 concentrations

were always below 0.3 wt. %, except at some outer fjord locations where they reached up to 2.4

wt. % (Fig. 5a-c; Table S2). The highest contents in organic matter, biogenic opal and CaCO3 all

occurred toward the outer section of the fjord system. In contrast, the lithogenic particle

concentrations were high along the entire fjord system (84.8–97.2 wt. %), with peak values

occurring near the Baker River mouth and Steffen Fjord (Fig. 5d). Station CO14-E6 in the outer

section of Martinez Channel (Fig. 1) contained high amounts of mollusk fragments, which explains

the particularly high CaCO3 concentrations (16.9%; Table S2). This station was therefore

considered as an outlier.

4.3 Bulk organic geochemistry

Corg is generally low along Martínez and Baker channels (<0.75%), with the lowest value

(0.16%) being recorded at station Gu13-E2, near Jorge Montt Glacier. Higher values (>1%) were

only reached towards the marine end of the fjord system (Fig. 6a). An exceptionally high value of

4.4% was observed at shallow station Qu14-E4Q (Fig. 6a). In general, the N/C values fell within the

range 0.08–0.20, with lower values at inner-fjord sites and near river mouths than towards the

17

open coast (Fig. 6b). Significantly higher N/C values (average: 0.2–0.5) were however observed in

glacial fjord Jorge Montt sediments, where samples have extremely low Corg contents (< 0.3 %;

Table S2; Figs. 6a, b).

Sedimentary stable isotopes of organic carbon (δ13Corg) showed depleted values (-25 to -29‰)

at mid- and inner fjord locations (e.g., heads of Baker and Martinez channels, and proglacial fjord

Jorge Montt) compared to sediments toward the outer fjords and in the Gulf of Penas (-21 to -

23‰, Fig. 6c). The proportion of terrestrial organic carbon varied between 15 and 100% and

obviously followed the same pattern as δ13 Corg (Fig. 6d).

4.4 n-alkanes

We compared the n-alkane composition of sediment samples from the Baker-Martinez fjord

system with living vegetation recovered near Caleta Tortel (Fig. 7, Table S3). n-alkanes of chain

length between C16 and C37 were recorded from trees, shrubs, and fern leaves (n=8) with total

concentrations varying from 1246 to 6073 µg g dw-1 (Figs. 7a and S1; Table S3). A marked

dominance of odd- over even-numbered n-alkanes was observed (Figs. 7a and S1). In sediments,

total n-alkane concentrations varied from 0.9 to 50.5 µg g dw-1 (Fig. 7b, Table S2) with chain

lengths between n-C14 and n-C38 (n=37) and a prevalence of odd- over even-numbered carbon

chains (Fig. 7b).

When comparing results from living vegetation with surface sediments (Fig. 7), it is evident that

the total concentration of n-alkanes preserved in the sediments is ~3 orders of magnitude lower

than for living vascular plants, which is consistent with the low sediment TOC values. Although

both exhibit a similar pattern in the odd- over even-numbered carbon prevalence, the pattern was

better expressed in the living vegetation.

18

For sediments, the ACL23–35 ranged from 27.9 to 29.5, with no clear spatial trend, although the

three lowest values were all located in the central part of the Baker-Martínez fjord system (Fig. 8a;

Table S2). The values for modern vegetation were similar (27.7–29.5), except for the shrub

Berberis ilicifolia (26.9) (Fig. S1, Table S3). CPI in surface sediments varied between 1.0 and 7.5

(Fig. 8b; Table S2), indicating a predominantly terrestrial source, and the values were positively

correlated with ACL (r=0.58; p< 0.001). For the vegetation samples, CPI ranged from 9.1−46.3, with

the most extreme value (46.3) being again associated with B. ilicifolia (Fig. S1, Table S3). TAR

values were high and particularly variable throughout the study area. They ranged from 2.7 to 18.7

in the middle and outer sections of the Baker-Martínez fjord system, and from 3.1 to 132.1 near

rivers and in glacial fjords (Fig. 8c, Table S2).

4.5 Siliceous microfossils

The analysis of siliceous microfossils in the surface sediment samples (Table S2) revealed a

dominance of diatoms, composing >99 % of the total siliceous assemblages. Sponge spicules,

silicoflagellates, and phytoliths were rare. The lowest diatom abundances were observed in glacial

fjord Jorge Montt (4 × 103 to 1.2 × 104 valves gdw-1), in contrast to outer fjord stations where peak

values were recorded (2 × 107 valves gdw-1 at station Qu14-E3Q) (Table S2). There was a significant

positive correlation between diatom abundance and biogenic opal values (r= 0.64, p<0.001).

Resting spores of the genus Chaetoceros (C. radicans/cinctus, C. coronatus, C. didymus, C.

diadema, C. debilis) accompanied by Thalassionema nitzschioides and Skeletonema japonicum,

represented almost 50% of the total diatom assemblage. Other diatom groups were also recorded:

coastal planktonic (Thalassiosira eccentrica, T. oestrupii, T. angulata, Rhizosolenia setigera f.

pungens); tycopelagic and benthic (Paralia sulcata, Cocconeis costata, Cocconeis pacifica); and

freshwater (e.g., resting spores of Chaetoceros muellerii, and several species of the genera

19

Tabularia, Fragilaria, Frustulia, Cymbella, Navicula, and Synedra) (Table S4). The contribution of

the freshwater group to the total diatom assemblage fluctuated from 3.2 to 100%; it was the most

abundant group in Jorge Montt fjord sediments and at stations close to the Baker and Pascua

rivers (Fig. 8d, Table S5). Low values were observed at outer sites.

4.6 Principal component analysis of sediment variables

The PCA of the 14 sediment variables showed that the first two principal components explain

68 % of the total variance (Fig. 9). PC1 accounts for 51% of the total variance and the variables

with the highest F1 loadings are diatom abundance (0.96), proportion of freshwater diatoms (-

0.91), δ13C (0.87), distance from the nearest river mouth/calving glacier front (0.80), and

proportion of lithogenic particles (-0.80).

5. Discussion

5.1 Spatial and temporal variations in fjord productivity

Depth-integrated (0−30 m) Chl-a values in spring 2014 (October) ranged from 8 to 114 mg Chl-

a m-2 (Fig. 4a; Table S1) with average values of 42 mg m-2 in Martínez Channel and 49 mg m-2 in

Baker Channel. The low depth-integrated values at COPAS station CO14-E1 near Tortel and at the

mouth of glacial fjord Steffen are comparable to those measured by Jacob et al. (2014) in the same

area. Similarly low values were observed at the head of Baker Channel, near the mouth of Pascua

River (Fig. 4a). In both cases, these low values correspond to sites with particularly low surface

salinities and with the highest suspended sediment concentrations (Fig. 2), in agreement with

González et al. (2013) and Quiroga et al. (2016), suggesting that primary production in the Baker-

Martínez system is limited by low light availability due to abundant suspended sediments in the

water column. This is also in agreement with the regional-scale results of Jacob et al. (2014), who

20

suggested that, in addition to low photosynthetically available radiation south of 46°S, low light

availability due to suspended detritus in the water column − particularly at sites with high

freshwater inputs of glacial origin − might limit primary production to the upper 15 m of the water

column.

In contrast to the inner-fjord sites, Chl-a estimates were particularly high at the outermost

sites during our spring 2014 survey (Fig. 4a, Table S1), in agreement with the summer 2015

(January) observations of Quiroga et al. (2016). This however contrasts with the pattern observed

by Quiroga et al. (2016) in winter 2014 (August), when Chl-a concentrations were overall lower,

and when the highest values occurred at mid-fjord locations in the Martínez Channel.

The clear covariance between fluorescence, which is commonly used as a proxy for

phytoplankton biomass, and salinity in the upper 20m of the water column over a 3-year period

(Fig. 3) suggests that salinity can be used to estimate the position of the productivity front across

the Baker-Martínez fjord system. This is in agreement with both published and unpublished

hydrographic observations for the study area (Meerhoff et al. 2014, Quiroga et al. 2016, Tapia et

al., submitted), which point to a correspondence between the surface salinity front – with

salinities changing rapidly from <10 to >20 PSU – and maximum chlorophyll-a values, which in all

likelihood correspond with areas of greater carbon fluxes from the water column to the fjord

sediments. In Martínez Channel (Fig. 4b), low salinities reached their westernmost position in

December-February (austral summer), which are the months of maximum freshwater discharge

from the Baker river. In fact, the distance between median positions of the 10 and 20 salinity

contours in January-February (Fig. 4b) point to a sharp surface salinity gradient in these months,

due to the high freshwater discharge associated with glacier melting. The other time of the year

when these two contours are close to each other is the month of October, when river influence

reaches its lowest point – due to minimal freshwater discharge in early spring – and marine

21

conditions are most influential (Fig. 4b). Considering that chlorophyll-a tends to significantly

increase seaward of the 10–20 PSU salinity front, the fjord section where productivity increases

significantly should be located between stations 4 and 5 through fall-winter, and farther out

(between stations 5 and 6) in the late spring and summer months (Fig. 4b).

5.2 Water column–surface sediments comparison

The spring-summer pattern of “inner fjord: low Chl-a – outer fjord: high Chl-a” was also

reflected in the Chl-a values measured in surface sediments by Quiroga et al., 2016. Interestingly,

our diatom abundance estimates in the sediments of the Baker-Martinez fjord system seem to

reflect the same pattern (Table S2).

The qualitative diatom analysis of phytoplankton net tows (0‒15 m) yielded 25 taxa (Table S5)

that are common members of the spring micro-phytoplankton assemblage in Patagonian fjords

and channels between ~47°S and 50°S (González et al., 2013). However, there is a striking

difference in the composition of the diatom assemblage present in the water column with the one

preserved in the sediments (Table S5), which was dominated by the robust resting spores of the

genus Chaetoceros. In contrast, the fragile frustules of i.e., Pseudo-nitzschia, Cerataulina, and

Chaetoceros vegetative cells were a minor or missing component of the surface sediments.

Unfortunately, we do not have measurements in the area to be able to discern among the various

factors (grazing, bottom currents, erosion, lateral advection; e.g., Sancetta and Calvert, 1988;

Muñoz and Wellner, 2016) that may have affected and modified the assemblages either in the

water column or at the water/sediment interface.

Freshwater diatoms were only recorded in sediments (ca. 30% of total assemblage on

average), mostly at sites near rivers and glacier fronts (Fig. 8d). We did not observe any freshwater

diatoms in the water column (Table S5), suggesting that freshwater species sink rather rapidly

22

after being discharged by rivers, and/or that the river supply of freshwater diatoms is seasonal (no

significant production of freshwater diatoms in spring). Because freshwater diatoms are washed

into the fjord environment from the continent (mainly by rivers), they should however constitute a

good proxy to reconstruct changes in terrestrial input, as suggested by the significant negative

covariance between the proportion of freshwater diatoms and distance from river mouths/calving

glacier fronts (Fig. 9).

5.3 Variations in sediment composition

Our biogenic opal values (1.1–10.2%; Fig. 5) extend the range of data previously given by

Aracena et al. (2011) for our study area and are in agreement with the ones reported by Silva et al.

(2009) and Bertrand et al. (2012b) for the fjords of northern Chilean Patagonia (1.4−12.6% and

1.9–17.7%, respectively). Spatial distribution of Corg in the surface sediments follows that of

biogenic opal (r=0.72; p<0.001) and diatoms (r=0.50; p<0.01), with the lowest values observed at

stations near glacial fjords and the highest values occurring in the seaward section of the Baker

and Martínez channels (Fig. 5; Table S2). Carbonate contents are about an order of magnitude

lower than opal values, providing evidence that the main producers in the Baker-Martinez fjord

system are silica-based organisms.

The similarity of trends in Corg, biogenic opal, and, to a lesser extent, carbonate contents seems

to largely reflect dilution by lithogenic particles, which are supplied by rivers and calving glaciers.

This is also clearly illustrated in Figure 9, where biogenic opal, carbonate and Corg concentrations

have clear positive PC1 loadings, whereas lithogenic particles are clearly negatively loaded.

Comparing biogenic silica and carbonate productivity between sites would therefore require

measuring accumulation rates at each site, which is beyond the scope of this study. These results

23

also imply that identifying variables of terrestrial input and productivity should focus on indicators

that are independent of accumulation rates.

5.4 Sources of sedimentary organic matter

The sediments of the Baker-Martinez fjord system are characterized by variable amounts of

terrestrial and marine organic matter. In sedimentary environments, bulk N/C and δ13C have been

equally used to estimate the proportions of terrestrial and marine organic matter since terrestrial

vegetation is characterized by lower N/C and δ13C than marine organic matter (Meyers, 1997;

Perdue and Koprivnjak, 2007). Both approaches have been successfully applied to modern and

Holocene sediments from the Chilean fjords (e.g., Sepúlveda et al., 2009, 2011; Rebolledo et al.,

2008; Silva et al., 2011; Bertrand et al., 2012a, 2012b, 2014, 2017; Lafon et al., 2014), for which

regional end-member values have been defined (terrestrial N/C = 0.0624, δ13C = -27.72‰, and

marine N/C = 0.130, δ13C = -19.86‰; Sepúlveda et al., 2011; Bertrand et al., 2012b). In the Baker-

Martinez system, both variables decrease towards the open ocean (Fig. 6b, c; see also the lower

density results of Quiroga et al., 2016), as they do for fjords in northern Patagonia (Silva et al.,

2011).

N/C values measured in the sediments of proglacial fjord Jorge Montt and in the prodeltas of

Pascua and Baker rivers are much higher than expected for proximal environments (Fig. 10). This is

likely due to the combination of very low TOC (~0.2%) in proglacial environments and the

abundance of inorganic nitrogen bound to glacial clays (Kumar et al., 2016), resulting in

particularly high N/C values that cannot be directly used to infer organic matter sources.

Additionally, the relatively high abundance of cyanobacteria (mostly Synechococcus) in Jorge

Montt Fjord (Gutiérrez et al., 2015) may contribute to the high N/C ratio since Synechococcus is

known to efficiently fix nitrogen in proglacial environments (Zhu et al., 2016). This contrasts with

24

sites that are not directly affected by glacial clay input, where N/C and δ13C are positively

correlated (Fig. 10), although glacial clays likely influence N/C values throughout the entire fjord

system since Baker and Pascua rivers, as well as Jorge Montt Glacier, are the main sources of

sediment to this fjord system. In the (non-glacial) fjords further to the north, the fraction of

terrestrial carbon calculated from N/C and δ13C is highly correlated (r=0.94, p < 0.001; Bertrand et

al., 2012b). These observations suggest that δ13C is better suited than N/C to calculate the fraction

of terrestrial carbon in fjord sediments, as it is independent of the fjord’s glacial status.

The TAR ratio for all studied locations was >>1, which can be interpreted as a predominantly

terrestrial origin of the n-alkanes throughout the Baker-Martínez fjord system (Bourbonniere and

Meyers, 1996), in agreement with the observation that the highest TAR values are found close to

rivers and in glacial fjords (Fig. 8c). This interpretation of the TAR values is, however, in

contradiction with the quantification of terrestrial vs. aquatic sources of organic carbon based on

δ13C (Fig. 6d), which shows that terrestrial organic carbon only dominates in the most proximal

environments. The likely reason behind this discrepancy is that land-plant organic matter typically

contains more n-alkanes than marine organic matter, resulting in a disproportionate weight

assigned to land-plant input (Peters et al., 2007). This bias is also likely responsible for the lack of

significant correlation between TAR and δ13C (r=-0.347, p=0.028).

The ACL values measured in our surface sediment samples all fell between 27.9 and 29.5, and

the three lowest values occurred in the central part of the Baker-Martínez fjord system (Fig. 8a).

The lack of a significant relation with distance from river mouths/calving glacier fronts (Fig. 9)

reflects the relatively homogenous nature of the evergreen vegetation along the fjord system

(Luebert and Pliscoff, 2006), and the absence of significant tributaries in the western and central

part of the fjord system, which is also the case for fjords further north (Sepúlveda et al., 2011). The

presence of the lowest ACL values (27.9–28.2) in the central part of the fjord may reflect the

25

particularly high precipitation that occurs on the western slope of the Andes immediately to the

west of the icefields (Garreaud et al., 2013; Lenaerts et al., 2014). Likewise, the high CPI values

indicate that the long-chain n-alkanes are predominantly derived from terrestrial sources, i.e.,

land-plants, suggesting a low contribution of petrogenic carbon to the fjord sediments, in

agreement with the granitoid nature of the regional bedrock (Sernageomin, 2003). The significant

correlation between CPI and ACL (r=0.58; p<0.001) results from the fact that both indexes are

obtained using n-alkane from higher plants (Jeng, 2006).

5.5. Proxies for terrestrial input

Several of the proxies analyzed here are designed to reflect terrestrial input: the N/C molar

ratio, δ13Corg, TAR, and the proportion of freshwater diatoms. All of them are independent of

accumulation rates. In this section, we compare the capability of these indicators to reflect the

terrestrial-to-marine gradient that occurs in the Baker-Martinez fjord system in order to identify

the best proxies for terrestrial input. To do so, we represent their variations with distance from

the river mouths/calving glacier fronts (Fig. 11). Distance is a good indicator of terrestrial input, as

most of the sediment is transported in the upper freshwater lens, making sediment transport

independent of variations in bathymetry (Fig. 2).

As discussed in the previous section, δ13Corg is much more appropriate than N/C to calculate

the proportions of terrestrial and aquatic organic carbon, mostly due to the significant influence of

lithogenic N on the N/C measurements. When looking at variations across the terrestrial-marine

continuum of our study area (Fig. 11), it is evident that the fraction of terrestrial organic carbon

calculated from the bulk δ13Corg measurements progressively decreases with distance from the

river mouth/calving glacier front (Fig. 11a). By comparison, N/C shows very little changes with

distance (Fig. 11b), confirming that the lithogenic N contribution renders the proxy inapplicable in

26

most of the fjord system. This is in contrast with the non-glaciated fjords of northern Chilean

Patagonia, where N/C and δ13Corg co-vary and can generally be used to estimate the proportion of

terrestrial organic carbon in sediments (Sepúlveda et al., 2010; Silva et al., 2011; Bertrand et al.,

2012b). These differences suggest that δ13Corg is much more universally applicable than N/C in

fjord environments to reconstruct variations in organic matter sources through space and time.

End-member mixing models for fjord sedimentary organic matter should therefore be based on

δ13Corg and not N/C. This observation also explains why differences were observed between the

fractions of terrestrial organic carbon calculated from δ13Corg and N/C measurements in sediment

cores from proglacial Chilean fjords (e.g., Bertrand et al., 2012a, 2017), and they suggest that the

discrepancies between the two variables may actually be used as an indicator of glacial activity.

Although the TAR values show a slightly decreasing trend with distance from the river

mouths/calving glacier front (Fig. 11c), the values remain high (~10) across the fjord system due to

land plants containing more n-alkanes than aquatic organic matter. Even at the most oceanic site,

which only contains 20% terrestrial organic matter, the TAR value is clearly terrestrial (7.8–15.2;

Fig. 11c). TAR is therefore rather insensitive to changes in terrestrial input, due to the signal being

almost entirely controlled by land plants. Its use in fjord sediments is therefore not recommended.

The distribution of freshwater diatoms in the sediments of the Baker-Martínez fjord system

reflects input from the continent (Fig. 8d). The major difference with the trend in the fraction of

terrestrial carbon (Fig. 11) is that freshwater diatoms quickly drop within the first 40 km of the

river mouths/calving glacier fronts, whereas the fraction of terrestrial carbon displays a gradual

decrease over the 180 km of the fjord transect. Further offshore than 40 km from river mouths

and calving glacier fronts, the proportion of freshwater diatoms remains relatively stable at ~20%,

suggesting that diatoms are most sensitive in proximal environments (Fig. 11d). Also, there seems

to be no major difference between proglacial and non-glacial fjords, making the freshwater diatom

27

proxy as usable as δ13C to reconstruct changes in terrestrial input, irrespective of the sources of

freshwater (river discharge/calving). The main advantage of this proxy is that it might be less

dependent on the evolution of terrestrial ecosystems (e.g., vegetation cover) than δ13C, making it

readily applicable on longer timescales. Due to its apparent insensitivity to terrestrial inputs at

mid- and outer-fjord sites, its use should be restricted to relatively proximal fjord environments.

Thus, it appears that δ13Corg and the proportion of the freshwater diatoms constitute reliable

proxies of terrigenous input in the Baker-Martínez fjord system. They are the best candidates to

reconstruct changes in river discharge and precipitation, or meltwater input, from fjord sediment

cores. One should however keep in mind that these proxies are relative, i.e., a decrease in marine

productivity would appear similar to as increase in terrestrial input.

6. Conclusions

The terrestrial-marine continuum of the Baker-Martinez fjord system allowed us to investigate

spatial variations in marine and terrestrial indicators, and to identify the best tracers of terrestrial

input. Results show that the signals occurring in the water column and recorded in sediments are

predominantly driven by the input of lithogenic particles and terrestrial organic carbon from rivers

and calving glaciers. As a consequence, organic carbon, biogenic silica, and carbonates increase

with increasing distance from the fjord head. Among all the terrestrial input indicators that were

investigated in this study, the fraction of terrestrial organic carbon calculated from bulk δ13C

measurements and the proportion of freshwater diatoms are by far the most promising proxies to

reconstruct changes in terrestrial input through time using fjord sediment cores. Freshwater

diatoms are most sensitive within the first 40 km from the fjord head, whereas the fraction of

terrestrial organic carbon is better suited for sediment cores collected in mid and outer fjords.

The main advantage of these two proxies is that they are independent of the glacial status of the

28

fjords and their watersheds, which means that they are applicable to reconstruct terrestrial input

in both proglacial and non-glacial fjords. By comparison, n-alkane-based proxies are less promising

owing to the consistent terrestrial bias of the signal due to the disproportionally high production

of n-alkanes by land plants. Likewise, our results show that the N/C ratio is mostly unsuitable to

reconstruct the sources of organic matter due to the relatively high contribution of lithogenic

nitrogen, especially in carbon-poor proglacial environments. Based on these results, we

recommend focusing on δ13C and/or diatom analysis to reconstruct past changes in terrestrial

input from fjord sediment cores. This study therefore sets a baseline for future sediment-based

research in the Baker-Martinez fjord system and in Patagonian fjords in general, and it supports

the use of fjord sediment archives for paleoclimate and paleoenvironmental reconstructions.

7. Acknowledgments

We thank Dr. Marcelo Gutiérrez (UdeC) for kindly providing the Jorge Montt sediment

samples, and Dr. Juan Placencia (UCSC) for sharing some of the CF20 samples used in this study.

Paola Reinoso and Francisco Gallardo (PUCV) are acknowledged for their help in collecting the

CF14 sediment samples for bulk organic geochemical analysis. We also thank Dr. Roberto

Rodríguez (UdeC) for his help with the identification of vascular plants from the Tortel area, and

Víctor Acuña (COPAS Sur-Austral) and Alejandro Ávila (IDEAL Center) for n-alkanes and biogenic

opal analyses, respectively. We are grateful to the captain and crew of the L/C Sur-Austral for their

help during sample collection, and to the CIMAR program (Cruceros de Investigación Marina en

Áreas Remotas; Marine Research Cruises in Remote Areas). The research presented here was

funded by the Center for Oceanographic Research COPAS Sur-Austral (CONICYT PIA PFB31 and

CONICYT PIA APOYO CCTE AFB170006) and by UGent BOF project HYDROPROX (01N02216). Partial

support from the IDEAL Center (grant FONDAP 15150003 to CBL and LR), and projects CONA-C14F

29

08-13 and CONA C20F 14-06 (to NS) is also acknowledged. The authors wish to thank Rolf Kilian

and an anonymous reviewer for their valuable comments.

30

References

Aiken, C.M., 2012. Seasonal thermal structure and exchange in Baker Channel, Chile. Dynamics of

Atmospheres and Oceans 58, 1–19.

Allan, J., Douglas, A.G., 1977. Variation in the content and distribution of n-alkanes in a series of

Carboniferous vitrinites and sporinites of bituminous rank. Geochimica et Cosmochimica Acta

41, 1223–1230.

Aracena, C., Lange, C.B., Iriarte, J.L., Rebolledo, L., Pantoja, S., 2011. Latitudinal patterns of export

production recorded in surface sediments of the Chilean Patagonian fjord (41−55°S) as a

response to water column productivity. Continental Shelf Research 31, 340–355.

Aracena, C., Kilian, R., Lange, C.B., Bertrand, S., Lamy, F., Arz, H.W., De Pol-Holz, R., Baeza, O.,

Pantoja, S., Kissel, C., 2015. Holocene variations in productivity associated with changes in

glacier activity and freshwater flux in the central basin of the Strait of Magellan.

Palaeogeography, Palaeoclimatology, Palaeoecology 436, 112−122.

Bertrand, S., Hughen, K.A., Lamy, F., Stuut, J.B.W., Torrejón, F., Lange, C.B., 2012a. Precipitation as

the main driver of Neoglacial fluctuations of Gualas glacier, Northern Patagonian Icefield.

Climate of the Past 8, 519−534.

Bertrand, S., Hughen, K.A., Sepúlveda, J., Pantoja, S. 2012b. Geochemistry of surface sediments

from the fjords of Northern Chilean Patagonia (44−47°S): Spatial variability and implications for

paleoclimate reconstructions. Geochimica et Cosmochimica Acta 76 (1), 125–146.

Bertrand, S., Hughen, K., Sepúlveda, J., Pantoja, S., 2014. Late Holocene covariability of the

southern westerlies and sea surface temperature in northern Chilean Patagonia. Quaternary

Science Reviews 105, 195–208.

31

Bertrand, S., Lange, C.B., Pantoja, S., Hughen, K., Van Tornhout, E., Wellner, J.L., 2017. Postglacial

fluctuations of Cordillera Darwin glaciers (southernmost Patagonia) reconstructed from

Almirantazgo fjord sediments. Quaternary Science Reviews 177, 265−275.

Bianchi, T.S., Canuel, E.A., 2011. Chemical Biomarkers in Aquatic Ecosystems. Princeton University

Press, Princeton, New Jersey.

Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction and purification. Canadian

Journal of Biochemistry and Physiology 37(8), 911–917.

Bourbonniere, R.A., Meyers, P.A., 1996. Sedimentary geolipid records of historical changes in the

watersheds and productivities of Lakes Ontario and Erie. Limnology and Oceanography 41(2),

352–359.

Bush, R.T., McInerney, F.A., 2015. Influence of temperature and C4 abundance on n-alkane chain

length distributions across the central USA. Organic Geochemistry 79, 65–73

Cáceres, M., 2007. Caracterización Oceanográfica. In: Proyecto Hidroeléctrico Aysén.

Determinación de línea de base de oceanografía flora y fauna marina y estuarina. Informe

Final. Resumen Ejecutivo. Facultad de Ciencias del Mar y Recursos Naturales. Universidad de

Valparaíso. Serie Informes Científicos y Técnicos N° 34/2007.

Caniupán, M., Lamy, F., Lange, C.B., Kaiser, J., Kilian, R., Arz, H.W., León, T., Mollenhauer, G.,

Sandoval, S., De Pol-Holz, R., Pantoja, S., Wellner, J., Tiedemann, R., 2014. Holocene sea-

surface temperature variability in the Chilean fjord region. Quaternary Research 82, 342−353.

Corner, E.D.S., 1979. Pollution Studies with Marine Plankton: Part I. Petroleum Hydrocarbons and

Related Compounds. Advances in Marine Biology 15, 289−380.

32

Cupp, E.E., 1943. Marine plankton diatoms of the west coast of North America. Bulletin of the

Scripps Institution of Oceanography 5, 1–238.

Cranwell, P.A., 1973. Chain-length distribution of n-alkanes from lake sediments in relation to

post-glacial environmental change. Freshwater Biology 3, 259−265.

Dussaillant, A., Benito G., Buytaert, W., Carling, P., Meir, C., Espinoza, F., 2010. Repeated glacial-

lake outburst floods in Patagonia: an increasing hazard? Natural Hazards 54(2), 469–481.

Eglinton, G., Hamilton, R.J., 1963. The distribution of alkanes. In: T. Swain (Ed.), Chemical Plant

Taxonomy, Academic Press, pp. 187-217 .

Eglinton, G., Hamilton, R.J., 1967. Leaf epicuticular waxes. Science 156, 1322−1335.

El Nemr, A., Moneer, A.A., Ragab, S., El Sikaily, A., 2016. Distribution and sources of n-alkanes and

polycyclic aromatic hydrocarbons in shellfish of the Egyptian Red Sea coast. Egyptian Journal

of Aquatic Research 42(2), 121–131.

Garreaud, R.D, Lopez, P., Minvielle, M., Rojas, M., 2013. Large-scale control on the Patagonian

climate. American Meteorological Society 26, 215–230.

Glasser, N.F., Jansson, K.N., Duller G.A.T., Singarayer, J., Holloway, M., Harrison, S., 2016. Glacial

lake drainage in Patagonia (13−8 Kyr) and response of the adjacent Pacific Ocean. Scientific

Reports 6, 21064; doi:10.1038/srep21064.

González, H.E., Castro, L.R., Daneri, G., Iriarte, J.L., Silva, N., Tapia, F., Teca, E., Vargas, C.A., 2013.

Land-ocean gradient in haline stratification and its effects on plankton dynamics and trophic

carbon fluxes in Chilean Patagonian fjords (47−50°S). Progress in Oceanography 119, 32–47.

33

Gutiérrez, M.H., Galand, P.E., Moffat, C., Pantoja, S., 2015. Melting glacier impacts community

structure Bacteria, Archaea and Fungi in a Chilean Patagonia fjord. Environmental

Microbiology 17(10), 3882−3897.

Hasle, G.R., Syvertsen, E.E., 1996. Marine diatoms, in: Tomas, C.R. (Ed.), Identifying marine

diatoms and dinoflagellates. Academic Press, Inc., San Diego, pp. 5–385.

Hebbeln, D., Marchant, M., Freudenthal, T. Wefer, G., 2000. Surface sediment distribution along

the Chilean continental slope related to upwelling and productivity. Marine Geology 164, 119–

137.

Hedges, J.L., Prahl, F.G., 1993. Early Diagenesis: Consequences for Applications of Molecular

Biomarkers. In: Engel M.H., Macko S.A. (Eds.), Organic Geochemistry. Topics in Geobiology, vol

11, Plenum Press, New York, pp 237–253.

Hijmans, R.J., Cameron, S.E, Parra, J.L., Jones, P.G., Jarvis, A., 2005. Very high resolution

interpolated climate surfaces for global land areas. International Journal of Climatology 25,

1965–1978.

Jacob, B.G., Tapia, F.J., Daneri, G., Iriarte, J.L., Montero, P., Sobarzo, M., Quiñones, R.A., 2014.

Springtime size-fractionated primary production across hydrographic and PAR-light gradients

in Chilean Patagonia (41−50°S). Progress in Oceanography 129, Part A, 75–84.

Jacquet, J., McCoy, S. W., McGrath, D., Nimick, D. A., Fahey, M., O’kuinghttons, J., Friesen, B. A.,

Leidich, J., 2017. Hydrologic and geomorphic changes resulting from episodic glacial lake

outburst floods: Rio Colonia, Patagonia, Chile, Geophysical Research Letters 44, 854–864.

Jeng, W.L., 2006. Higher plant n-alkane average chain length as an indicator of petrogenic

hydrocarbon contamination in marine sediments. Marine Chemistry 102, 242−251.

34

Kilian, R., Lamy, F., 2012. A review of Glacial and Holocene paleoclimate records from

southernmost Patagonia (49–55°S). Quaternary Science Reviews 53, 1−23.

Kumar, V., Tiwari, M., Nagoji, S., Tripathi, S., 2016. Evidence of anomalously low δ13C of marine

organic matter in an Arctic Fjord. Scientific Reports 6, 36192; doi:10.1038/srep36192.

Lafon, A., Silva, N., Vargas, C.A., 2014. Contribution of allochthonous organic carbon across the

Serrano River Basins and the adjacent fjord system in Southern Chilean Patagonia: Insights

from the combined use of stable and fatty acid biomarkers. Progress in Oceanography 129,

Part A, 98–113.

Lara, A., Bahamondez, A., González-Reyes A., Muñoz, A.A., Cuq, E., Ruiz-Gómez, C., 2015.

Reconstructing streamflow variation of the Baker River from tree-rings in Northern Patagonia

since 1765. Journal of Hydrology 529, 511–523.

Lenaerts, J.T.M., van den Broeke, M.R., van Wessem, J.M., van de Berg, W.J., van Meijgaard, E.,

van Ulft, L.H., Schaefer, M., 2014. Extreme Precipitation and Climate Gradients in Patagonia

Revealed by High-Resolution Regional Atmospheric Climate Modeling. Journal of Climate 27,

4607-4621.

Liu, H., Liu, W., 2016. n-Alkane distributions and concentrations in algae, submerged plants and

terrestrial plants from Qinghai-Tibetan Plateau. Organic Geochemistry 99, 10–22.

Luebert, F. and Pliscoff, P. 2006. Sinopsis bioclimática y vegetacional de Chile. Editorial

Universitaria, Santiago de Chile.

Marín, V.H., Tironi, A., Paredes, M.A., Contreras, M., 2013. Modelling suspended solids in a

Northern Chilean Patagonia glacier-fed fjord: GLOF scenarios under climate change

conditions. Ecological Modelling 264, 7–16.

35

Mayr, C.C., Rebolledo, L., Schulte, K., Schuster, A., Zolitschka, B., Fösterra, G., Häusserman, V.,

2014. Responses of nitrogen and carbon deposition rates in Comau Fjord (42°S, Southern

Chile) to natural and anthropogenic impacts during the last century. Continental Shelf

Research 78, 29–38.

Meerhoff, E., Castro, L., Tapia, F.J., 2013. Influence of freshwater discharges and tides on the

abundance and distribution of larval and juvenile Munida gregaria in the Baker river estuary,

Chilean Patagonia. Continental Shelf Research 61-62, 1–11.

Meerhoff, E., Tapia, F.J., Castro, L., 2014. Spatial structure of the meroplankton community along a

Patagonian fjord − The role of changing freshwater inputs. Progress in Oceanography 129, Part

A, 125–135.

Meerhoff, E., Tapia, F.J., Sobarzo, M., Castro, L., 2015. Influence of estuarine and secondary

circulation on crustacean larval fluxes: a case study from a Patagonian fjord. Journal of

Plankton Research 37(1), 168−182.

Meerhoff, E., Castro, L.R., Tapia, F.J., Pérez-Santos, I., 2018. Hydrographic and biological impacts of

a Glacial Lake Outburst Flood (GLOF) in a Patagonian fjord. Estuaries and Coasts.

Meyers, P.A., 1997. Organic geochemical proxies of paleoceanographic, paleolimnologic and

paleoclimatic processes. Organic Geochemistry 27(5-6), 213–250.

Moffat, C., 2014. Wind-driven modulation of warm supply to a proglacial fjord, Jorge Montt

Glacier, Patagonia. Geophysical Research Letter 41, 3943−3950.

Moffat, C., Tapia, F.J., Nittrouer, C.A., Hallet, B., Bown, F., Iturra, C., Boldt, K., 2018. Seasonal

evolution of ocean heat supply and freshwater discharge from a rapidly retreating tidewater

glacier: Jorge Montt, Patagonia. Journal of Geophysical Research: Oceans. In press.

36

Mortlock, R.A., Froelich, P.N., 1989. A simple method for the rapid determination of biogenic opal

in pelagic marine sediments. Deep-Sea Research, Part A, 36, 1415–1426.

Muñoz, Y.P., Wellner, J., 2016. Local controls on sediment accumulation and distribution in a fjord

in the West Antarctic Peninsula: Implications for paleoenvironmental interpretations. Polar

Research, 35:1, 25284.

Müller, P.J.; Schneider, R., 1993. An automated leaching method for the determination of opal in

sediments and particulate matter. Deep Sea Research, Part I, 40, 425−444.

Parsons, T.R., Maita, Y., Lalli, C.M., 1984. A Manual of Chemical and Biological Methods for

Seawater Analysis. Pergamon Press, Oxford.

Pendoley, K., 1992. Hydrocarbons in Rowley Shelf (Western Australia) oysters and sediments.

Marine Pollution Bulletin 24, 210–215.

Perdue, E.M., Koprivnjak, J.F., 2007. Using the C/N ratio to estimate terrigenous inputs of organic

matter to aquatic environments. Estuarine Coastal and Shelf Science 73, 65–72.

Peters, K. E., Walters, C. C., Moldowan, J. M., 2007. The Biomarker Guide, Volume 2. Biomarkers

and Isotopes in Petroleum Systems and Earth History. Cambridge University Press, Cambridge.

Pinto, L.A., Bonert, C., 2005. Origen y distribución especial de hidrocarburos alifáticos en

sedimentos de Seno Aysén y Canal Moraleda, Chile Austral. Ciencia y Tecnología Marina 28(1),

35−44.

Piret, L., Bertrand, S., Vandekerkhove, E, Harada, N., Moffat, C., Rivera, A., 2017. Gridded

bathymetry of the Baker-Martinez fjord complex (Chile, 48°S) v1. Fig share.

https://doi.org/10.6084/m9.figshare.5285521.v3.

37

Placencia, J., Llanos, G., Contreras, S., 2017. Organic carbon sources across salinity gradients in

Chilean Fjords: Reloncaví Fjord (∼41ºS) and Southern Patagonian ice fields area (∼48°S).

Geophysical Research Abstracts 19, 9216.

Powell, R.L., Still, C.J., 2009. Biogeography of C3 and C4 vegetation in South America. Anais XIV

Simpósio Brasileiro Sensoriamento Remoto, 2935–2942.Poynter, J. G., Farrimond, P.,

Robinson, N., Eglinton, G., 1989. Aeolian-Derived Higher Plant Lipids in the Marine

Sedimentary Record: Links with Palaeoclimate. In: Leinen M., Sarnthein M. (Eds.),

Paleoclimatology and Paleometeorology: Modern and Past Patterns of Global Atmospheric

Transport. NATO ASI Series (Series C: Mathematical and Physical Sciences). Springer,

Dordrecht, vol 282, pp. 435–462.

Quiroga E., Ortiz, P., Reid, B., Villagrán, S., Gerdes, D., Quiñones, R., 2012. Organic enrichment and

structure of macrobenthic communities in the glacial Baker Fjord, northern Patagonia. Chile.

Journal of the Marine Biological Association UK 92, 73−83.

Quiroga, E., Ortiz, P., González, R., Tapia, F., Pérez-Santos, I., Rebolledo, L., Reid, B., Mansilla, R.,

Pineda, C., Cari, I., Salinas, N., Montiel, A., Gerdes, D., 2016. Seasonal patterns in the benthic

realm of a glacial fjord (Martinez Channel, Chilean Patagonia): the role of suspended

sediment and terrestrial organic matter. Marine Ecology Progress Series 56, 31−50.

Rebolledo, L., Lange, C.B., Figueroa, D., Pantoja, S., Muñoz, P., Castro, R., 2005. 20th century

fluctuations in the abundance of siliceous microorganisms preserved in the sediments of the

Puyuhuapi Channel (44ºS), Chile. Revista Chilena de Historia Natural 78(3), 469–488.

38

Rebolledo, L., Sepúlveda, J., Lange, C.B., Pantoja, S., Bertrand, S., Hughen, K., Figueroa, D., 2008.

Late Holocene marine productivity changes in Northern Patagonia-Chile inferred from a multi-

proxy analysis of Jacaf channel sediments. Estuarine, Coastal and Shelf Science 80, 314–322.

Rebolledo, L., Lange C.B., Bertrand, S., Muñoz, P., Salamanca, B., Lazo, P., Iriarte, J.J., Vargas, G.,

Pantoja, S., Dezileau, L., 2015. Late Holocene precipitation variability recorded in the

sediments of Reloncaví Fjord (41°S, 72° W). Quaternary Research 84(1), 21–36.

Rivera, P., 1981. Beiträge zur Taxonomie und Verbreitung der Gattung Thalassiosira Cleve.

Bibliotheca Pycologica 56, 1–220.

Rivera, A., Koppes, M., Bravo, C., Aravena, J.C., 2012. Little Ice Age advance and retreat of Glaciar

Jorge Montt, Chilean Patagonia. Climate of the Past 8, 403–414.

Rodríguez, R., Marticorena, A., Teneb, E., 2008. Plantas vasculares de los ríos Baker y Pascua,

Región de Aisén, Chile. Gayana Botánica 65(1), 39–70.

Romero, O.E., Hebbeln, D., 2003. Biogenic silica and diatom thanatocoenosis in surface sediments

below the Peru-Chile Current: controlling mechanisms and relationship with productivity of

surface waters. Marine Micropaleontology 48(1-2), 71−90.

Rosell-Melé, A., McClymont, E.L., 2007. Biomarkers as Paleoceanographic Proxies. In: Hillaire-

Marcel, De Vernal, A. (Eds.), Proxies in Late Cenozoic Paleoceanography, vol. 1. Developments

in Marine Geology. Elsevier. Amsterdam, pp. 441−-490.

Ross, L., Pérez-Santos, I., Valle-Levinson, A., Schneider, W., 2014. Semidiurnal internal tides in a

Patagonian fjord. Progress in Oceanography 129, Part A, 19−34.

39

Ross, L., Valle-Levinson, A., Pérez-Santos, I., Tapia, F.J., Schneider, W., 2015. Baroclinic annular

variability of internal motions in a Patagonian fjord. Journal of Geophysical Research: Oceans

120, 5668−5685.

Round, F.E., Crawford, R.M., Mann, D.G., 1990. The Diatoms: Biology & Morphology of the Genera.

Cambridge University Press, Cambridge, 760 p.

Sancetta, C., Calvert, S.E., 1988. The annual cycle of sedimentation in Saanich Inlet, British

Columbia. Implications for the interpretation of diatom fossil assemblages. Deep-Sea

Research, Part A, 35, 71–90.

Schlitzer, R., 2017. Ocean Data View, http://odv.awi.de

Schrader, H., Gersonde, S., 1978. Diatoms and silicoflagellates. In: Zachariasse, W.J., et al.

Microplaeontological counting methods and techniques - an exercise on an eight metres

section of the lower Pliocene of Capo Rossello. Sicily. Utrecht Micropaleontological Bulletins

17, pp. 129−-176.

Sepúlveda, J., Pantoja, S., Hughen, K.A, Lange, C.B., González, F., Muñoz, P., Rebolledo, L., Castro,

R., Contreras, S., Ávila, A., Rossel, P., Lorca, G., Salamanca, M., Silva, N., 2005. Fluctuations in

export productivity over the last century from sediments of a southern Chilean fjord (44°S).

Estuarine Coastal and Shelf Science 65, 587–600.

Sepúlveda, J., Pantoja, S., Hughen, K.A., Bertrand, S., Figueroa, D., Leon, T., Drenzek, N.J., Lange, C.,

2009. Late Holocene sea-surface temperature and precipitation variability in northern

Patagonia, Chile (Jacaf Fjord, 44⁰S). Quaternary Research 72, 400−409.

40

Sepúlveda, J., Pantoja, S., Hughen, K.A., 2011. Sources and distribution of organic matter in

northern Patagonian fjords, Chile (~44–47°S): A multi-tracer approach for carbon cycling

assessment. Continental Shelf Research 31(3-4), 315−329.

Sernageomin, 2003. Mapa geológico de Chile versión digital, escala 1/1.000.000.

Silva, N., Calvete, C., Sievers, H.A., 1998. Masas de agua y circulación general para algunos canales

australes entre Puerto Montt y Laguna San Rafael, Chile (Crucero Cimar-Fiordo 1). Ciencia y

Tecnología del Mar 21, 17–48.

Silva, N., Haro, J., Prego, R., 2009. Metals background and enrichment in the Chiloé Interior Sea

sediments (Chile). Is there any segregation between fjords, channels and sounds? Estuarine,

Coastal and Shelf Science 82, 469–476.

Silva, N., Vargas, C., Prego, R., 2011. Land-ocean distribution of allochthonous organic matter in

surface sediments of the Chiloé and Aysén interior seas (Chilean Northern Patagonia).

Continental Shelf Research 31(3-4), 330−339.

Sims, P.A. (Ed.), 1996. An Atlas of British Diatoms. Biopress Ltd, Bristol, United Kingdom.

Smith, R.W., Branchi, T., Allison, M., Svage, G., Galy, X., 2015. High rates of organic carbon burial in

fjords sediments globally. Nature Geoscience 8, 450–453.Tapia, F.J., D. Henriquez, C. Moffat,

V. Manriquez. Hydrographic structuring of larval abundance and stage structure for the squat

lobster Munida gregaria in a Patagonian fjord. Submitted to Estuaries and Coasts.

Torres, R., Silva, N., Reid, B., Frangopulos, M., 2014. Silicic acid enrichment of subantarctic surface

water from continental inputs along the Patagonian archipelago interior sea (41–56°S).

Progress in Oceanography 129, Part A, 50–61.

41

Thornton, S.F., McManus, J., 1994. Application of organic carbon and nitrogen stable isotope and

C/N ratios as source indicators of organic matter provenance in estuarine systems: evidence

from the Tay Estuary, Scotland. Estuarine Coastal and Shelf Science 38, 219–233.

Vargas, C.A., Martínez, R.A., San Martín, V., Aguayo, M., Silva, N., Torres, R., 2011. Allochthonous

subsidies of organic matter across a lake-river-fjord landscape in the Chilean Patagonia:

Implications for marine zooplankton in inner fjord areas. Continental and Shelf Research 31(3-

4), 187–201.

Vidal, O.J., Bannister, J., Sandoval, V., Perez, J., Ramirez, C., 2011. Woodland communities in

Chilean cold-temperate zone (Baker and Pascua basins): Floristic composition and morpho-

ecological transitions. Gayana Botanica 68 (2), 141–154.

Willis, M.J., Melkonian, A.K., Pritchard, M.E., Ramage, J.M., 2012. Ice loss rate of Northern

Patagonian Icefield derived using a decade of satellite remote sensing. Remote Sensing of

Environment 117, 184–198.

Witkowski, A., Lange-Bertalot, H., Metzeltin, D., 2000. Diatom Flora of Marine Coasts I. In: Lange-

Bertalot, H. (Ed.), Iconographia Diatomologica, vol. 7, Diversity – Taxonomy - Identification.

Koeltz Scientific Books, Königstein, Germany, 925 pp.

Zhu, Z.Y., Wu, Y., Liu, S.M., Wenger, F., Hu, J., Zhang, J., Zhang, R.F., 2016. Organic carbon flux and

particulate organic matter composition in Arctic valley glaciers: examples from the Bayelva

River and adjacent Kongsfjorden. Biogeosciences 13, 975−987.

42

Figure captions

Figure 1 – Location of the Baker-Martínez fjord system in Chilean Patagonia, with indication of the

main glaciers and rivers influencing the fjord system. Sediment sampling stations are represented

by red circles (see Table S2 for details). SPI = Southern Patagonian Icefield.

Figure 2 – Bathymetry of the Baker-Martínez fjord system (from Piret et al., 2017) and location of

the CTD stations investigated during the October 2014 COPAS Sur-Austral campaign (top). The

cross-sections of temperature, salinity, dissolved oxygen concentration, turbidity, and

fluorescence (from top to bottom) are shown for Martínez Channel (left) and Baker Channel

(right). CTD data were plotted using Ocean Data View (Schlitzer, 2017). The bathymetry is not

shown on the cross-sections due to its relatively low resolution but shallow (<100m) sills likely

occur in the center of Martínez Channel. The color scales are open-ended. Note the change in

depth scale for the lowermost (turbidity and fluorescence) cross-sections.

Figure 3 – Principal Component Analysis (PCA) of the four main environmental variables

(temperature, salinity, turbidity, and fluorescence) measured in the upper 20m of the water

column at the 17 stations shown in Fig. 2, during four campaigns in spring (October 2014, October

2015) and summer (January 2016, March 2016). For each variable, the averages of all

measurements obtained in the upper 20m of the water column were used. The results show high

covariance between fluorescence and salinity, suggesting that salinity is the main environmental

variable controlling the distribution of phytoplankton biomass in the Baker-Martínez fjord system.

Figure 4 – Spatial and temporal variability in phytoplankton biomass in the Baker-Martínez fjord

system. (a) Chlorophyll-a concentrations in October 2014 (austral spring) integrated over the

43

upper 30 m of the water column (COPAS 2014 campaign; data in Table S1). Underlined labels

indicate stations where phytoplankton samples were collected in the water column and analyzed

(results in Table S5). (b) Monthly statistics for the position of the 10 and 20 PSU surface salinity

contours along Martínez channel, based on hourly records of salinity gathered from five stations

(numbers 2 to 6) along the channel from October 2012 through November 2013. Symbols

correspond to the median position of the contours for each month in the record. Error bars

correspond to the first and third quartiles (25% and 75% percentiles) of the distribution for each

month. Based on the strong covariance between fluorescence and salinity (Fig. 3), a significant

increase in phytoplankton biomass is expected to occur between stations 4 and 5 through most of

the year, except for summer months (Dec–Feb) when it moves to the outer section between

stations 5 and 6.

Figure 5 – Spatial distribution of the four main components of surface sediments in the Baker-

Martínez fjord system. a) Organic matter (wt %), b) biogenic opal (wt %), c) CaCO3 (wt %), and d)

lithogenic particles (wt %). Data presented in Table S2.

Figure 6 – Spatial distribution of bulk organic geochemical parameters in surface sediments of the

Baker-Martínez fjord system. a) Corg (wt %), b) N/C (molar ratio), c) δ13Corg (‰), and d) proportion

of terrestrial organic carbon (wt %). Data presented in Table S2.

Figure 7 – Comparison of n-alkanes distribution and concentration between a) living vegetation at

Caleta Tortel, and b) surface sediments of the Baker-Martínez fjord system. Individual plant

species are shown in Fig. S1. The error bars indicate the 1σ standard deviation.

44

Figure 8 – Spatial distribution of n-alkanes and diatom-based indices in the Baker-Martínez fjord

system. a) ACL, b) CPI, c) TAR, and d) freshwater diatoms (%). The main rivers and glaciers are

indicated. Data are presented in Table S2.

Figure 9 – Spearman Principal Component Analysis (PCA) of the main variables measured in the

surface sediments of the Baker-Martínez fjord system: Corg, biogenic opal, total diatoms, CaCO3,

lithogenic particles (LP), δ13Corg, δ15 N, N/C ratio, total n-alkanes, carbon preference index (CPI),

average chain length (ACL), terrigenous/aquatic ratio (TAR), and proportion of freshwater diatoms.

Figure 10 – Molar N/C versus δ13Corg of surface sediment samples from the Baker-Martínez fjord

system. The end-member values are from Hebbeln et al. (2000)and Bertrand et al. (2012b):

terrestrial: N/C = 0.0624 and δ13C = -27.72‰; aquatic: N/C = 0.130 and δ13C = -19.86‰. The N/C

and δ13C measurements obtained on river sediment sample RS11-05 confirm the selection of the

terrestrial end-member values.

Figure 11 – Variations of the main terrestrial proxies with distance from the nearest river mouth or

calving glacier front: (a) Fraction of terrestrial carbon (fT) based on δ13Corg (%), (b) molar N/C ratio;

(c) n-alkane-based terrestrial-to-aquatic ratio (TAR) and (d) proportion of freshwater diatoms (%).

All proxies were plotted with the most terrestrial values towards the top of the figure. The open

symbols correspond to samples from proglacial fjord Jorge Montt.

45

46

47

48

49

50

51

52

53

54

55

Highlights

The Baker-Martinez fjord system covers a complete terrestrial-to-marine continuum

Variability in sediment composition is mostly explained by distance from river mouths

Ntot/Corg measurements are strongly affected by glacial clay input

δ13Corg and freshwater diatoms are the best indicators of terrestrial input

56

Graphical abstract