SPATIAL VARIABILITY IN PROGLACIAL FJORD SEDIMENT COMPOSITION ALONG THE SOUTHERN PATAGONIAN ICEFIELD (47–51°S)
Matthias Troch Student number: 01205240
Promotor: Prof. Dr. Sébastien Bertrand
Jury: Dr. Inka Meyer, Dr. Juan Placencia
Master’s dissertation submitted in partial fulfilment of the requirements for the degree of Master of Science in Geology
Academic year: 2016 - 2017
Acknowledgements
Before reading my master dissertation, I would like to seize the opportunity to thank some people who helped and supported me during this study.
First of all, I would like to thank my promotor, Prof. Dr. Sébastien Bertrand. You gave me the opportunity to study the Patagonian fjords, a pristine area of natural beauty. I greatly appreciate your support and readiness to answer my questions. I am very thankful for all your advice and constructive discussions we had. From you, I learned a lot of how to work and think as a geologist. For this, I am very grateful for having you as a promotor.
I would also like to thank Dr. Juan Placencia and Dr. Lorena Rebolledo for providing samples from the Cimar 20 and Copas 2014 cruises in the Chilean fjords. Without these samples, and the additional data provided by them, this study would not have been possible.
Furthermore, I would like to thank the Renard Centre of Marine Geology for the usage of the equipment in their laboratory and computer lab. I would also like to acknowledge all the people working at the RCMG, who taught me how to work with different kinds of software. I am especially grateful to Veerle Vandenhende from the laboratory staff of the Department of Geology, for helping me in the laboratory, performing ICP- analysis on my samples, and providing answers to all my questions.
I would like to express my gratitude to the people who work in the group of Prof. Dr. Sébastien Bertrand: Loic Piret, Elke Vandekerkhove and Dawei Liu. They helped me to solve my problems in the laboratory and practice lots of analytical techniques.
I would like to thank all my fellow students. We supported each other during our entire study and, especially, during our final, hard times of writing our master dissertations. I enjoyed the great times we had during the classes, field trips and non-geological related activities. I am grateful to all French wine farmers, who made our evenings during the Alps excursion.
I would also like to thank Bruno Vanderborght and Freddy Van Puyenbroeck for reviewing my master thesis. Their comments and suggestions helped me to improve this dissertation.
Finally, I would really like to thank my mother, Kristin Vanderborght, and my girlfriend, Hannah Dermaux, for their great support and endless encouragement during my student years. To my mother and father, thank you for providing me the chance of education and your unconditional help and love. To my sister, Hannah Troch, after this experience, I might consider to become a doctor, but probably not the type you want… I would like to thank my girlfriend for all the Netflix evenings helping me relax after a stressful day of writing. This being said, I would also like to thank Netflix for the free month, this came in pretty handy.
Thank you and enjoy reading!
Matthias
2
The Story of my Research
As library books provide information regarding human history, geological records, e.g., ice cores, tree rings, and fjord, lake and marine sediments, provide information about our climate history. The language in which this information is written is however not entirely clear yet. For my master dissertation, I investigated sediments from the fjords in Chilean Patagonia, where glaciers from the Southern Patagonian Icefield terminate, to understand how glacier size variations are written in sediments. More specifically, the goal of my research was to identify fjord sediment properties that can be used to reconstruct past glacier advance and retreat. Since no similar research has been done in this study area, the results of this master dissertation provide a valuable asset to reconstruct past glacier size variability in Chilean Patagonia.
Since the size of glaciers is affected by various climate parameters such as temperature and precipitation, reconstructing past variations in glacier size can provide important indications on past climate variations in the glacier’s region. Comprehending past climate change is essential to predict future climate change. Glacier advance might by the result of decreased temperature and/or increased precipitation. Glacier retreat might be caused by increased temperature and/or decreased precipitation.
The Patagonian fjords form a pristine area of natural beauty. In the future, human influence in this region may become critical, disturbing local fauna and flora. Scientific research in this area is only in its early stages. Learning more about this region can help preserve it.
3
Abstract
An accurate understanding and interpretation of fjord sediment records is essential to reconstruct glacier and climate variability. In order to predict the evolution of glaciers in a changing climate, one must comprehend the dynamic response of glaciers to climate change. Fjord sediments can provide a detailed record of past glacier variability. To better interpret such proxy records, accurate proxies of past glacier mass balance variations are needed. With this in mind, this study investigates the spatial variability of physical properties, sediment composition and bulk organic and inorganic geochemistry of surface sediments within the fjords north and west of the Southern Patagonian Icefield (SPI). The studied surface sediment samples are located along transects from the fjord heads towards the open ocean. The main research objective is to identify bulk organic and inorganic geochemical properties to estimate past changes in terrestrial sediment supply towards these fjords systems. In the case of glaciated fjords, terrestrial sediment supply can serve as an indicator for glacier variability. This study makes use of two fjord systems: the Baker Martinez Fjord Complex (BMFC), which is located to the north of the SPI and mostly composed of river-fed fjords, and the fjord system west of the SPI, which is mainly influenced by calving glaciers. The main difference between both fjord systems is the presence and absence of calving glaciers, hence the presence and absence of icebergs and ice rafted debris.
Our results indicate that ice rafted debris can serve as a sedimentological proxy to differentiate glaciated fjords, i.e., fjords with calving glaciers, from non glaciated fjords, i.e., fjords without calving glaciers, in Chilean Patagonia. The spatial variability of the bulk inorganic geochemistry is significantly different between fjords with different sources of terrestrial sediment, i.e., calving glaciers or rivers. Therefore, the use of bulk inorganic geochemical proxies should be confined to one specific fjord, and not over an entire fjord system. Within the BMFC, the elemental log-ratios Fe/Al, Zr/Al, Ti/Al, Mg/Al and litho-Si/Al show promising results to estimate past changes in terrestrial sediment supply. Due to a low sampling density, no such proxies were identified within the fjord system west of the SPI. Throughout both fjord systems, carbon and nitrogen isotopic composition of fjord sediments are well suited to estimate past changes in terrestrial sediment supply. More depleted δ13C (‰) and δ15N (‰) values are found near glacier fronts and river outlets, while less depleted δ13C (‰) and δ15N (‰) values represent more marine environments. The spatial variability of the atomic N/C ratio does not show a significant terrestrial to marine trend in the studied fjord systems. In addition, our results show that the mass-specific magnetic susceptibility signal of fjord sediments can be used to differentiate terrestrial sediment supply from the Northern and Southern Patagonian Icefields, due to the distinct lithologies on which these ice masses occur. The results of this master dissertation should help improve the interpretation of physical, sedimentological and geochemical data from fjord sediment cores in terms of sediment provenance and glacier variability.
4
Table of content Acknowledgements ...... 2 The Story of my Research ...... 3 Abstract ...... 4 Table of content ...... 5 1. Introduction ...... 7 2. Regional Setting ...... 10 2.1 Patagonian Geology ...... 10 2.2 Tectonic Setting ...... 11 2.3 Patagonian Volcanism ...... 12 2.4 Patagonian Climate ...... 13 2.5 Patagonian icefields ...... 14 2.5.1 Quaternary ...... 14 2.5.2 Holocene...... 15 2.5.3 Recent ...... 16 2.6 The fjord system ...... 17 2.6.1 Fjords influenced by glaciers ...... 17 2.6.2 Fjords influenced by rivers...... 18 2.6.3 Water masses within the fjord system ...... 20 3. Material & Methods ...... 23 3.1 Material ...... 23 3.2 Methods ...... 24 3.2.1 Magnetic susceptibility ...... 24 3.2.2 Grain size ...... 24 3.2.3 Ice Rafted Debris ...... 25 3.2.4 Loss on ignition ...... 25 3.2.5 Bulk organic geochemistry ...... 26 3.2.6 Carbonate analysis ...... 28 3.2.7 Biogenic silica ...... 29 3.2.8 Lithogenic Particles ...... 29 3.2.9 Bulk inorganic geochemistry ...... 30 3.2.10 Statistical analyses ...... 31 4. Results ...... 32 4.1 Physical properties ...... 32 4.1.1 Magnetic Susceptibility ...... 32
5
4.1.2 Grain Size and Ice Rafted Debris ...... 33 4.2 Sediment composition ...... 34 4.2.1 Biogenic opal ...... 34 4.2.2 Carbonates ...... 34 4.2.3 Organic matter ...... 35 4.2.4 Lithogenic Particles ...... 35 4.3 Bulk organic geochemistry ...... 37 4.3.1 Carbon isotopic composition ...... 37 4.3.2 Nitrogen isotopic composition ...... 38 4.3.3 Atomic N/C ratio ...... 38 4.3.4 Terrestrial and marine fractions of organic carbon ...... 38 4.4 Bulk inorganic geochemistry ...... 40 4.5 Principal Component Analysis ...... 44 5. Discussion ...... 46 5.1. Spatial variability ...... 46 5.1.1 Grain Size ...... 47 5.1.2 Sediment Composition ...... 47 5.1.3 Bulk organic Geochemistry...... 50 5.1.4 Bulk inorganic Geochemistry ...... 51 5.2. Glaciated fjords versus non glaciated fjords ...... 57 5.3. Provenance of terrestrial sediment supply ...... 59 5.4. Future research ...... 63 6. Conclusion ...... 63 7. References ...... 66 8. Appendix ...... 72
6
1. Introduction
Glaciers are currently retreating at an alarming rate worldwide as a consequence of anthropogenic climate change. Often, glacier retreat is used as an indicator of global warming (Bertrand et al., 2012a). Predicting the future evolution of glaciers in our changing world requires an accurate understanding of how glaciers respond to changes in climate on timescales that extend beyond instrumental records. Such an understanding can be obtained by comparing proglacial sediment records of glacier mass balance to independent climate reconstructions (Bertrand et al., 2012a). Interpreting proglacial sediment records in terms of glacier variability, however, is not as straightforward as it may seem (Bertrand et al., 2012a). As a baseline for an improved interpretation of geochemical data from fjord sediment cores, this study aims to investigate the spatial variability of the physical properties, sediment composition and bulk organic and inorganic geochemistry of fjord surface sediments to identify geochemical proxies of terrestrial sediment supply into the Patagonian fjords.
Previous research, using a similar approach, was already conducted by Bertrand et al. (2012b), Faust et al. (2014), Munoz & Wellner (2016) and Rebolledo et al. (2015). Except for Munoz & Wellner (2016), these studies were conducted in river-fed fjords. Munoz & Wellner (2016) investigated the local controls on sediment accumulation and distribution in Flanders Bay, West Antarctic Peninsula. Munoz & Wellner (2016) discuss grain size as an indicator of proximity to the ice margin. Within their surface sediment samples, they observed an increase in grain size from the inner to the outer part of the bay (Munoz & Wellner, 2016). Munoz & Wellner (2016) conclude that their findings highlight the variability in sedimentation patterns within fjords, which provides valuable evidence of the complexity that may occur in the sedimentary record found within fjord environments (Munoz & Wellner, 2016).
Faust et al. (2014) studied the organic and inorganic geochemistry of the surface sediments within Trondheimsfjord, central Norway. They showed that within Trondheimsfjord, the origin of organic matter as well as the distribution of CaCO3 can be used as a proxy for marine primary productivity, variable inflow of Atlantic water and changes in river run-off (Faust et al., 2014). In addition, grain size independent elemental ratios of Ni/Al, K/Al and K/Ni are possible proxies for variations in river discharge and precipitation in the Trondheimsfjord region (Faust et al., 2014).
Within the surface sediments of the Reloncavi Fjord (Northern Chilean Patagonia), geochemical parameters (biogenic silica, total organic carbon, N/C and δ13C) revealed a spatial gradient of allochthonous and autochthonous organic matter with a clear terrigenous signal near the mouths of major rivers (Rebolledo et al., 2015). Rebolledo et al. (2015) showed that freshwater diatom concentrations and the mass accumulation rates of terrigenous organic carbon can be used as proxies for precipitation-driven river discharge in Reloncavi Fjord sediments.
Results from Bertrand et al. (2012b) demonstrate that, under the cold climate conditions of Patagonia, chemical weathering is weak and the inorganic geochemical composition of the fjord sediments is primarily controlled by hydrodynamic mineralogical sorting, i.e., the intensity of river discharge. Bertrand et al. (2012b) state that the elemental ratios Fe/Al, Ti/Al and Zr/Al can be used to estimate changes in the energy of terrestrial sediment supply into the Patagonian fjords through time. Within the same fjords of northern Patagonia (44°S–47°S), Sepúlveda et al. (2011) found a strong gradient (84% difference) of organic carbon sources between open ocean areas and those located towards inner fjord and/or river outlets using a 13 15 combination of Corg/Ntot, δ C and δ N data.
7
Chilean Patagonia constitutes one of the most important and extensive fjord regions worldwide, involving the area from Reloncavi Fjord (41.5°S) to Cape Horn (55.9°S), covering almost 240,000 km2 (Silva et al., 2011). This region is characterized by a highly fragmented coastline and myriad islands, channels, and fjords (Silva et al., 2011). Within this region, two major icefields are located: the Northern Patagonian Icefield (NPI) and the Southern Patagonian Icefield (SPI) (Fig. 1.1). The Southern Patagonian Icefield is the largest mid-latitude icefield in the Southern Hemisphere. A comprehensive survey of Patagonian glaciers by Lopez et al. (2010) found that the majority of Patagonia’s glaciers have retreated, some of them quite dramatically, during the second part of the twentieth century and the NPI and SPI have shrunk considerably in the last decades. Nevertheless, a few glaciers in the SPI have advanced or remained stable (Lopez et al., 2010).
This study comprises two areas within the complex Patagonian fjord system: 1) the Baker Martinez Fjord Complex (BMFC), and 2) the channels west of the SPI: Canal Messier, Canal Wide, Canal Trinidad, Canal Piclon, and Canal Fallos (Fig. 1.1). The BMFC comprises the Martinez Channel and the Baker Channel, and separates the NPI and SPI at approximately 48°S (Fig. 1.1). The BMFC receives freshwater from the Baker and Pascua rivers, plus contributions from the Jorge Montt glacier and the Bravo and Huemuleis rivers (Fig. 1.1) (Aiken, 2012). The Baker and Pascua rivers are fed by large amounts of precipitation and glacial meltwater from the NPI and SPI, respectively. Exchange of saline water with the open ocean occurs through Golfo de Penas (Fig. 1.1). Over the coming decades, the river discharge regime into the BMFC may significantly alter due to reduced precipitation and accelerated icefield melting associated to climate change, plus possible hydroelectric dam construction on the Pascua and Baker rivers (Aiken, 2012; Dussaillant et al., 2012; Garreaud et al., 2013).
The second study area comprises the main channels west of the SPI, with a focus on Canal Messier and Canal Wide (Fig. 1.1). These channels are less influenced by river discharge. Meltwater discharge from calving and land-based glaciers forms the most important source of freshwater and terrestrial material within this area. Canal Messier is connected to the open ocean through Golfo de Penas, while Canal Wide is connected to the open ocean through Canal Trinidad, Canal Concepción and Golfo Trinidad (Fig. 1.1). The most important difference between both study areas, is that calving glaciers are absent in the BMFC, except for Jorge Montt Fjord, and present in the channels west of the SPI.
As mentioned above, most of the existing work on Patagonian surface sediments is based on river-fed fjords (Bertrand et al., 2012b; Rebolledo et al., 2015; Sepúlveda et al., 2011). However, terrestrial sediment supply is significantly different between glacier-fed and river-fed fjords. With this in mind, this study has three main research objectives: 1) investigating the geochemistry of surface sediments within both study areas in order to evaluate the ability of bulk organic and inorganic geochemical properties to estimate past changes in terrestrial sediment supply, which, in the case of glaciated fjords, might be used as an indicator for glacier variability, 2) to identify sedimentological properties which can be used to differentiate fjords influenced by calving glaciers, i.e., glaciated fjords, and fjords that are not, i.e., non glaciated fjords, and 3) to identify the provenance of sediment reaching the BMFC, i.e., Northern or Southern Patagonian Icefield. This will be based on the investigation of the physical properties, sediment composition and bulk organic and inorganic geochemistry of surface sediment samples, generally along transects from the fjord heads towards the open ocean (Fig. 1.1). The surface sediment samples that will be used in this thesis were collected during the Cimar 20 and Copas 2014 research cruises in the Chilean fjords (Fig. 1.1). Both cruises took place in October 2014 (Fig. 1.1). In addition, biogenic opal and bulk organic geochemical data from the Gutierrez 2013 and Copas 2014 research cruises were provided by Rebolledo et al. (in prep.) (Fig. 1.1 & Appendix 8.1).
8
Fig. 1.1. A) Geographical location of the study area (modified from Aiken (2012)), B) Location of the two main study areas: 1. Baker Martinez Fjord Complex (BMFC), and 2. Channels and fjords west of the SPI (dashed red rectangles), and sampling sites of the Cimar 20 and Copas 2014 research cruises, presented on a simplified map of the Patagonian Andes. In addition, biogenic opal data from the Gutierrez 2013 research cruise was provided by Rebolledo et al. (in prep.). Some of the Gutierrez 2013 samples are located in Jorge Montt Fjord (third study area, dashed red rectangle). NPI = Northern Patagonian Icefield, SPI = Southern Patagonian Icefield.
9
2. Regional Setting
2.1 Patagonian Geology
The Andean Cordillera is a 7,500 km long mountain chain along the western margin of South America. The cordillera is subdivided in the Northern (12°N–5°S), Central (5°S–33°S) and Southern (33°S–56°S) Andes (Stern, 2004). The Southern Andes comprises the Patagonian Andes and Fuegian Andes. The study area is located in the Patagonian Andes.
The geology of the Southern Andes is dominated by the Patagonian batholith (40°S–56°S) (Fig. 2.1), which is a plutonic complex related to subduction processes along the Andean continental margin (Hervé et al., 2007; Pankhurst et al., 1999). The batholith formed episodically from the Late Jurassic until Late Cenozoic times (Pankhurst et al., 1999).
The Patagonian batholith is subdivided into three parts: the Northern Patagonian batholith (NPB; 39°S–47°S), the Southern Patagonian batholith (SPB; 47°S–53°S) and the Fuegian batholith (FB; 53°S–55°S) (Hervé et al., 2007; Moreno & Gibbons, 2007).
The NPB formed on the western edge of the South America plate, from the Late Jurassic until the Pleistocene, in response to the eastward subduction of the Nazca plate (Fig. 2.1) (Pankhurst et al., 1999). The NPB is predominantly composed of granodiorites and tonalities, and smaller amounts of granite, diorite and gabbro (Pankhurst et al., 1999).
The SPB formed from the Late Jurassic until the Neogene, by the subduction of the Antarctic plate beneath the South America plate (Fig. 2.1) (Hervé et al., 2007). The SPB has a more variable magmatic composition containing gabbro, gabbro-diorite, diorite, tonalite, granodiorite and granite (Hervé et al., 2007). Fig. 2.1. Patagonian geology and tectonic setting (Hervé et al., 2007) The Fuegian batholith developed over the present day Scotia microplate (Hervé et al., 2007) and consists mainly of gabbros and granitoids (Peroni et al., 2009).
The SPB intruded through the late Paleozoic to Mesozoic metamorphic complexes of the Patagonian Andes (Fig. 2.1) (Allamand et al., 2008). These metamorphic complexes contain the Coastal accretionary complexes and the Eastern Andes metamorphic complex (EAMC) (Fig. 2.1) (Allamand et al., 2008). The EAMC consists mainly of polydeformed turbidite successions, with minor amounts of limestone, metabasites and low-grade metapelites (Allamand et al., 2008; Moreno & Gibbons, 2007). Its regional metamorphic grade is in the greenschist facies (Allamand et al., 2008). The Cordillera Darwin metamorphic complex (CDMC) is located in the Fuegian Andes and consists of metasedimentary and metavolcanic units (Allamand et al., 2008; Moreno & Gibbons, 2007). These units have a late Paleozoic to early Mesozoic supposed age (Allamand et al., 2008; Moreno & Gibbons, 2007).
10
2.2 Tectonic Setting
The large scale tectonic setting of the region includes three plates that meet at the Chile Triple Junction: the Nazca and Antarctic plates, separated by the South Chile Ridge, and the South American plate, under which the two others subduct (Fig. 2.1) (Cande & Leslie, 1986; Lagabrielle et al., 2000). In the past, this subduction process has lead to enormous earthquakes, such as the Mw 9.5 Valdivia earthquake in 1960 (Melnick et al., 2009). Within the study area, the Liquine-Ofqui fault system (LOFS) forms the dominating regional tectonic feature (Fig. 2.2). This NNE-SSW trending fault system has a trench-linked dextral strike-slip structure (Cembrano et al., 1996). Multiple authors explain the occurrence of the LOFS to be created in response to the oblique subduction of the Nazca plate beneath the South America plate (Cembrano et al., 1996; Cembrano et al., 2000; Diraison et al., 1998; Nelson et al., 1994; Thomson, 2002). In addition, the tectonic activity of the southern segment of the LOFS results from the subduction of the South Chile Ridge beneath the South America plate (Cembrano et al., 2000; Murdie et al., 1993; Nelson et al., 1994; Thomson, 2002). Cembrano et al. (1996) proposed a minimum Miocene-Pliocene age for the onset of the LOFS deformation. Through plate motion modeling, Pardo‐Casas & Molnar (1987) suggested a minimum Eocene-Miocene age. It has been estimated that the deformation by the LOFS spanned from ca. 11 to ca. 3 Myr, although the system is still partly active to the present day (Glasser & Ghiglione, 2009). Fig. 2.2. Tectonic and volcanic setting. Fault systems are Besides the LOFS, the studied fjord system is cut indicated by red, dashed lines (modified from by multiple smaller faults (Fig. 2.2) (Glasser & SERNAGEOMIN (2003)). Earthquakes with a minimum Ghiglione, 2009; SERNAGEOMIN, 2003). magnitude of 2.5 that occurred within the study area in a period of 14 years before the sampling expeditions (2000- 2014) are indicated by circles (source: It has been observed by multiple authors that the earthquake.usgs.gov). Volcanoes within the study area are sediments within the Patagonian fjords can be indicated by volcano symbols (source: volcano.si.edu). disturbed by earthquake-triggered deposits resulting from this tectonic activity (Chapron et al., 2006; Sepúlveda et al., 2010; St-onge et al., 2012; Van Daele et al., 2013). Within the Baker and Martinez Channels, care should therefore be taken when interpreting the sedimentological data in the scope of palaeoclimatological investigations, since these sediments are possibly disturbed by (mega)turbidites (Piret, 2016). The earthquakes that occurred within the study area with a minimum magnitude of 2.5 in a period of 14 years before the sampling expeditions (2000–2014) are visualized in figure 2.2. The two strongest earthquakes occurred in 2002 and 2009, with a
11
magnitude of 5.6 and 5.1, respectively (Fig. 2.2). The most recent earthquake took place on July 9th 2014, in the southwestern part of the study area, with a magnitude of 4.7 (Fig. 2.2).
Agurto-Detzel et al. (2014) studied the local seismicity of the area in the vicinity of the Chile Triple Junction between 44°S and 49°S during the years 2004–2005. A total of 519 events with magnitudes varying between 0.5–3.4 ML were detected (Agurto-Detzel et al., 2014). The significant majority of these events took place north of the study area, beyond the limits of figure 2.2 (Agurto-Detzel et al., 2014). This seismic activity preceded an earthquake sequence during 2007 in this area (44°S–49°S) (Agurto-Detzel et al.,
2014). This earthquake sequence was related to the reactivation of the LOFS, which peaked with a 6.2 MW earthquake (Agurto-Detzel et al., 2014). This 6.2 MW earthquake had an epicenter far north of the study area (45.243°S 72.648°W) (source: earthquake.usgs.gov ).
From this data, it can be inferred that the seismic influence on the surface sediment samples of this study is rather low to negligible.
2.3 Patagonian Volcanism
Four active volcanoes are located within the study area: Lautaro, Viedma, Aguilera and Reclus (Fig. 2.2). They are all located within the Andean Austral Volcanic Zone (AVZ) (Stern, 2008). The last eruptions from the Lautaro, Viedma, Aguilera and Reclus volcanoes date from 1979 CE, 1988 CE, 3,000 ± 100 yr BP (Stern, 2008) and 1908 CE, respectively. The 1979 CE and 1988 CE eruptions of the Lautaro and Viedma volcanoes reached a Volcanic Explosivity Index (VEI) of 2, the 1908 CE eruption of Reclus reached a VEI of 1. The above mentioned volcanic eruption information was obtained from the Smithsonian Institution (source: volcano.si.edu).
Stern (2008) discusses the five most important Holocene explosive eruptions in the southern Andes. The tephra units resulting from these eruptions were assigned to four different volcanoes: Aguilera, Mt Burney, Hudson and Reclus (Stern, 2008). Isopach maps from Stern (2008) and Vandekerkhove et al. (2016) show that the volcanic ash from these large eruptions did not reach the study area. Vandekerkhove et al. (2016) state that volcanic ash from volcanoes in northern Chilean Patagonia is predominantly deposited on the eastern side of the volcanoes, reflecting the influence of the prevailing westerly winds on the distribution of pyroclastic material. This is valid for the entire southern Andes: ‘ash plumes originating from volcanoes of the Southern Andes are typically transported and deposited eastward due to the prevailing strong westerly winds’ (Caniupán et al., 2011; Kilian et al., 2003).
Andosols, i.e., postglacial volcanic ash soils, are very widespread in Chile. They occur between 36°S and 47°S (Vandekerkhove et al., 2016). Andosols do not occur within the study area (Gut, 2008). The soil map constructed by Gut (2008) clearly shows that the study area is dominated by histosols (peaty, organic soils) and podzols (stratified, organic rich soils, typical for boreal forests) (Covelo et al., 2008; Gut, 2008; Rosling et al., 2003).
From the above discussion, it can be inferred that the influence of pyroclastic material and volcanic ash soils on the surface sediment samples is low to negligible.
12
2.4 Patagonian Climate
The climate of southern South America is dominated by the humid southern westerly winds coming from the Pacific Ocean (Fig. 2.3a). The interplay between the Westerlies and the orographic high of the Andes results in a strong zonal asymmetry in the precipitation patterns (Fig. 2.3a) (Garreaud et al., 2013). In western Patagonia (Chilean Patagonia), the annual mean precipitation ranges between 5,000 and 10,000 mm, resulting in a hyperhumid climate with a modest seasonal cycle (Garreaud et al., 2013). In Chilean Patagonia, the precipitation is due to the uplift of moist air provided by the Westerlies (Garreaud et al., 2013). In contrast, the mean annual precipitation decreases to less than 300 mm just a few kilometers eastward of the Andes (Garreaud et al., 2013). In this region, Eastern Patagonia (Argentinean Patagonia), forced subsidence causes very dry conditions (Garreaud et al., 2013). This leads to arid, highly evaporative conditions and a strong seasonal cycle in Argentinean Patagonia (Garreaud et al., 2013). These observations by Garreaud et al. (2013) confirm the statement of Mercer (1982); Western Patagonia has a maritime climate and Eastern Patagonia has a continental climate. The precipitation gradient in the study area is visualized in figure 2.3c (Hijmans et al., 2005).
Temperature in Patagonia doesn’t show the strong West-East gradient as for precipitation (Fig. 2.3b) (Garreaud et al., 2013). An overall decreasing temperature towards the South can be observed (Fig. 2.3b). From 42°S to 49°S, an area containing the NPI and the northern part of the SPI, the mean annual temperature is about 2.8°C (Sagredo & Lowell, 2012). South of 49°S, an area containing the southern part of the SPI, the mean annual temperature is about 3.7°C (Sagredo & Lowell, 2012). In both areas the mean monthly temperature varies between -2°C in austral winter (July) and 8°C in austral summer (January) (Sagredo & Lowell, 2012). C
Fig. 2.3 A) Annual precipitation (modified from Casanova et al. (2013)) and B) mean annual temperature (modified from Sagredo & Lowell (2012)) maps of southern South America. Major wind patterns across the west coast of southern South America are indicated by blue arrows in (A) (modified from Sagredo & Lowell (2012)). The study area is indicated by black rectangles in (A) and (B). C) Mean annual precipitation across the study area. This map was created using data from the Worldclim database (Hijmans et al., 2005). Red line in (C) indicates the profile visualized in figure 2.8.
13
2.5 Patagonian icefields
The two largest temperate icefields of the Southern Hemisphere are located within the Patagonian Andes: the Northern Patagonian Icefield (NPI) and the Southern Patagonian Icefield (SPI) (Fig. 1.1) (Aniya, 2013; Rignot et al., 2003; Rivera et al., 2007). A smaller temperate icefield is located within the Fuegian Andes: the Cordillera Darwin Icefield (CDI; centered on 54°30’S) (Caniupán et al., 2011; Warren & Aniya, 1999). The NPI (47°00’S, 73°39’W) is 120 km long and 40–60 km wide, covering an area of ~4,200 km2 and capping the Andes between altitudes of 700–2,500 m a.s.l. (Bennett & Glasser, 2011; Glasser et al., 2008). The larger SPI (~13,000 km2) stretches north to south for 360 km between 48°50’ and 51°30’S, with a mean width of ~40 km (Bennett & Glasser, 2011; Glasser et al., 2008).
On the western side of the NPI, the annual precipitation increases from 3,700 mm at sea level to an estimated maximum of 6,700 mm at 700 m a.s.l. (Bennett & Glasser, 2011). On the eastern side, the precipitation decreases sharply (Bennett & Glasser, 2011). As the NPI, the SPI has very strong west to east precipitation contrasts (Bennett & Glasser, 2011; Garreaud et al., 2013). These precipitation contrasts are due to the climate setting as described above (section 2.4 ‘Patagonian Climate’). As a result, the western glaciers have high mass-balance gradients and are very dynamic, while the eastern glaciers are less dynamic (Bennett & Glasser, 2011). This means that the accumulation areas of the glaciers on the westerns side contain more ice mass than those on the eastern side. This results in more pronounced glacier movements and higher mass release in the ablation zone.
Two types of glaciers dominate the NPI and SPI: 1) tidewater glaciers calving into the Chilean fjords, mostly to the west of the icefields, 2) outlet glaciers terminating in proglacial lakes, east of the icefields (Mercer, 1970; Warren & Aniya, 1999; Warren & Sugden, 1993). Warren & Sugden (1993) suggest that the mass-balance of the glaciers on the western side of the NPI and SPI is controlled by precipitation, while that of the glaciers on the eastern side is controlled by temperature. This is countered by Harrison & Winchester (2000), since they observed similar accumulation rates and patterns on both sides of the NPI. Bertrand et al. (2012a) confirm the dominant control of precipitation on the mass-balance of the Gualas glacier, located on the western side of the NPI, during the last 5,400 years. In addition, Bertrand et al. (2012a) question the precipitation/temperature control on the mass-balance of western NPI glaciers in a warmer future climate. Ackert et al. (2008) investigated moraines east of the SPI. They suggest that the response of the SPI, during the Younger Dryas, was in function of varying amounts of easterly-sourced precipitation, rather than regional cooling (Ackert et al., 2008).
2.5.1 Quaternary
The NPI and SPI contracted and expanded several times during the Quaternary (Glasser et al., 2016; Harrison & Glasser, 2011; Kaplan et al., 2004; Kaplan et al., 2005; Mercer, 1976). During glacial periods they converged, together with the CDI, to form the large Patagonian Ice Sheet (PIS; 38°S–56°S) (Glasser et al., 2008; Glasser & Jansson, 2008). The PIS was at its largest extent during the Last Glacial Maximum (LGM; 25.6–20.4 kyr BP) (Caniupán et al., 2011; Sugden et al., 2005). After the LGM and before the Antarctic Cold Reversal (ACR; 14.6–12.8 kyr BP), an unknown amount of regional glacier retreat occurred (García et al., 2012), probably resulting in the disintegration of the PIS into the NPI and SPI. Turner et al. (2005) state that the final separation of the NPI and SPI occurred ca. 12.8 kyr BP.
14
During the last deglaciation (18.0–11.5 kyr BP), the variations of the Patagonian icefields are still highly debated (Fig. 2.4) (García et al., 2012). García et al. (2012) report that SPI glaciers advanced during the ACR, reaching a maximum extent by ~14,200 ± 560 yr BP. Glacier advances during the ACR are supported by numerous authors (Fig. 2.4) (Menounos et al., 2013; Moreno et al., 2009; Strelin et al., 2011; Turner et al., 2005). This was followed by glacier regression and deglaciation by 12.5 kyr BP, early in the Younger Dryas (YD; 12.9–11.7 kyr BP) (García et al., 2012; Mercer, 1982; Sugden et al., 2005). Numerous authors discuss whether the Patagonian icefields retreated or advanced during the Younger Dryas (Fig. 2.4). García et al. (2012) report about a retreating SPI, while Glasser et al. (2012) state an advance of the NPI, during the Younger Dryas (Fig. 2.4). Glasser et al. (2012) do not reject that the advance of the NPI already started during the ACR. Both García et al. (2012) and Glasser et al. (2012) hypothesize that these advances are due to changes in the position of the southern westerly winds, resulting in a period of cooling and regionally increased precipitation. Menounos et al. (2013) support the glacier advances during the ACR and YD (Fig. 2.4), based on 10Be dated valley moraines in the Fuegian Andes. This is countered by 14C and 10Be exposure ages from the eastern flank of the NPI, implying a deglaciation around 12.8 kyr BP (Turner et al., 2005) (Fig. 2.4). The glacier retreat during the YD is supported by Strelin et al. (2011) (Fig. 2.4), based on 14C dated moraines from Lago Argentino, east of the SPI. Ackert et al. (2008) investigated moraines from the same Lago Argentino, they observed a readvance of the SPI at, or shortly after, the end of the YD (Fig. 2.4).
By the Pleistocene-Holocene transition the large PIS separated into the NPI and SPI (Glasser et al., 2004; Turner et al., 2005), which caused a new drainage route to open to the west (Glasser et al., 2016). This route drained paleo glacial lake General Carrera through the BMFC to the Gulf of Penas (Glasser et al., 2016).
Fig. 2.4. Several studies proposing different onsets for the re-advance of glaciers of the NPI and SPI (modified from Vandekerkhove (2014)). LGM = Last Glacial Maximum, ACR = Antarctic Cold Reversal, and YD = Younger Dryas.
2.5.2 Holocene
Two major schemes for the Holocene glaciations, i.e., the Neoglaciations, have been proposed (Fig. 2.5) (Aniya, 2013; Glasser et al., 2004; Mercer, 1982). The older ‘Mercer-type’ chronology contains three Neoglaciations: I) 4,500–4,000 yr BP, II) 2,700–2,000 yr BP, and III) during the Little Ice Age (LIA; 17th–19th centuries) (Aniya, 2013; Glasser et al., 2004; Mercer, 1970, 1976, 1982). This chronology is based on 14C dated moraines from the NPI, SPI and the Chilean Lake District (Mercer, 1970, 1976, 1982). The second chronology, the ‘Aniya-type’ chronology, contains four Neoglaciations: I) ca. 3,600 yr BP, II) ca. 2,400– 2,000 yr BP, III) ca. 1,600–900 yr BP, and IV) during the LIA (17th–19th centuries) (Aniya, 1995, 1996, 2013;
15
Glasser et al., 2004). This chronology is based on 14C dated moraines from the SPI (Aniya, 1995, 1996, 2013). Glasser et al. (2004) state that these chronologies should be regarded as broad regional trends, since there are also dated examples of glacier advances outside these time periods. The main differences between the two chronologies are: 1) the age of Neoglaciation I, and 2) the existence of a Neoglaciation at 1,600–900 yr BP (Aniya, 2013). Based on dates given by other studies, Aniya (2013) proposed a new scheme of five Neoglaciations (Fig. 2.5): I) 4,500–4,000 yr BP, II) 3,600–3,300 yr BP, III) 2,700–2,000 yr BP, IV) 1,600–900 yr BP, and V) during the LIA (17th–19th centuries). In addition, there were two earlier Holocene glaciations, probably at 5,700–5,000 yr BP and 8,100–6,800 (or 7,500) yr BP (Aniya, 2013). Two older ages, 8,800–8,500 yr BP and 9,700–9,100 yr BP, remain uncertain (Aniya, 2013).
During the LIA, there were regional differences in the glacier advances: in the NPI the LIA maximum mostly occurred during the 19th century, whereas in the SPI it occurred one or three centuries earlier (Aniya, 2013; Masiokas et al., 2009). Aniya (2013) explains this lag between the NPI and SPI to be linked to the north- south shift and oscillation of the southern westerly winds.
Fig. 2.5. Neoglacial chronologies, i.e., the ‘Mercer-type’ and ‘Aniya-type’ chronologies, for the fluctuations of Patagonian glaciers in the Holocene (Glasser et al., 2004). Shaded areas represent periods of glacier expansion (Glasser et al., 2004), blue bars represent the five Neoglaciations as proposed by Aniya (2013). The overall ice volume is presented schematically and not intended to represent specific ice volumes at any one moment in time (Glasser et al., 2004). The indicated ages are in 14C yr BP. (Modified from Glasser et al. (2004))
2.5.3 Recent
From 1870 to 2011, a general glacier retreat has been observed in the NPI and SPI (Davies & Glasser, 2012). This has been confirmed for the 20th and 21st centuries (Aniya, 1999; Aniya et al., 1997; Lopez et al., 2010; Masiokas et al., 2009; Mercer, 1982; Rignot et al., 2003; Rivera et al., 2007; Rivera et al., 2012). Both Aniya et al. (1997) and Mercer (1982) recognized greater retreating rates on the eastern side of the icefields than on the western side. Numerous authors stated multiple causes for these retreating glaciers: increasing temperature, decreasing precipitation, surface gradient around the equilibrium line, calving status, basin geometry, ratio of the accumulation area to the total area, subglacial topography, response time, glacier dynamics… (Aniya, 1999; Aniya et al., 1997; Lopez et al., 2010; Masiokas et al., 2009; Rignot et al., 2003). By 2011, the NPI and SPI reached retreating rates of 9.4 km2 yr-1 (0.23% yr-1) and 20.5 km2 yr- 1 (0.15% yr-1), respectively (Davies & Glasser, 2012). From 1945 to 1996, the NPI and SPI contributed 0.038 mm yr-1 to global sea-level rise, which is 3.6% of the total sea-level change (Aniya, 1999). Rignot et
16
al. (2003) estimated that the two icefields jointly contributed 0.042 ± 0.002 mm yr-1 to global sea-level in the period 1968/1975 to 2000 but that this has doubled to 0.105 ± 0.011 mm yr-1 in the more recent years 1995-2000.
2.6 The fjord system
2.6.1 Fjords influenced by glaciers
The western side of the SPI is dominated by tidewater glaciers calving into the Chilean fjords (Mercer, 1970; Warren & Aniya, 1999; Warren & Sugden, 1993). Tidewater glaciers are grounded ice masses with their margins partially submerged in the ocean and often terminating in fjords (Fig. 2.6) (Mugford & Dowdeswell, 2011). Meltwater flows to the glacier bed through englacial channels (Fig. 2.6). Once at the bed, the meltwater flows through the subglacial drainage system (Fig. 2.6) (Mugford & Dowdeswell, 2011). As the meltwater flows along the base of the glacier, it entrains and transports glacially eroded basal sediment (Mugford & Dowdeswell, 2011). Eventually, the sediment-loaded meltwater reaches the glacier front, it emerges from one or more glacial conduits and it enters the adjacent fjord (Fig. 2.6) (Mugford & Dowdeswell, 2011).
The three main primary processes of sediment transport in glaciated fjords are: 1) glacial meltwater plumes, 2) underflows (hyperpycnal flows), and 3) iceberg rafting (Fig. 2.6) (Chu, 2014; Mugford & Dowdeswell, 2011; Mulder et al., 2003). In addition, gravitational processes including slumps, slides, debris flows and turbidity currents are processes which play an important role in reworking fjord-floor deposits (Mugford & Dowdeswell, 2011). Sea ice also transports sediment from littoral to deep water, by actively freezing sediment within the ice and by passive loading from colluvial, aeolian and fluvial processes (Mugford & Dowdeswell, 2011).
Fig. 2.6. Calving glaciers and sediment transport processes (modified from Chu (2014)).
17
When the fresh, turbid glacial meltwater emerges from the glacial conduits into the adjacent fjord, it is almost always less dense than the ambient saline water within the fjord (Mugford & Dowdeswell, 2011). The combination of the buoyancy forces with a pressure drop results in the formation of a rising meltwater plume (Mugford & Dowdeswell, 2011). When the glacial meltwater plume reaches the water surface it spreads as a radial surface gravity current (Fig. 2.6) (Mugford & Dowdeswell, 2011). Sedimentation from these glacial meltwater plumes occurs when the sediment fall velocity exceeds the entrainment velocity (Mugford & Dowdeswell, 2011). When these glacial meltwater plumes move further down fjord, sediment dispersal and settling rates are further influenced by tides, wind and sea ice (Chu, 2014). To create an underflow (hyperpycnal flow; Fig. 2.6), the emerging glacial meltwater must have a sediment load exceeding 30 kg/m3 (Mugford & Dowdeswell, 2011). In polar waters it must exceed 40 kg/m3 (Mugford & Dowdeswell, 2011; Mulder et al., 2003). These limits depend on convective instability and local hydrodynamic and climatic conditions (Mulder et al., 2003). Hyperpycnal flows can only transport suspended material, i.e., particles finer than medium sands, over long distances (Mulder et al., 2003). Ice rafting is another process responsible for sediment transport into glaciated fjords (Fig. 2.6) (Kuijpers et al., 2016). All grain sizes, from clay to boulder size, may potentially be transported by icebergs or sea ice drift (Kuijpers et al., 2016). Due to this process, coarser particles can be transported further in the fjords than they could through glacial meltwater plumes or underflows (Caniupán et al., 2011; Kuijpers et al., 2016). Grain size is another important factor determining sediment transport through fjords (Boldt, 2014; Syvitski & Shaw, 1995). Based on Stokes law of settling velocity, coarser particles have greater settling velocities in comparison to finer particles. Coarser particles will settle more easily from the glacial meltwater plumes and underflows than finer particles. Therefore, fine-grained sediment will be transported further in the fjord system, while coarser sediment will settle down more easily close to the glaciers fronts (Boldt, 2014; Syvitski & Shaw, 1995).
2.6.2 Fjords influenced by rivers
The Baker and Martinez Channels are influenced by major river systems: the Pascua river and the Baker river, respectively (Fig. 1.1). Within these fjord systems, the rate of sediment accumulation is directly related to river dynamics (Syvitski et al., 1987; Syvitski & Shaw, 1995). When the freshwater of a river discharges into a fjord, containing large volumes of saline water, a freshwater plume is formed (Syvitski et al., 1987; Syvitski & Shaw, 1995). Due to density differences, a seaward flowing surface layer is formed (Syvitski & Shaw, 1995). Along its path, the surface layer entrains saline water, and new sea water must enter the fjord at depth (Syvitski & Shaw, 1995). This circulation of deeper waters may be relatively decoupled from river discharge in fjords where the brackish surface layer only occupies a small fraction of the water column (Aiken, 2012). The salinity of the surface layer increases down-fjord (seaward) and with depth (Syvitski et al., 1987; Syvitski & Shaw, 1995).
Within a river-influenced fjord, the freshwater plume, resulting from the river discharge, spreads laterally to a width determined by the narrow parts of the fjord (Syvitski & Shaw, 1995). During its lateral spread, the freshwater plume passes through a zone of deceleration, which is a function of both spreading and mixing between the discharged river water and the surrounding brackish/saline water within the fjord (Syvitski & Shaw, 1995). As a consequence, the turbidity of the surface layer decreases seaward (Syvitski et al., 1987). In the outer part of the fjord, external agents such as tidal currents, wind, shoreline morphology, and the earth’s rotation influence the river plume circulation (Syvitski et al., 1987; Syvitski & Shaw, 1995).
Within a river-influenced fjord, the sediment load supplied by the river separates into two components seaward of the river mouth, i.e., the bedload and suspended load (Syvitski & Shaw, 1995). The bedload (traction and saltation) material settles quickly, close to the river mouth (Bennett & Glasser, 2011; Syvitski & Shaw, 1995). The suspended load is carried seaward within the river plume (Bennett & Glasser, 2011; Syvitski & Shaw, 1995). The suspended load is composed mostly of sand to clay-size mineral grains, and
18
is often referred to as glacial or rock flour (Syvitski & Shaw, 1995). When the surface layer mixes with the ambient saline water, the suspended particles undergo enhanced settling (Syvitski & Shaw, 1995). The settling enhancement is due to flocculation (particles join together), pelletization or agglomerative processes (Syvitski & Shaw, 1995). Due to flocculation, particles smaller than 10 µm attain settling velocities of around 100 m per day (Syvitski & Shaw, 1995). This settling rate is some 10 to 1000 times larger than if the particles settled solo and as predicted by Stoke’s settling theory (Syvitski & Shaw, 1995). The down-fjord sedimentation rate decreases exponentially with distance from the river mouth, which is associated with a concomitant decrease in the size of particles that settle out (Syvitski & Shaw, 1995). However, new sediment sources can completely alter the grain size properties of the fjord-floor sediments as laid down from the surface layer (Syvitski & Shaw, 1995). Variations in the velocity of the surface layer will affect the ability of the surface layer to carry particles with higher settling velocities (Syvitski & Shaw, 1995). The higher the river discharge is, the higher the influx of fluvially transported sediment is (Syvitski et al., 1987).
The BMFC, with the Baker Channel as principal branch, receives an annual mean freshwater flux over 1,600 m3/s from the Baker and Pascua rivers, plus contributions from the Jorge Montt glacier and the Bravo and Huemuleis rivers (Fig. 1.1) (Aiken, 2012). Dussaillant et al. (2012) report a mean annual discharge rate of approximately 1,100 m3/s for the Baker river. Pantoja et al. (2011) report a mean annual discharge rate of 1,133 m3/s for the Baker river, 753 m3/s for the Pascua river, and 112 m3/s for the Bravo river. Within the BMFC, the total river discharge remains above 700 m3/s on average throughout the year (Aiken, 2012). As a result, a brackish surface layer forms a permanent feature throughout the BMFC (Aiken, 2012). The Baker river is fed by glacier meltwater from the NPI, precipitation (Fig. 2.3c) and by draining water from the General Carrera lake passing through Bertrand lake (Dussaillant et al., 2012). The Pascua river originates in Lake O’Higgins and, after traveling 67 km, flows into Mitchell Fjord (Vargas et al., 2011). In addition, the river receives glacier meltwater from the SPI and large amounts of precipitation (Fig. 2.3c). Both rivers receive an amount of monthly precipitation which remains relatively stable throughout the year (Fig. 2.7b) (Bertrand et al., 2014; Hijmans et al., 2005; Vandekerkhove et al., 2016). Within this fjord system, the peak in riverine freshwater discharge occurs during late summer, due to the snow and glacial melt maximum (Fig. 2.7a) (Aiken, 2012; Vandekerkhove et al., 2016).
Fig. 2.7 A) Daily mean river discharge from the Baker (blue) and Pascua (green) rivers, and combined daily mean river discharge (grey). Peak in river discharge occurs during austral summer (yellow bar) (modified from Aiken (2012)). B) Monthly precipitation (1980–2010) data for meteorological station at Faro San Pedro (modified from Bertrand et al. (2014)). The location of Faro San Pedro is indicated on figure 1.1 (Bertrand et al., 2014).
19
2.6.3 Water masses within the fjord system
Sievers & Silva (2008) analyzed the water masses of the 800 m deep, western part of Golfo de Penas (Fig. 1.1). This showed the presence of three water masses: 1) Subantarctic Water (SAAW), between the surface and 150 m depth, 2) Equatorial Subsurface Water (ESSW), with a core between 200 and 300 m depth, and 3) Antarctic Intermediate Water (AAIW), with its core around 600 m depth (Sievers & Silva, 2008). To the south, in the area of Golfo Trinidad (Fig. 1.1), ESSW was not observed (Sievers & Silva, 2008). However, a salinity maximum attributed to the Western Pacific Subsurface Water (WPSSW) was observed in Golfo Trinidad (Sievers & Silva, 2008). Within the study area, a general two-layer structure, with the upper layer reaching depths of 20–30 m, was observed by Sievers & Silva (2008) (Fig. 2.8). The upper layer consists of Estuarine Water (EW), while the lower layer consists of SAAW (Sievers & Silva, 2008). Within the studied fjord systems, the SAAW mixes with fresh water in different proportions, according to the contributions from rivers, glaciers, coastal run-off, pluviosity, and the distance or proximity of the fresh water sources (Sievers & Silva, 2008). This mixing process results in the production of the Modified Subantarctic Water (MSAAW), with salinities between 31 and 33 psu, and the lower-salinity water of the EW (Sievers & Silva, 2008). Sievers & Silva (2008) constructed a three-category classification for these resulting EW: 1) estuarine-saline water (66% sea water, 21–31 psu), 2) estuarine-brackish water (33–66% sea water, 11–21 psu), and 3) estuarine-fresh water (<33% sea water, 2–11 psu). This classification indicates that when the surface layer of EW moves out from fresh water sources and approaches the ocean, its salinity increases (Sievers & Silva, 2008).
Within the channel connecting Golfo de Penas and Golfo Trinidad, a damming situation occurs between Canal Messier and Canal Wide due to the 80 m deep Angostura Inglesa constriction-sill (Fig. 2.8) (Sievers & Silva, 2008). This sill divides the channel into two separate microbasins: the northern one over 1,300 m deep and the southern one over 750 m deep (Sievers & Silva, 2008). The same water masses (EW, SAAW and MSAAW) occur within both microbasins, although they have a different origin and they do not mix due to the Angostura Inglesa sill (Fig. 2.8) (Sievers & Silva, 2008). The different geographic origin of these waters can be seen in their different characteristics: the southern waters are warmer and have more dissolved oxygen, but lower salinity and density values (Sievers & Silva, 2008). Within the deeper part of the northern basin, Agua Ecuatorial Subsuperficial (AESS) and ESSW are present (Fig. 2.8) (Valdenegro & Silva, 2003).
Fig. 2.8. Schematic vertical circulation model from Golfo de Penas towards Canal Concepción, throughout Canal Messier and Canal Wide. The profile is indicated on figure 2.3c by the red line. EW = Estuarine Water, SAAW = Subantarctic Water, MSAAW = Modified Subantarctic Water, AESS = Agua Ecuatorial Subsuperficial, and ESSW = Equatorial Subsurface Water. C. = Canal, and G. = Golfo. Distance on the horizontal axis is expressed in nautical mile (nm) and depth on the vertical axis in meter (m). (Modified from Sievers & Silva (2008)).
20
For the BMFC area, the bathymetric map of Piret (2016) visualizes the oceanic connection between Golfo de Penas and the fjords (Fig. 2.9). At this oceanic connection, Golfo de Penas reaches depths of 100 m and less, limiting the exchange between the ocean and the BMFC (Aiken, 2012; Piret, 2016). Since the Baker Channel reaches a maximum depth of 1250 m (Fig. 2.9) (Piret, 2016), the general two-layer structure will be conserved within the Baker Channel. This is confirmed by the observations of the EW, SAAW and MSAAW within the Baker Channel by Aiken (2012). The Martinez Channel is much shallower than the Baker Channel, reaching a maximum depth of 535 m (Fig. 2.9) (Piret, 2016). However, this does not complicate the presence of the general two-layer structure.
Fig. 2.9. Bathymetric map of the Baker Martinez Fjord Complex. Depth is indicated by the color scale, distance by the red/white scale. White areas in the North represent fjords where no bathymetric data is available. C: Canal (channel), E: Estero, and R: Rio (river). (modified from Piret (2016))
Within the BMFC, Aiken (2012) identified a River Water (RW) layer with a maximum salinity of 15 psu. The thickness of the RW layer is strongly linked to the river run-off seasonal cycle, which is consistent with the fact that the degree of turbulent mixing depends upon the river discharge (Aiken, 2012). During austral summer, peak flow conditions are reached, and the RW covers the upper 4–7 m of the water column in the BMFC, thinning with distance downstream (Aiken, 2012). The RW layer is at its thinnest during austral autumn and austral winter, which is consistent with the reduction in river discharge (Aiken, 2012). During austral spring, the RW layer is similarly deep as in summer but it thins faster with distance (Aiken, 2012). Despite this seasonal cycle, a permanent strong stratification with the presence of the RW prevails throughout the entire year in the BMFC (Aiken, 2012).
CTD data (Fig. 2.10a) acquired during the Copas Sur-Austral expedition in October 2014 (Rebolledo et al., in prep.) was used to construct salinity profiles through the Baker and Martinez Channels (Fig. 2.10 b & c), similar to the profiles from Rebolledo et al. (in prep.). In the salinity profiles through the Baker and Martinez Channels, a general trend of increasing salinity with depth can be observed (Fig. 2.10 b & c). A thin, low salinity (up to 25 psu) water layer is present in the inner part of the Baker Channel between casts 10 to 17 (Fig. 2.10b). This water layer thins toward the ocean and eventually disappears at cast 4 (Fig. 2.10b). Within the Martinez Channel, a similar thin, low salinity (up to 25 psu) water layer is present from cast 5 up to cast 6 (Fig. 2.10c). This water layer thins significantly between casts 6 and 3, eventually disappearing at cast 4 (Fig. 2.10c). Through the Baker and Martinez Channels, the sea surface salinity shows an increasing trend towards the ocean (Fig. 2.10 b & c).
21
A
B
C
Fig. 2.10. A) Location of the CTD profiles (casts) from the Copas Sur-Austral expedition in October 2014, B) Salinity (psu) profile through the Baker Channel, and C) Salinity (psu) profile through the Martinez Channel. In all profiles (B and C), the casts are represented by vertical black lines and their number is presented above the upper, horizontal axis. (modified from Rebolledo et al. (in prep.))
22
Fig 2.11 Sea surface salinity (SSS; psu) map. To construct this SSS map, the SSS data of the ST- samples (Cimar 20 cruise, October 2014) was combined with the salinity data at 1 m depth of the CTD casts (Fig. 2.9a) located at the position of the E-samples (Copas 2014 cruise) (Fig. 1.1). Salinity data from the Cimar 20 cruise was measured at 1 m depth by CTD scanning. Exceptions are samples ST4 and ST5 which were measured at 0 and 2 meters depth, respectively. Sample ST4 was measured with a hand-held conductivity meter. Salinity data from the Cimar 20 cruise was provided by Dr. Juan Plancencia.
The seaward increasing sea surface salinity trend presented in figures 2.10b and 2.10c is confirmed by the sea surface salinity (SSS; 1 m depth) data measured during the Cimar 20 cruise (October 2014; Fig. 2.11). For the channels and fjords west of the SPI, different salinity trends can be observed (Fig. 2.11). From the outlet of the Bernardo glacier (ST21B) throughout Canal Messier, SSS increases towards Golfo de Penas (Fig. 2.11). From the outlet of the Pio XI glacier (ST28) throughout Canal Wide, SSS decreases towards the open ocean (Fig. 2.11). Samples located in a more coastal environment (ST80, ST83, ST86, ST91 and ST93) have the highest salinity values (Fig. 2.11).
3. Material & Methods
3.1 Material
For this study, surface sediment samples obtained during two separate cruises were used. The first set (samples labeled “ST”) was taken with a grab sampler during the Cimar (Cruceros de Investigacion Marina) 20 cruise. The ST-samples were provided by Dr. Juan Placencia (UCSC, Concepción, Chile). The second set (samples labeled “E”) was taken during the Copas 2014 cruise, and was provided by Dr. Lorena Rebolledo (INACh, Punta Arenas, Chile). These samples were taken either using a grab sampler or from the upper centimeter of sediment cores. Both cruises took place in October 2014, and, in both cases, the samples were already freeze-dried before arrival at Ghent University. In addition to these samples, biogenic opal and bulk organic geochemical data from the Gutierrez 2013 and Copas 2014 cruises were provided by Rebolledo et al. (in prep.) (Appendix 8.1).
23
3.2 Methods
3.2.1 Magnetic susceptibility
Magnetic susceptibility (MS) is one of the most frequently measured magnetic parameters on sediment samples. Measuring MS is easy, non-destructive, fast and cost effective (Sandgren & Snowball, 2001). MS is a measure of the net contribution of all constituents of sediment. Bulk magnetic susceptibility is usually used as an indicator for the concentration of allochtonous mineral matter present in the terrigenous fraction of the sediment (Sandgren & Snowball, 2001). Diamagnetic constituents (organic matter, carbonates, biogenic silica, water, etc.) dilute the MS signal of the sample. MS reacts to the net contribution of ferromagnetic and paramagnetic minerals in the sediment, the contribution of paramagnetic minerals becomes more important when ferromagnetic minerals occur in very low concentrations (Sandgren & Snowball, 2001).
The volume-specific magnetic susceptibility of all samples was measured with a Bartington MS2G single- frequency (1.3 kHz) sensor, connected to a Bartington MS3 meter. The sediment was packed into 1 mL plastic vials with a length of 30 mm and a diameter of 8 mm. Every sample was analyzed in duplicate. The sediment weight within each vial was measured with an Ohaus Voyager Pro balance, in order to calculate the mass-specific magnetic susceptibility .
The mass-specific MS was determined to eliminate the density dependency of the volume-specific MS. Mass-specific MS can be calculated by dividing the volume-specific MS by the dry sample weight (Sandgren & Snowball, 2001). Replicate mass-specific MS measurements reached an average relative standard deviation (rsd, %) of 0.31%.
3.2.2 Grain size
Grain size is the most fundamental physical property of sediments. It is used to reconstruct sedimentary processes, especially (hydro)dynamic transport and depositional processes. In addition, grain size has an important effect on the geochemical and mineralogical composition of the sediment.
Grain size was measured on the terrigenous fraction of the sediment samples using a Malvern Mastersizer 3000 laser diffraction particle size analyzer. This instrument measures the size of sediment grains based on the laser diffraction principle. When a sediment grain passes through the laser of the instrument, the laser beam scatters. The intensity and angle of the scattered laser beam is measured by the instrument. The angle of the scattered beam is inversely proportional to the particle size (Ryzak & Bieganowski, 2011). Subsequently, Mastersizer 3000 software uses the Mie and Fraunhofer theories to calculate grain size and grain size distribution (GSD) (Ryzak & Bieganowski, 2011). This software expresses the size of a grain as the diameter (µm) of a sphere with an equivalent volume as the sediment grain.
Prior to the analysis, the optimal sample amount was determined by measuring a few chemically-untreated samples. Results varied between 15 and 25 mg of freeze-dried sediment, except for the coarser samples for which 300 to 500 mg was needed.
The terrigenous fraction was isolated using a chemical treatment with boiling H2O2, HCl and NaOH, to remove organic matter, carbonates and biogenic silica, respectively (Mulitza et al., 2008). Prior to analysis, the samples were treated shortly with boiling Sodium Hexametaphosphate (Calgon), to ensure complete disaggregation of the particles.
24
Within the scope of this study, the Malvern Mastersizer 3000 was combined with a HydroMV dispersion unit resulting in a practical measuring range of 1 nm to 1 mm. A stirrer, operating at 2500 rpm, retained the sediment in suspension. A continuous ultrasound, operating at 10% of its maximum capacity, prevented agglomeration of the particles. Subsequently, a red and blue laser passed through the suspended sediment. Each sample was measured min. 3 times during 12 seconds, by the red and blue laser. Due to insufficient obscuration rates, too low consistency between measurements and weighted residual values above 1, samples ST30, ST80, ST83, ST86, ST93, E6 and E13 were re-measured min. 2 times during 60 seconds, using larger sample amounts.
Eventually, 2 grain size parameters and the GSD plots were exported from the Mastersizer 3000 software. The first parameter is the mean grain size (µm) of the fraction below 150 µm. The second parameter is the proportion of grains larger than 150 µm (volume %). Why these grain size parameters were chosen is explained in result section 4.1.2 ‘Grain Size and Ice Rafted Debris’.
3.2.3 Ice Rafted Debris
Ice Rafted Debris (IRD) is a classic proxy of glacial variability and calving activity (Kuijpers et al., 2016). IRD is a terrigenous material transported within a matrix of ice and deposited when the ice matrix melts (Kuijpers et al., 2016). No standard technique for IRD analysis exists. However, IRD is commonly measured as the weight percentage of the detrital material with grain sizes larger than 63 µm or 150 µm (Kuijpers et al., 2016).
IRD was measured based on the method used by Caniupán et al. (2011). IRD analysis was performed only on the samples containing enough material to not jeopardize other analyses. All samples from the Cimar 20 cruise were analyzed, except for ST8 and ST11 because too little sample material was available. For each sample, 5 gram of sediment was precisely weighted using a Mettler-Toledo XSE204 balance (precision of 0.1 mg) and placed in a 250 mL glass vial. Organic matter, carbonates and biogenic silica were removed using 3.5% hydrogen peroxide, 10% acetic acid and 2N boiling sodium hydroxide, respectively. Subsequently, the samples were wet-sieved at 63 µm and 150 µm and the dry-weight of the 63–150 µm and >150 µm fractions were determined. Eventually, the weight percentages (wt%) of the >63 µm and >150 µm fractions were calculated (Appendix 8.1).
3.2.4 Loss on ignition
The Loss On Ignition method (Heiri et al., 2001) was used to estimate the sediment organic matter and carbonate content. This in order to estimate the ideal amount of sample material needed for bulk organic geochemical and total inorganic carbon analysis. All samples were analyzed, except for samples E3, E4, E8 and Isla Irene because too little material was available. 0.5 g of freeze dried sediment was placed in pre-weighted porcelain crucibles. These were heated in a muffle furnace at 105°C, 550°C and 950°C.
First, the samples were heated at 105°C for 24h, to remove water. Afterwards, the dry weight (DW105) was measured. In a second step, the samples were heated at 550°C for 4h, and the dry weight (DW550) was measured. At 550°C the organic matter oxidizes and CO2 is expelled. The resulting weight loss is an indication for the organic matter content. LOI550 provides an estimate of the organic matter weight percentage.
25
In the last step, the samples were heated at 950°C for 2h, and the dry weight (DW950) was measured. At 950°C, CaCO3 is converted into CaO and CO2. The resulting weight loss is mostly due to the loss of CO2. The carbonate content (weight %) can be estimated by multiplying LOI950 by 2.2.
After each heating step the samples were placed in a desiccator to cool down and to avoid moisture. LOI550 is a good indicator for the organic matter content but it is influenced by the loss of structural water from clay minerals (Howard & Howard, 1990). LOI950 provides a good estimate for the carbonate content, however it is still influenced by organic matter remaining after heating at 550°C.
3.2.5 Bulk organic geochemistry
The total organic carbon (Corg, wt%) and total nitrogen (N, wt%) content and the stable isotope composition (δ13C and δ15N, ‰) were determined with bulk organic geochemical analysis. The samples were prepared at Ghent University and sent to the Stable Isotope Facility (SIF) at UC Davis (University of California) for analysis. For sample preparation the method of Bertrand et al. (2012b; 2014) was followed. The organic matter content (OM, wt%) of the surface sediment samples was calculated by multiplying Corg (wt%) by 2.2 (Bertrand et al., 2012b).
The amount of sediment needed for bulk organic geochemical analysis was calculated based on the relationship between LOI550 and Corg determined on lake and river sediments by Granon (2015): Corg = 0.3566 * LOI550 – 0.6822. We aimed for Corg values between 0.8 mg and 1 mg, with a maximum sample weight of 60 mg. Samples E3, E4, E8 and Isla Irene were not analyzed because too little material was available. However bulk organic geochemical data of these samples was provided by Rebolledo et al. (in prep.) (Appendix 8.1).
The samples were placed in silver capsules and precisely weighted using a Mettler-Toledo XP56 microbalance with a precision of 0.001 mg. Once all capsules were filled, 60 µl of deionised water (Milli-Q water) and 60 µl of sulphurous acid H2SO3 (6–8%) was added to remove inorganic carbon (Verardo et al, 1990). The samples were placed under a fume hood for one hour to react and dried overnight under an infra-red lamp. Finally, the capsules were folded and shaken to avoid spills. During the entire procedure, tweezers were used to avoid contact with the capsules and contamination of the samples.
The samples were analyzed at the SIF using an elemental analyzer (Elementar Vario EL Cube) interfaced to a continuous flow isotope ratio mass spectrometer (PDZ Europa 20-20) (EA-IRMS).
The accuracy and analytical precision was calculated from the analysis of standard samples at the UC Davis. The standard deviation of δ13C (‰) and δ15N (‰) was calculated from standard G-21 (enriched alanine). The sd was 0.10 ‰ for δ13C and 0.08 ‰ for δ15N.
26
The atomic C/N ratio can be used to distinguish the amounts of sedimentary organic matter originating from aquatic and terrestrial sources, based on the characteristic C/N ratios of algae and vascular plants (Fig. 3.1) (Lamb et al., 2006; Meyers & Teranes, 2001; Sepúlveda et al., 2011). Phytoplankton has characteristic C/N ratios between 4 and 10, whereas vascular land plants have characteristic ratios of 20 and greater (Fig. 3.1) (Meyers & Teranes, 2001; Parplies et al., 2008; Sepúlveda et al., 2011). Carbon isotopic composition (δ13C) contributes to the identification of the sources of the sedimentary organic matter (Fig. 3.1) (Meyers & Teranes, 2001; Sepúlveda et al., 2011). In addition, δ13C is a proxy for aquatic productivity and nutrient availability in surface waters; increased primary 13 production yields higher δ C values (Meyers & Fig. 3.1. Typical δ13C and atomic C/N ranges for organic 13 Teranes, 2001). The δ C of phytoplankton inputs to coastal environments (Lamb et al., 2006). organic matter is influenced by secondary factors: pH, temperature, nutrient limitations and growth rate. In specific situations one or more of these secondary factors can become important, providing additional paleoenvironmental significance to the δ13C of organic matter (Meyers & Teranes, 2001). Nitrogen isotopic composition (δ15N) is a proxy, similar to δ13C, for identifying sources of organic matter and primary productivity (Meyers & Teranes, 2001; Sepúlveda et al., 2011). However, care should be taken when interpreting sedimentary δ15N records since the nitrogen biogeochemical cycle is more complicated than the carbon cycle (Meyers & Teranes, 2001). Two important 15 nitrogen sources are coastal marine plankton and C3 land plants, which have δ N values approximating +8.5‰ and +0.5‰ respectively (Meyers & Teranes, 2001). As δ13C, δ15N yields higher values with greater primary productivity (Meyers & Teranes, 2001; Parplies et al., 2008; Robinson et al., 2012; Talbot & Laerdal, 2000; Teranes & Bernasconi, 2000).
Bertrand et al. (2012b) and Sepúlveda et al. (2011) investigated the provenance of organic matter within the fjords of northern Patagonia (~44–47°S). In their studies, the N/C ratio was preferred over the C/N ratio because it better represents the fraction of terrestrial derived organic carbon (Perdue & Koprivnjak, 2007). The marine and terrestrial end-members, as described by Sepúlveda et al. (2011), have δ13C values of -19.8 ± 0.3‰ and -29.3 ± 2.1‰, δ15N values of 9.9 ± 0.5‰ and 0.2 ± 3.0‰, and atomic N/C ratios of 0.127 ± 0.010 and 0.040 ± 0.018, respectively. Bertrand et al (2012b) state slightly different end-member values. The marine and terrestrial end-members, as described by Bertrand et al (2012b), have δ13C values of -19.86‰ and -27.72‰, and atomic N/C ratios of 0.130 and 0.0624, respectively. These values are different because Bertrand et al. (2012b) took the degradation of organic matter during transport in rivers and weathering in soils into account.
The contribution of marine and terrestrial sources to the organic carbon in the surface sediments was calculated by a mixing equation based on the carbon isotopic content δ13C (Sepúlveda et al., 2011). A mixing equation based on the atomic N/C ratios (Bertrand et al., 2010; Perdue & Koprivnjak, 2007) could not be used since the majority of the samples have atomic N/C ratios outside the end-member ranges (see section 4.3.3 ‘Atomic N/C ratio’). This is due to differences in Corg and N content between the study areas.
27
The mixing equation from Sepúlveda et al. (2011) was used:
13 13 13 13 FM = (δ CS δ CT) / (δ CM δ CT) and FT = 1 FM
Here, FM (%) is the fraction of marine organic carbon and FT (%) is the fraction of terrestrial organic carbon. 13 13 13 13 13 13 δ CS is the δ C of a given sample, δ CT is the δ C of the terrigenous end-member and δ CM is the δ C of the marine end-member. The marine and terrestrial end-members of Bertrand et al. (2012b) were used 13 13 in these calculations, i.e., δ CM = -19.86‰ and δ CT = -27.72‰.
3.2.6 Carbonate analysis
Calcium carbonate (CaCO3) within fjord sediments mainly results from aquatic productivity (foraminifera, cocollitophores …) (Bertrand et al., 2012b; Faust et al., 2014). Depending on the regional lithology, detrital carbonates from minerals like calcite, aragonite, dolomite and siderite, can also significantly contribute to the total carbonate content of sediments (Last & Smol, 2002). The distribution of CaCO3 in fjord sediments can be used as a proxy for the variable inflow of certain water masses (Bertrand et al., 2012b; Faust et al., 2014).
The weight percentage of Total Inorganic Carbon (TIC) in the surface sediment samples was determined using an UIC CM5017 coulometer equipped with a CM5330 acidification module (Fig. 3.2). This coulometer has a working range from less than one microgram C up to 10,000 micrograms of C for a single sample. All samples were analyzed, except sample E3 because too little sample material was available. For each sample, 50 mg of sediment was precisely weighted using a Mettler-Toledo XSE204 balance with a precision of 0.1 mg. The sediment was inserted into a reaction flask (glass vial), in which 5 mL of 2N H3PO4 was added by an acid dispenser (Fig. 3.2). The phosphoric acid caused the inorganic carbon to evolve into
CO2 (CaCO3 CaO + CO2). This reaction product was transported by a CO2-free carrier gas into the reaction cell within the Coulometer (Fig. 3.2). Eventually, the CO2 was measured using absolute coulometric titration. Since this method only measures the carbon from the calcium carbonate, the carbonate content had to be calculated using the following equation: CaCO3 (wt%) = TIC (wt%) * 8.33 (Bertrand et al., 2012b). This equation is based on the assumption that 100% of the measured CO2 is derived from dissolution of calcium carbonate (Bertrand et al., 2012b). The standard deviation, determined from three entirely separate analyses of sediment sample ST25, was 0.03 wt% CaCO3.
Fig. 3.2. Schematic overview of the UIC CM5017 coulometer.
28
3.2.7 Biogenic silica
The biogenic silica (bio-Si) content of sediments can be used to chemically estimate the siliceous microfossil abundance (Conley & Schelske, 2001). Siliceous microfossils consist primarily of diatoms and to a lesser extent of radiolarians, sponge spicules, silicoflagellates, chrysophyte and phytoliths (Fiers, 2016; Rebolledo et al., 2015). Bio-Si can provide an index for diatom abundance and productivity (Conley & Schelske, 2001).
Biogenic silica concentration was measured using the alkaline extraction technique (Fiers, 2016). The procedure corresponds to the analysis described by Bertrand et al. (2012b; 2014), which in turn was adapted from Carter & Colman (1994) and Mortlock & Froelich (1989). Prior to analysis, all centrifuge tubes and pipette tips were washed with nitric acid (5% HNO3) in order to avoid leaching of elements from the plastic material. First, 50 mg of precisely weighted (Mettler-Toledo XSE204) freeze-dried sediment was placed in 50 mL centrifuge tubes. Samples E3 and E8 were not analyzed since too little material was available. However, biogenic silica data of these samples was provided by Rebolledo et al. (in prep.). 5 mL of 10% H2O2 and 5 mL of 1N HCl were added to remove organic matter and carbonates, respectively. These chemicals aid in disaggregating the sediment and exposing the opal surfaces to dissolution (Mortlock & Froelich, 1989). When the reactions ended, 20 mL of deionised water (Milli-Q water) was added in each centrifuge tube. The tubes were agitated using a vortex mixer and centrifuged at 2400 rpm for 5 minutes, then shaken and re-centrifuged for 10 minutes. Afterwards, the supernatant was decanted in order to remove residual acid and peroxide (Mortlock & Froelich, 1989). During a few days, the solid samples dried under an infra-red lamp. Next, 30 mL of 0.2N NaOH was added to the dry sediments. The samples were agitated and sonicated for 5 minutes. This procedure was repeated five times to ensure full disaggregation of the sediment. Next, the centrifuge tubes were placed in a heated water bath at 95°C for 5 hours and agitated after 2 and 4 hours. Subsequently, the samples were agitated and centrifuged (5 minutes at 2400 rpm) one last time. Finally, 5 mL was pipetted to new acid-washed centrifuge tubes containing 35 mL 5% HNO3. The samples were stored overnight and Si and Al were analyzed the next day using an ICP-OES Varian 720-ES.
The Al analysis was necessary to correct the measured Si concentrations for lithogenic Si: bio-Si = measured Si – 2 x Al (Bertrand et al., 2012b). The 2:1 ratio for Si:Al accounts for Si leached from volcanic glass and clay minerals (Bertrand et al., 2012b; Carter & Colman, 1994; Ohlendorf & Sturm, 2008). This correction assumes that all Al originates from the dissolution of lithogenic particles (Bertrand et al., 2012b;
Ohlendorf & Sturm, 2008). Biogenic opal content (bio-opal, SiO2•nH2O, wt.%) was obtained by multiplying the bio-Si values by 2.4 (Mortlock & Froelich, 1989). Accuracy and analytical precision were calculated from the duplicate analysis of sample ST7. The standard deviation is 0.15 wt%.
In addition to our measurements, Rebolledo et al. (in prep.) provided biogenic opal data from the Gutierrez 2013 research cruise (Appendix 8.1).
3.2.8 Lithogenic Particles
The four main constituents of the surface sediment samples are: 1) biogenic opal, 2) carbonates, 3) organic matter and 4) lithogenic particles (Bertrand et al., 2012b). Since the weight percentages of the biogenic opal, CaCO3 and organic matter were analyzed, the wt% of the lithogenic particles was calculated as 100 wt% – biogenic opal (wt%) – carbonates (wt%) – organic matter (wt%) (Bertrand et al., 2012b). No CaCO3 data was available for E3, so its wt% of lithogenic particles could not be calculated. The results are presented in Appendix 8.1.
29
3.2.9 Bulk inorganic geochemistry
All samples were prepared for inorganic geochemical analyses using the Li Metaborate fusion technique (Murray et al., 2000). The main advantage of this technique is a complete dissolution, allowing determination of all elements, including Si and the refractory elements (Murray et al., 2000). Ten major elements (Al, Ca, Fe, K, Mg, Mn, Na, P, Si and Ti) and three minor elements (Zr, Ba and Sr) were measured by Inductively Coupled Plasma-Atomic Emission Spectrometry (ICP-AES). The limit of detection was 0.01 wt% for the major elements and 10 ppm for the minor elements. The advantage of ICP-AES is that rapid and quantitative analysis of a variety of sample types can be conducted relatively easily with a single instrument (Murray et al., 2000).
Sample preparation consisted in mixing 200 ± 0.5 mg of Li-metaborate/Li-tetraborate (80:20 wt%) with 50 ± 0.5 mg of sediment. This mixture was placed in Pt:Au crucibles. The crucibles were placed in a muffle furnace for 12 min at 1050°C, in order to fuse the mixture. This resulted in the formation of glass beads within the crucibles. Subsequently, these glass beads cooled down for 2 min. The glass beads were detached from the crucibles, and poured into 50 mL glass beakers containing 25 mL solution of 5% HNO3. This solution was magnetically stirred to dissolve the glass bead. The glass beads were fully dissolved after 60 to 90 minutes. Afterwards, the solution was filtered and 5 mL was pipetted into a centrifuge tube containing 35 mL of 5% HNO3. The final dilution was 4000x, the exact dilution factor was calculated from the precise weight of sediment used for fusion.
The measured concentrations (wt% for the major elements and ppm for the minor elements) were corrected for instrumental drift using the concentration of a lab control sample (Geostandard WS-E) (Govindaraju et al., 1994), which was measured before and after every sample. A blank solution was prepared identically to the samples in order to matrix-match the blank with the samples. A critical aspect of the preparation is the matrix-matching between the lab control sample, blank and samples in terms of composition, total dissolved solids (TDS) and acid concentration of the solution (Murray et al., 2000). In addition to the samples, lab control sample and blank, the reference sediments PACS-2 and STSD-3 were analyzed. PACS-2 was selected because its geochemical composition matches the average composition of the fjord sediment samples (Bertrand et al., 2012b). The ICP-AES method measured the elemental concentration at various wavelengths (Appendix 8.5). The accuracy of these wavelengths was determined based on the measurements of PACS-2 and STSD-3. The precision of the wavelengths was determined as the rsd (relative standard deviation, %) of the 45 replicate measurements of the lab standard sample (Appendix 8.5). Eventually, per element certain wavelengths were chosen and averaged to represent the final result (Appendix 8.1 & 8.5). This decision was based on the best combination of accuracy and precision (Appendix 8.5).
Lithogenic silica (litho-Si, wt.%) was calculated by difference: litho-Si = Si – bio-Si (Bertrand et al., 2012b). The total amount of Si (Si) was measured by bulk inorganic geochemical analysis.
The results will not be discussed in terms of elemental concentrations (wt%), instead elemental log-ratios will be used (Weltje, 2012; Weltje & Tjallingii, 2008). In sedimentary geochemistry, the dilution of elemental concentrations (wt%) by organic and/or biogenic phases is often overcome by using elemental ratios (Bertrand et al., 2012b; Weltje, 2012). The use of elemental ratios has an important disadvantage: it is insufficient to permit rigorous statistical analysis of compositional data (Weltje & Tjallingii, 2008). Elemental ratios have the undesirable property of asymmetry, i.e., conclusions based on evaluation of A/B cannot be directly translated into equivalent statements about B/A (Weltje & Tjallingii, 2008). This implies that the results of statistical analysis of elemental ratios depend on arbitrary decisions, since there is no ‘law’ to suggest which element should be the numerator or the denominator (Weltje & Tjallingii, 2008). This problem can be solved by one additional step: compositions should be expressed in terms of logarithms of
30
ratios of component abundances (so-called log-ratios) (Weltje & Tjallingii, 2008). From a mathematical point of view, it does not matter whether to choose to analyze log(A/B) or log(B/A): their variances and their means are identical, apart from their sign (Weltje, 2012). Based on Bertrand et al. (2012b), it was chosen to use Al-based elemental log-ratios. The only disadvantage of using log-ratios is that one easily loses sense of the real elemental concentrations.
3.2.10 Statistical analyses
Statistical analyses, including Pearson correlation coefficients (r), two-tailed test of significance (p) and Principal Component Analysis (PCA) were conducted with IBM SPSS Statistics 24. Except where indicated, correlations with p < 0.01 were considered significant. The entire dataset, consisting of 28 variables, was used for PCA (Appendix 8.1). PCA was also performed on separate datasets for the Baker Channel, Martinez channel and the study area west of the SPI. In addition, correlation matrices were constructed. In order to evaluate the spatial variability of the investigated variables in function of terrestrial sediment supply, surface salinity (psu) and the distance (km) between the nearest river outlet or glacier front and each sample location were included in the statistical analyses. Distance was measured using Google Earth (Appendix 8.2).
31
4. Results
4.1 Physical properties
4.1.1 Magnetic Susceptibility
Since volume-specific MS is dimensionless, mass-specific MS has a dimension of m3/kg (SI-unit) (Dearing, 1994). The resulting mass-specific MS values were multiplied by 106 to facilitate the interpretation. The results are mapped in figure 4.1. Three samples reached high MS values: ST80, ST93 and E6 Draga have MS values of 4,265 x 10-6 m3/kg, 2,582 x 10-6 m3/kg and 1,590 x 10-6 m3/kg, respectively. These samples, which are all located towards the open ocean, were indicated as outliers (black dots) for mapping purposes (Fig. 4.1).
Within Baker Channel, an increasing trend from the outlet of the Pascua river towards Golfo de Penas can be observed (Fig. 4.1). This increasing trend towards Golfo de Penas is not present within the Martinez Channel, instead a decreasing trend is observed (Fig. 4.1). From the outlet of the Baker river throughout the Martinez Channel, mass-specific MS decreases (Fig. 4.1).
Observing a trend within Canal Messier and Canal Wide is more difficult due to the lower sampling density (Fig. 4.1). The outlets of the Bernardo and Pio XI glaciers are characterized by moderate mass-specific MS values (Fig. 4.1). Throughout Canal Wide, the mass- specific MS decreases towards the open ocean (Fig. 4.1). This decreasing gradient is not observed in Canal Messier (Fig. 4.1). The distal samples (ST83 and ST86) have similar, moderate mass- specific MS values as the glacier outlets. Fig. 4.1 Mass-Specific Magnetic Susceptibility. Outliers are indicated by black dots.
32
4.1.2 Grain Size and Ice Rafted Debris
The GSD results were grouped into 4 categories of GSD (Fig. 4.2). Category 1 contains one mode within the range of 1 to 10 µm. Category 2 is characterized by bimodal distributions with one mode above 10 µm, in addition to the mode between 1 and 10 µm. Category 3 also contains a mode between 1 and 10 µm, but it has a broader GSD towards coarser grain sizes, often reaching 1 mm. Category 4 contains all the samples with a more complex and coarser GSD, which does not fit into any of the other previous categories. As a consequence, it was chosen to determine two grain size parameters to describe all samples: 1) mean grain size (µm) of the fraction below 150 µm, i.e., the fraction that is mostly transported in suspension, and 2) proportion of grains larger than 150 µm (volume %). For the first grain size parameter, the mean was chosen instead of the mode since multiple samples had more than one mode in the grain size fraction below 150 µm. The purpose of the second grain size parameter was to represent the amount of IRD, i.e., grains larger than 150 µm. This approach was inaccurate since (1) the laser diffraction method uses an amount of sample that is too low to accurately represent IRD, (2) the measurement duration was too short to properly quantify very coarse particles, and (3) the upper limit of the Malvern Mastersizer 3000 is 1 mm. As a consequence, only the IRD values obtained by sieving will be used. Based on the different GSD categories (Fig. 4.2), the IRD fraction above 150 µm (wt%) was chosen instead of the IRD fraction above 63 µm (wt%). From now on, ‘IRD’ will refer to the ice rafted debris fraction above 150 µm (wt%). Grain size will be discussed as the mean grain size (µm) of the fraction below 150 µm. To this variable will be referred to as ‘mean IRD-free grain size’. These results are mapped figure 4.3. ST83, ST86 and ST93 are very coarse compared to the other samples. As a consequence, these samples were identified as outliers for mapping purposes (Fig. 4.3a).
Fig. 4.2. Four categories of grain size distribution identified in the analyzed surface sediment samples. Category 1 is represented by the GSD of sample ST11, category 2 by sample ST3, category 3 by sample ST30 and category 4 by sample E13.
The mean IRD-free grain size is low in both Baker and Martinez Channels, varying between 4 and 7 µm (Fig. 4.3a). Towards Golfo de Penas, higher values can be observed: 20.5 µm at E6 Draga and 11.7 µm at ST3 (Fig. 4.3a). Also, at the outlets of the Baker and Pascua rivers, larger grain sizes prevail: 23.3 µm at E13 and 14.7 µm at E1 (Fig. 4.3a). Within Canal Messier, an increasing grain size towards Golfo de Penas can be observed (Fig. 4.3a). The outlet of the Pio XI glacier is characterized by a high mean grain size of 13.6 µm (ST28), which decreases towards 7.66 µm at ST25, but increases again towards the open ocean at ST31 (7.98 µm) and further at ST80, ST83 and ST86 (Fig. 4.3a).
With respect to IRD, figure 4.3b displays very low values (< 1 wt%) of IRD in the Baker and Martinez Channels and in Golfo de Penas. Only three samples have IRD above 1 wt%: ST3 (1.48 wt%), ST96 (2.54 wt%) and ST97 (1.29 wt%) (Fig. 4.3b). Within the channels west of the SPI, only 2 samples have IRD below 1 wt%: ST25 and ST28 (Fig. 4.3b). The highest IRD contents can be observed in samples: ST19, ST20, ST30, ST80 and ST86 (Fig. 4.3b).
33
Fig. 4.3. Physical Properties: A) mean IRD-free grain size (µm), and B) ice rafted debris >150 µm (wt%). Outliers are indicated by black dots.
4.2 Sediment composition
4.2.1 Biogenic opal
Throughout the BMFC, bio-opal content increases towards Golfo de Penas (Fig. 4.4a). Low bio-opal contents are observed near the front of Jorge Montt glacier and river mouths (Fig. 4.4a).
Throughout Canal Messier, samples with a high biogenic opal content, in the order of ± 8 wt%, prevail (Fig. 4.4a). Within this channel, biogenic opal content increases little from ST18 (8.04 wt%) towards ST20 (8.69 wt%) (Fig. 4.4a). Only a small decrease in bio-opal content towards Bernardo glacier is observed (7.99 wt% at ST21B) (Fig. 4.4a). Throughout Canal Wide, the biogenic opal content of the surface sediments increases with distance from the glacier fronts (Fig. 4.4a). ST28 represent the outlet of the Pio XI glacier, having a biogenic opal content of 4.16 wt% (Fig. 4.4a). Estero Falcon is represented by a low biogenic opal content of 3.51 wt% at ST30 (Fig. 4.4a). ST23 has the highest biogenic opal content of all samples: 9.71 wt%.
4.2.2 Carbonates
From figure 4.4b, a first important observation can be made: except for five samples, the carbonate content is low, varying between 0.01 wt% and 2.42 wt%. The five samples that do not fit this observation are: ST80,
ST83, ST86, ST91 and E6 Draga, with values between 10 to 35 wt% CaCO3 (Appendix 8.1). These samples are identified as outliers for mapping purposes and are indicated by black dots in figure 4.4b.
34
Throughout the Baker Channel, the carbonate content in the surface sediment samples increases towards Golfo de Penas (Fig. 4.4b). From values as low as 0.02 wt% at the outlet of the Pascua river, towards values of 2.42 wt% at ST3 (Fig. 4.4b). A similar trend is observed within the Martinez Channel, with values of 0.01 wt% at the outlet of the Baker river to 1.14 wt% (ST4) at the confluence with the Baker Channel (Fig. 4.4b).
This increasing carbonate content towards Golfo de Penas can also be observed within Canal Messier
(Fig. 4.4b). Here, the CaCO3 content varies between 0.01 and 0.10 wt% (Appendix 8.1). Further south, it is more difficult to observe a trend (Fig. 4.4b). The carbonate content is already relatively high at the outlet of the Pio XI glacier: 0.13 wt% at ST28. The carbonate content decreases little towards ST25 (0.11 wt%). Higher carbonate contents can be observed in Canal wide at ST31 and in Estero Falcon at ST30 (Fig. 4.4b). As for the biogenic opal content, the carbonate content in ST23 is high (2.25 wt%) (Fig. 4.4b).
4.2.3 Organic matter
For samples ST80, ST83, ST86, ST91 and E6 Draga, the Corg values were incorrect (explained in section 4.3 ‘Bulk Organic Geochemistry’), so the OM content was calculated based on the correlation between
LOI550 and Corg (Appendix 8.3).
Throughout the BMFC, the OM content increases towards Golfo de Penas (Fig. 4.4c). Sample E13, located nearest to the outlet of the Pascua river, has an OM content of 1.72 wt%. This rapidly decreases towards 0.98 wt% in sample ST10 (Fig. 4.4c). At the outlet of the Baker river, a similar observation can be made: a rapidly decreasing OM content from E1 towards ST14 (Fig. 4.4c). In the Baker Channel from ST10 towards E8, the OM content remains relatively stable (Fig. 4.4c). From sample ST6 onwards, the OM content increases towards Golfo de Penas (Fig. 4.4c). Throughout the Martinez Channel, it is more difficult to observe an increasing trend towards Golfo de Penas. Here, the OM content varies between 0.37 wt% and 1.76 wt% (Fig. 4.4c).
In Canal Wide, an increasing OM content away from the glacier front can be observed from ST28 towards ST31 (Fig. 4.4c). The high OM contents of ST80 and ST83 contribute to this observation (Fig. 4.4c). At the outlet of Bernardo glacier, the OM content is 1.42 wt% in ST21B (Fig. 4.4c). Within Canal Messier, the OM content varies between 1.33 wt% (ST19) and 2.18 wt% (ST18) (Fig. 4.4c). This observation can indicate that OM content increases away from the glacier front. Sample ST23 has an OM content of 4.89 wt%.
4.2.4 Lithogenic Particles
Throughout the BMFC, the amount of lithogenic particles within the surface sediments is constantly above 90 wt% (Fig. 4.4d). At the outlet of the Baker river, a slightly higher content of lithogenic particles (97.16 wt% at E1 Draga and 96.54 wt% at ST12) compared to the Pascua river outlet (94.26 wt% at E13 and 94.96 wt% at ST10) can be observed (Fig. 4.4d). From E13 until ST7 in the Baker Channel, the lithogenic particle content remains above 94 wt% (Fig. 4.4d). Starting from ST7, the lithogenic content within the surface sediments decreases throughout the Baker Channel towards Golfo de Penas (84.82 wt% at ST3) (Fig. 4.4d). This is not observed in the Martinez Channel. Here, the lithogenic content remains above 92 wt% (Fig. 4.4d).
35
Fig. 4.4. Sediment Composition: A) Biogenic Opal (wt%), B) Carbonates (wt%), C) Organic Matter (wt%), and D) Lithogenic Particles (wt%). Outliers are indicated by black dots.
36
At the front of the Bernardo glacier, the lithogenic particles represent 90.58 wt% of sample ST21B. This decreases towards 89.69 wt% in sample ST18, located in Canal Messier (Fig. 4.4d). A similar observation can be made from the front of the Pio XI glacier towards Canal Wide (Fig. 4.4d). In ST28, the lithogenic particles represent 95.18 wt% of the sample. This decreases towards 92.30 wt% in ST31. The more distal samples (ST80, ST83, ST86, ST91) contain the least lithogenic particles (Fig. 4.4d). ST23 only contains 83.15 wt% of lithogenic particles, due to its high biogenic opal, carbonate and organic matter content (Fig. 4.4). 4.3 Bulk organic geochemistry
δ13C is plotted versus the atomic N/C ratio in figure 4.5. For five samples (ST80, ST83, ST86, ST91 and E6 Draga), the δ13C values lie outside the terrestrial-marine end-member range (Fig. 4.5), which is unlikely in this study area. These are the same samples for which high CaCO3 contents were observed (Fig. 4.4b). From this observation, it seems that the preparation procedure of the bulk organic geochemical analysis was insufficient to fully decarbonate such
CaCO3 rich samples. As a consequence, the bulk organic geochemical data of samples ST80, ST83, ST86, ST91 and E6 Draga is considered incorrect. Eventually, 31 13 samples have correct Fig. 4.5. δ C (‰) versus Atomic N/C ratio bulk organic geochemical data. This data is presented in Appendix 8.1 and mapped in figure 4.6.
4.3.1 Carbon isotopic composition
Throughout the BMFC, δ13C values increase towards Golfo de Penas (Fig. 4.6a). Samples located at the river outlets of the Pascua and Baker river (E1, E13, ST10 and ST12) show the lowest δ13C, approximating the terrestrial end-members described above (Fig. 4.6a). At ST3, a δ13C value of -19.92‰ is observed, approximating the marine end-members described above (Fig. 4.6a). Throughout Canal Messier and Canal Wide, carbon isotopic content increases towards the open ocean (Fig. 4.6a).
37
4.3.2 Nitrogen isotopic composition
Throughout the BMFC, δ15N values increase towards Golfo de Penas (Fig. 4.6b). At the outlets of the Pascua and Baker rivers, low δ15N values are observed (Fig. 4.6b) At ST3, a δ15N value of 9.26‰ is observed (Fig. 4.6b). Within the Martinez Channel, two samples do not fit this observation: E3 has a very low δ15N value of 2.82‰ and E4 has a very high δ15N value of 11.96‰ (Fig. 4.6b).
Throughout Canal Messier and Canal Wide, the nitrogen isotopic content increases towards the open ocean (Fig. 4.6b).
4.3.3 Atomic N/C ratio
Disregarding the incorrect values of samples ST80, ST83, ST86, ST91 and E6 Draga, figure 4.5 indicates a positive relation between δ13C and the atomic N/C ratio. However, over the entire study area, these variables are not significantly correlated (r = 0.407; p = 0.023).
The atomic N/C ratios range from 0.08 to 0.18, except sample E3 (Fig. 4.5). The majority of the samples have atomic N/C ratios outside the end-member ranges (Fig. 4.5).
In figure 4.6c, no clear terrestrial to marine trend can be observed. Sample E3 has an extreme low atomic N/C ratio (Fig. 4.5 & 4.6c), due to its low nitrogen content (Appendix 8.1).
4.3.4 Terrestrial and marine fractions of organic carbon
Throughout the BMFC, the proportion of terrestrial organic carbon decreases towards Golfo de Penas (Fig. 4.6d). As expected, the outlets of the Baker and Pascua rivers are represented by samples with a high proportion of terrestrial organic carbon, i.e., 80 to 95% (Fig. 4.6d).
Throughout Canal Messier and Canal Wide, the proportions of terrestrial organic carbon decrease towards the open ocean (Fig. 4.6d). The proportion of terrestrial organic carbon is lower at the outlets of the Bernardo and Pio XI glacier in comparison with the river outlets: 70.06% at ST21B and 57.63% at ST28 (Fig. 4.6d).
38
Fig. 4.6. Bulk organic geochemistry: A) δ13C (‰), B) δ15N (‰), C) Atomic N/C ratio, and D) Fraction of terrestrial organic carbon (%).
39
4.4 Bulk inorganic geochemistry
The results of the bulk inorganic geochemical analysis are presented in Appendix 8.1 and mapped in figure 4.7. The measured and certified values of PACS-2 and STSD-3 are presented in Appendix 8.5. First, the spatial distribution of Al (wt%) will be discussed, next the Al-based log-ratios of the lithophile elements (Fe, Ti, Zr, Si, litho-Si, Ca, Mg, Na, K, Ba and Sr), and last the Al-based log-ratios of the partly lithophile, partly siderophile elements (Mn and P).
Within the BMFC, Canal Messier and Canal Wide, Al (wt%) varies between 9 and 12 wt% (Fig. 4.7a). Lower Al concentration can be observed in the distal/coastal samples (ST80, ST83, ST86, ST81 and ST93) (Fig. 4.7a). From all the analyzed lithophile elements, Al (wt%) has the best correlation with the lithogenic fraction of the sediment over the entire study area (r = 0.804; p < 0.001).
Within the BMFC, log(Fe/Al) increases towards Golfo de Penas (Fig. 4.7b). Throughout Canal Messier and Canal Wide, log(Fe/Al) remains stable with a small increase towards the open ocean (Fig. 4.7b). Throughout the Baker Channel and Canal Messier, log(Ti/Al) increases towards the open ocean (Fig. 4.7c). In contrast to this observation, log(Ti/Al) decreases towards the open ocean throughout the Martinez Channel and Canal Wide (Fig. 4.7c). Within the BMFC, log(Zr/Al) decreases towards Golfo de Penas (Fig. 4.7d). From Pio XI glacier throughout Canal Wide, log(Zr/Al) decreases towards the open ocean (Fig. 4.7d). Within Canal Messier, no trend is observed (Fig. 4.7d).
Throughout the BMFC, log(Si/Al) and log(litho-Si/Al) are low (Fig. 4.7 e & f). Higher values can be observed near the mouths of the Baker and Pascua rivers, and in Golfo de Penas (Fig. 4.7 e & f). Towards the open ocean, both log-ratios have a small increasing trend in Canal Messier and a small decreasing trend in Canal Wide (Fig. 4.7 e & f). Higher values can be observed in the distal samples (ST80, ST83, ST86 and ST91) (Fig. 4.7 e & f). These trends are similar to what is observed for the mean IRD-free grain size (Fig. 4.3a). This similarity is confirmed by statistical analysis. Over the entire study area, log(Si/Al) is correlated with the mean IRD-free grain size (r = 0.932; p < 0.001) and log(litho-Si/Al) is correlated with the mean IRD-free grain size (r = 0.932; p < 0.001 ).
Within the BMFC, log(Ca/Al) increases towards Golfo de Penas throughout the Baker Channel, but remains stable within the Martinez Channel (Fig. 4.7g). In the study area west of the SPI, log(Ca/Al) is low in Canal Messier and Canal Wide and high in the distal/coastal samples (ST80, ST83, ST86 and ST91) (Fig. 4.7g). Throughout the BMFC, log(Mg/Al) increases towards Golfo de Penas (Fig. 4.7h). Throughout Canal Messier and Canal Wide, a small increase towards the open ocean is observed (Fig. 4.7h).
Throughout the BMFC, log(Na/Al) increases towards Golfo de Penas (Fig. 4.7i), however this trend is more complicated throughout the Baker Channel. Throughout Canal Messier and Canal Wide, log(Na/Al) increases towards the open ocean (Fig. 4.7i). Log(K/Al), log(Ba/Al) and log(Sr/Al) remain stable in most channels of both study areas (Fig. 4.7 j, k & l). Within the BMFC, log(K/Al) is higher than in the channels west of the SPI (Fig. 4.7j). In the distal samples west of the SPI, log(Ba/Al) is significantly lower and log(Sr/Al) higher compared to their values in Canal Messier and Canal Wide (Fig. 4.7 k & l).
Throughout the BMFC, log(Mn/Al) increases towards Golfo de Penas, although this trend is less clear throughout the Baker Channel (Fig. 4.7m). Throughout the BMFC, log(P/Al) increases towards Golfo de Penas (Fig. 4.7n). West of the SPI, log(Mn/Al) has an extreme high value in ST20 (Fig. 4.7m). Log(Mn/Al) remains stable throughout Canal Messier and increases towards the open ocean in Canal Wide, although low values are again observed in the distal samples (ST80, ST83, ST86) (Fig. 4.7m). Throughout the channels west of the SPI, log(P/Al) increases towards the open ocean (Fig. 4.7n).
40
41
42
43
Fig. 4.7. Bulk inorganic geochemistry: A) Al (wt%), B) log(Fe/Al), C) log(Ti/Al), D) log(Zr/Al), E) log(Si/Al), F) log(litho-Si/Al), G) log(Ca/Al), H) log(Mg/Al), I) log(Na/Al), J) log(K/Al), K) log(Ba/Al), L) log(Sr/Al), M) log(Mn/Al), and N) log(P/Al).
4.5 Principal Component Analysis
The four PCA biplots (entire study area, Baker Channel, Martinez Channel and the study area west of the 13 SPI) are presented in figure 4.8. On all PCA biplots, FM (fraction of marine organic carbon) and δ C plot nearly on the same spot (Fig. 4.8). This is due to the fact that FM was calculated by an equation only based on δ13C, which results in nearly identical loadings of these variables on both PCA axes. On the PCA biplots, the more marine samples and more terrestrial samples plot oppositely on the first (horizontal) PCA axis (F1) (Fig. 4.8). The more marine samples have positive loadings on F1, while the more terrestrial samples have negative loadings on F1 (Fig. 4.8). This is best expressed in the PCA biplot of the Martinez Channel (Fig. 4.8c). This observation becomes more complicated in the PCA biplot of the study area west of the SPI (Fig. 4.8d). Here, the marine to terrestrial distribution plots more diagonally between both PCA axes (F1 and F2) (Fig. 4.8d).
On the PCA biplot of the entire study area (Fig. 4.8a), the first two PCA axes account for 63.98% of the total variance (F1: 37.92%; F2: 26.06%), the third axis accounts for 17.19%. The first axis primarily 13 indicates the variance between: 1) distance, δ C, FM and log(Ca/Al), variables representing the more marine samples (positive F1 loadings), and 2) FT, Al (wt%) and the lithogenic fraction of the sediment, variables representing the more terrestrial samples (negative F1 loadings) (Fig. 4.8a). CaCO3, organic matter, δ15N, log(Na/Al) and log(P/Al) form a relative close cluster on the PCA biplot, and have high positive 15 13 loadings on F1 (> 0.720) (Fig. 4.8a). Log(Ca/Al), δ N, δ C and CaCO3 are indicative of an increased marine influence (Fig. 4.8a). Log(Zr/Al) has a strong, negative F2 loading of -0.944 (Fig. 4.8a).
44
On the PCA biplot of the Baker Channel (Fig. 4.8b), the first two PCA axes account for 90.62% of the total variance (F1: 71.36%; F2: 19.26%), the third axis accounts for 5.50%. The first axis primarily indicates the 13 15 variance between: 1) distance, mass-specific MS, bio-opal, δ C, δ N, FM, log(Ca/Al), log(Mg/Al), log(P/Al), log(Ti/Al) and log(Sr/Al), these variables have F1 loadings above 0.940 and represent the more marine samples (Fig. 4.8b), and 2) FT, the lithogenic fraction of the sediment, atomic N/C, log(Zr/Al), Al (wt%) and log(Ba/Al), these variables have F1 loadings below -0.800 and represent the more terrestrial samples (Fig. 13 15 4.8b). Distance, δ C, δ N, FM, and log(Mg/Al) form a relatively close cluster on the PCA biplot, and have high, positive F1 loadings (>0.950) (Fig. 4.8b). The mean IRD-free grain size has a strong, positive F2 loading of 0.966 (Fig. 4.8b).
Fig. 4.8. Principal component analysis (PCA) biplots showing the relationship between the 28 measured variables. A) PCA biplot of the entire study area, B) PCA biplot of the Baker Channel, C) PCA biplot of the Martinez Channel, and D) PCA biplot of the study area west of the SPI. Brown represents more terrestrial and blue more marine sediments. Ft = fraction of terrestrial organic carbon, Fm = fraction of marine organic carbon, and GrainSize = mean IRD-free grain size.
On the PCA biplot of the Martinez Channel (Fig. 4.8c), the first two PCA axes account for 87.52% of the total variance (F1: 71.81%; F2: 15.71%), the third axis accounts for 10.14%. The first axis primarily 13 15 indicates the variance between: 1) distance, bio-opal, δ C, δ N, atomic N/C, FM, log(Fe/Al) and log(K/Al), these variables have F1 loadings above 0.930 and represent the more marine samples (Fig. 4.8c), and 2) mass-specific MS, the lithogenic fraction of the sediment, FT, log(litho-Si/Al) and log(Zr/Al), these variables have high, negative F1 loadings of -0.910 or lower and represent the more terrestrial samples (Fig. 4.8c). Three variables have high F2 loadings: log(Ca/Al) (0.979), log(Sr/Al) (0.864) and log(Ba/Al) (-0.984) (Fig. 4.8c).
45
On the PCA biplot of the study area west of the SPI (Fig. 4.8d), the first two PCA axes account for 70.64% of the total variance (F1: 45.53%; F2: 25.11%), the third axis accounts for 14.24%. As mentioned above, the terrestrial to marine distribution is more complicated in this study area and does not plot clearly along the first PCA axis (F1) as in the other PCA biplots (Fig. 4.8d). This can be due the low sampling density in 13 the study area (Fig. 1.1). Distance, δ C, FM, CaCO3 and log(Na/Al) form a relatively close cluster, representing the more marine samples (Fig. 4.8d). These variables have positive, but moderate loadings (0.543 to 0.735) on both PCA axes (F1 and F2) (Fig. 4.8d). The more terrestrial samples are not represented by a dense cluster of variables (Fig. 4.8d). Al (wt%) (-0.929) and surface salinity (0.918) are the only variables with strong F2 loadings (Fig. 4.8d).
5. Discussion
5.1. Spatial variability
As stated in the introduction, the main research objective of this master dissertation is to evaluate the ability of certain geochemical and sedimentological properties to estimate past changes in terrestrial sediment supply, from river- and/or glacial meltwater discharge, within two different study areas: 1) the Baker Martinez Fjord Complex (BMFC), north of the SPI, and 2) the channels and fjords west of the SPI, with a focus on Canal Messier and Canal Wide (Fig. 1.1). In the case of glaciated fjords, terrestrial sediment supply can serve as an indicator for glacier variability. To evaluate the relations between the geochemical and sedimentological results and terrestrial sediment supply, a reference variable must be identified. Within the BMFC, surface salinity can be used as a reference since terrestrial input/river discharge clearly affects surface salinity (Davila et al., 2002). However, surface salinity data was obtained during October 2014 (austral spring). The river discharge graphs in figure 2.7a indicate that the mean discharge from Baker and Pascua rivers peaks during late austral summer (January, February and March), due to the snow and glacial melt maximum (Aiken, 2012; Vandekerkhove et al., 2016). Since the Baker and Pascua rivers are mainly fed by glacial meltwater, surface salinity data should be obtained during late summer to better represent freshwater and sediment input. However, salinity data from the studied sample locations obtained during austral summer is not available.
Figure 2.11 indicates that surface salinity increases throughout the BMFC and Canal Messier towards Golfo de Penas. Throughout Canal Wide, surface salinity decreases from the glacier front of the Pio XI glacier towards the open ocean (Fig. 2.11). This is probably due to the increase in precipitation towards the southwest in Canal Wide (Fig. 2.3c) and/or the salinity data being obtained outside the glacial melt season (Aiken, 2012; Vandekerkhove et al., 2016). This reasoning indicates that surface salinity data from October 2014 should not be used as a reference.
Another potential reference variable is the distance (km) between the nearest river outlet or glacier front and each sample location. With distance away from the river mouth or glacier front, the thickness of the freshwater lens, resulting from river/glacier discharge, and terrestrial sediment supply decrease. In general, on a transect from a river outlet or glacier front towards the open ocean, terrestrial influence decreases and marine influence increases. An example is the decrease in the fraction of terrestrial organic carbon and the increase in the fraction of marine organic carbon towards the open ocean (Fig. 4.6d). This is supported by the PCA biplots (Fig. 4.8), visualizing the control of distance on the investigated properties. In these PCA biplots, distance has a high positive F1 loading (> 0.882) (Fig. 4.8). Here, distance is closely related to 13 variables representing the more marine samples (e.g. δ C, FM and log(Ca/Al)). Surface salinity has consistently lower F1 loading than distance, thus less representing the more marine samples (Fig. 4.8). Within the PCA biplot of the study area west of the SPI, distance has a lower F1 loading (0.617). However, 13 distance still clusters with variables representing the more marine samples (e.g. δ C, FM and CaCO3), while surface salinity reaches a negative F1 loading (-0.323) (Fig. 4.8). The PCA biplots (Fig. 4.8) indicate
46
that high distance values represent more marine samples, while low distance values represent more terrestrial samples. Therefore, it was chosen to use distance as a reference to evaluate the spatial variability of the investigated properties in function of the terrestrial sediment supply.
5.1.1 Grain Size
Within the BMFC, mean IRD-free grain size results indicate the presence of relatively coarse sediments near the outlets of the Baker and Pascua rivers, while finer sediments prevail throughout the Baker and Martinez Channels (Fig. 4.3a). The coarse sediment near the river outlets can originate from the rapid deposition of the bedload and coarse particles from the suspended load which have high settling velocities (Syvitski & Shaw, 1995). The finer sediments within the Baker and Martinez Channels originate from the surface river plume (Syvitski et al., 1987; Syvitski & Shaw, 1995). At the outer part of the BMFC, near Golfo de Penas, again coarser sediments are observed (Fig. 4.3a). This can be explained by the presence of currents strong enough to carry clay and fine silt-sized particles away and leaving behind a lag deposit of coarser material (Faust et al., 2014; Munoz & Wellner, 2016). As such, it is likely that these deposits are the result of winnowing of fine grains and sediment re-suspension (Faust et al., 2014; Munoz & Wellner, 2016).
The fine grained sediments within Canal Messier and Canal Wide (Fig. 4.3a) are interpreted as the fine glacial clays and silts deposited from the glacial meltwater plumes originating at the glacier fronts. The mean IRD-free grain size data indicate that the distal samples (ST80, ST83, ST86 and ST91) are coarser than the samples within Canal Messier and Canal Wide (Fig. 4.3a). Again, this can be explained by the winnowing of fine grains and sediment re-suspension due to strong currents (Faust et al., 2014; Munoz & Wellner, 2016) and by the presence of IRD (Fig. 4.3b).
5.1.2 Sediment Composition
At the outlets of the Baker and Pascua rivers, the organic matter content in the surface sediments rapidly decreases with distance from the river mouths (Fig. 4.4c & 5.1). This is due to the rapid deposition of terrestrial organic matter transported by the rivers (Aracena et al., 2011; Faust et al., 2014). This hypothesis is confirmed by the calculation of the terrestrial fraction of organic carbon (FT) (Fig. 4.6d). Further down the Baker and Martinez Channels, the organic matter content increases towards Golfo de Penas (Fig. 5.1). Within Canal Messier and Canal Wide, the organic matter content also increases towards the open ocean (Fig. 4.4c & 5.1). These increasing gradients are due to less dilution by the lithogenic fraction and/or by the increase in marine productivity.
The increase of primary productivity towards the open ocean is confirmed by the results of Aracena et al. (2011). The low productivity near the river mouths and glacier fronts can be due to the high suspended sediment load in the river and glacial meltwater discharge, limiting light penetration (Aracena et al., 2011; Munoz & Wellner, 2016) and creating a high energy environment in which turbulence is high. Towards the open ocean, the turbidity of the surface water layer decreases and its salinity increases (Fig. 2.10 & 2.11) (Syvitski et al., 1987; Syvitski & Shaw, 1995), which promotes marine productivity.
Within the BMFC, Canal Wide and Jorge Montt Fjord, the biogenic opal content increases with distance from the river mouth or glacier front (Fig. 4.4a & 5.1). These increasing gradients towards the open ocean can be linked to less dilution by the lithogenic particles (Fig 5.1) and/or to an increase in siliceous organisms productivity (Bertrand et al., 2012b). In order to decipher these two options, bio-opal accumulation rates within these fjords should be known to quantify productivity. Within Canal Messier, the biogenic opal content remains high and does not increase towards the open ocean (Fig. 4.4a). Within this channel, the lithogenic fraction of the surface sediments varies little, with a maximum difference of 1.29
47
wt% between ST21B and ST20 (Appendix 8.1). This indicates a possible covariance between bio-opal content and lithogenic fraction of the sediments through Canal Messier. Another explanation can be linked to primary productivity, for which accumulation rates must be known. Figure 5.1 shows a decrease in bio- opal towards the open ocean within the study area west of the SPI, which does not agree with the increase throughout Canal Wide (Fig. 4.4a). Probably, the trend in figure 5.1 is due to the low bio-opal content in two distal samples (ST80 and ST86) (Fig. 4.4a). A higher sampling density within this study area could reveal a more clear trend.
Over the entire study area, a general increase towards the open ocean in carbonate content within the surface sediments can be observed (Fig. 5.1). Again, this can be explained by the decreasing dilution effect of the lithogenic fraction of the sediment (Fig. 5.1) and/or an increase in carbonate marine productivity towards the open ocean (Bertrand et al., 2012b; Faust et al., 2014). To decipher these options, carbonate accumulation rates within the studied fjords should be known. The extremely high CaCO3 content of some samples (ST80, ST83, ST86, ST91 and E6 Draga) can be explained by the presence of dolomite. However, this hypothesis is not confirmed by the geological map (SERNAGEOMIN, 2003). The high 13 CaCO3 content in ST80 can be caused by marble contamination (SERNAGEOMIN, 2003). The high δ C values of these CaCO3 rich samples (Fig. 4.4b & 4.5) indicates the presence of a carbonate constituent with a marine origin, which was not removed by the sulphurous acid during the preparation procedure of the bulk organic geochemical analysis.
The lithogenic fraction of the surface sediments reaches the fjords through rivers and glaciers (Aracena et al., 2011). Within the fjords, the lithogenic particles are transported by river plumes or glacial meltwater plumes (Mugford & Dowdeswell, 2011; Syvitski et al., 1987; Syvitski & Shaw, 1995). On all four PCA biplots, the lithogenic fraction plots high and negative on the F1 axis, representing the more terrestrial samples (Fig. 4.8). Distance plots high and positive on the F1 axis, representing the more marine samples (Fig. 4.8). This indicates that the lithogenic fraction is inversely correlated with distance, which is also visualized in figure 5.1. Both variables are negatively correlated in the Baker Channel (r = -0.689; p = 0.013), Martinez Channel (r = -0.924; p < 0.001) and in the fjords west of the SPI (r = -0.603; p = 0.029). Although, the lithogenic fraction of the sediment reaches the fjords through rivers and glaciers (Aracena et al., 2011) and its concentration decreases towards the open ocean (Fig. 5.1), its concentration is also affected by primary productivity. This means that the lithogenic fraction cannot be used to estimate past changes in terrestrial sediment supply. However changes in the composition of the lithogenic fraction, i.e., log-ratios of lithophile chemical elements potentially can.
48
Fig. 5.1. Spatial variability of the bulk sediment composition, expressed as weight percentages
(wt%) of organic matter (calculated as 2.2 x Corg), carbonate, biogenic opal and lithogenic fraction. Within the BMFC, samples with carbonate content > 1 wt% are not included for representation purposes.
49
5.1.3 Bulk organic Geochemistry
Within all four PCA biplots (Fig. 4.8), δ13C is closely related to distance. This relation is visualized in figure 5.2. More depleted values can be observed near the river mouths and glacier fronts, and less depleted values towards the open ocean (Fig. 5.2). The correlation matrices (Appendix 8.4) confirm the significant correlation between δ13C and distance (Fig. 5.2). δ13C and distance are correlated within the Baker Channel (r = 0.978; p < 0.001), Martinez Channel (r = 0.966; p < 0.001) and in the fjords west of the SPI (r = 0.936; p < 0.001). The δ13C results approach the terrestrial and marine end-members defined by Bertrand et al. (2012b) and Sepúlveda et al. (2011) (Fig. 4.5). In both study areas, the FT results show a decreasing proportion of terrestrial organic carbon, within the fjord surface sediments, towards the open ocean (Fig. 4.6d). A similar observation is described by Aracena et al. (2011). They observed a west-east gradient in terms of δ13C values in the Chilean Patagonian fjord region (41–56°S); more depleted δ13C values (-26‰) correspond to areas close to rivers and glaciers (Aracena et al., 2011). The results of Aracena et al. (2011) show that δ13C decreases towards the fjord heads; at these sites the main input of organic matter comes from fluvial discharge, whereas at sites with an oceanic influence, the main source is marine. The Chilean Fig. 5.2. Spatial variability of the bulk organic geochemistry, expressed as fjord system receives organic δ13C and δ15N, over the entire study area. matter input in the form of marine primary production and land-plant remains carried from the surrounding temperate evergreen rainforest by rivers and overland run-off (Aracen et al., 2011; Sepúlveda et al., 2011). Silva & Prego (2002) state that the edaphic input doesn’t play an important role. The δ13C results of this study indicate the potential of the carbon isotopic composition as a proxy to reconstruct past changes in terrestrial sediment supply towards Chilean fjords (Fig. 5.5).
Over the entire study area, δ15N has a similar gradient as δ13C: increasing values towards the open ocean (Fig. 5.2). As stated by Meyers & Teranes (2001) and Sepúlveda et al. (2011), δ15N is a proxy, similar to δ13C, for identifying sources of organic matter. Two important nitrogen sources are coastal marine plankton 15 and C3 land plants, which have δ N values approximating +8.5‰ and +0.5‰, respectively (Meyers & Teranes, 2001). Although, no direct relation between δ15N and distance can be observed within the PCA biplots (Fig. 4.8), the correlation matrices (Appendix 8.4) reveal the significant correlation between δ15N and distance. δ15N and distance are correlated within the Baker Channel (r = 0.982; p < 0.001), Martinez Channel (r = 0.873; p = 0.001) and in the fjords west of the SPI (r = 0.774; p = 0.014).
50
δ13C is better correlated with distance compared to δ15N, because the nitrogen biogeochemical cycle is more complicated than the carbon cycle (Meyers & Teranes, 2001). Thus, care should be taken when interpreting the sedimentary δ15N record (Meyers & Teranes, 2001). It is not recommended to use the δ15N record as a stand-alone proxy, better is to use it in multi-proxy based studies to evaluate the δ13C signal.
Figure 4.6d illustrates the decrease in the proportion of terrestrial OM and increase in marine OM proportion towards the open ocean. Within both study areas, δ13C and δ15N can be used to estimate past changes in the terrestrial sediment supply (Fig. 5.5.). Based on the δ13C and δ15N results of this study, δ13C has a greater potential as a proxy of terrestrial sediment supply.
5.1.4 Bulk inorganic Geochemistry
Based on Bertrand et al. (2012b), it was chosen to use Al (wt%) as a normalizer in the elemental log-ratios. This choice is justified since our results confirm Al (wt%) has the best correlation with the lithogenic fraction of the sediment (r = 0.804; p < 0.001) of all analyzed lithophile elements. In addition, it is common that Al is overwhelmingly of detrital origin and immobile during diagenesis (Tribovillardet al., 2006).
Over the entire study area, log(Ca/Al) is significantly positively correlated with CaCO3 (wt%) (r = 0.915; p< 0.001). This correlation is due to the general increase in carbonate content towards the open ocean within the surface sediments (Fig. 5.1). Since the carbonate content is controlled by dilution by the lithogenic fraction of the sediment and/or marine carbonate productivity, log(Ca/Al) cannot be used as a proxy to reconstruct past changes in terrestrial sediment supply.
Over the entire study area, the bio-opal content (wt%) and log(Si/Al) are not significantly correlated (r = - 0.379; p = 0.023), while log(litho-Si) and log(Si/Al) are (r = 0.972; p < 0.001) (Fig. 4.8). This indicates that the log(Si/Al) signal is primarily controlled by the lithogenic silica content of the surface sediments. In addition, both log(litho/Si) and log(Si/Al) are significantly positively correlated with the mean IRD-free grain size and with IRD (Appendix 8.4, table 8.4.1). This indicates that log(litho-Si/Al) and log(Si/Al) can be used as proxies for grain size variations. However, these proxies cannot be used to distinguish ice rafted debris from bulk grain size. Log(litho-Si/Al) has a slightly better correlation with IRD compared to log(Si/Al) (Appendix 8.4, table 8.4.1). Log(Si/Al) and log(litho-Si/Al) have the same correlation with the mean IRD-free grain size (r = 0.932; p < 0.001) (Appendix 8.4, table 8.4.1). Since log(Si/Al) is influenced by the bio-opal content and grain size is measured on the lithogenic fraction of the sediment, log(litho-Si/Al) is a better proxy for grain size. The relation of log(Si/Al) and log(litho-Si/Al) with grain size is interpreted as quartz being concentrated in the coarse fraction of the sediment (Bertrand et al., 2012b). However, this hypothesis should be investigated by a mineralogical analysis of the surface sediment samples.
Over the entire study area, log(P/Al) is significantly negatively correlated with the Al content (wt%) (r = - 0.664; p < 0.001). This indicates that P has a non-detrital source in the studied fjord systems (Tribovillard et al., 2006). A significant positive correlation is present between log(P/Al) and the organic matter content (OM, wt%) (r = 0.722; p < 0.001). This indicates that log(P/Al) might be used as a proxy for OM abundance and primary productivity (Tribovillard et al., 2006). Although the distribution of P in sediments is linked to OM supply, possibly resulting from primary productivity, the use of P as a proxy is not straightforward (Tribovillard et al., 2006). Phosphorous can escape from the sediment by bacterial reactions consuming 3- sedimentary OM (Tribovillard et al., 2006). This reaction regenerates organically bound P to aqueous PO4 (Tribovillard et al., 2006). Under anoxic conditions, P then generally diffuses upward from the sediment and returns to the water column (Tribovillard et al., 2006). Under certain conditions, P released to pore water can reach relatively high concentrations and authigenic phases, i.e., apatite, can precipitate (Tribovillard et al., 2006). Due to these P escape and trapping processes, P is not a reliable productivity proxy (Tribovillard et al., 2006).
51
Some elemental log-ratios (e.g. log(Ti/Al), log(K/Al) and log(Zr/Al)) have different terrestrial to marine gradients throughout different channels (Fig. 4.7 & 4.8). This indicates that the spatial variability signal of certain elemental log-ratios differs between fjords with different terrestrial sources, i.e., calving glaciers or rivers. As a consequence, the ability of elemental log-ratios to reconstruct past changes in terrestrial sediment supply will be analyzed for each channel individually.
Baker Channel Within the Baker Channel, four elemental log-ratios have a significant correlation (absolute value of r > 0.810; p ≤ 0.001; Appendix 8.4) with distance: log(Fe/Al), log(Ti/Al), log(Mg/Al) and log(Zr/Al) (Fig. 5.3a). For log(Ti/Al) and log(Mg/Al) the relation with distance is best described by a positive linear function (Fig. 5.3a). The variation of log(Zr/Al) in function of distance is best described by a decreasing logarithmic function, and for log(Fe/Al) by an increasing logarithmic function (Fig. 5.3a). However, one outlier (sample ‘E6 Draga’) does not fit in the logarithmic model of log(Fe/Al) (Fig. 5.3a).
Martinez Channel Within the Martinez Channel, five elemental log-ratios have a significant correlation (absolute value of r > 0.820; p ≤ 0.001; Appendix 8.4) with distance: log(Fe/Al), log(K/Al), log(Mg/Al), log(Zr/Al) and log(litho-Si/Al) (Fig. 5.3b). In the case of log(K/Al), the relation with distance is best described by a positive linear function (Fig. 5.3b). While for log(Fe/Al), log(Zr/Al) and log(litho-Si/Al), the variations with distance are best described by logarithmic functions (Fig. 5.3b). In the case of log(Mg/Al), a logarithmic function can describe its variation in function of distance. However two outliers (samples ‘Isla Irene’ and ‘E4’) do not fit in this model (Fig. 5.3b). This indicates that the relation between log(Mg/Al) and distance becomes more complicated from the middle fjord environment towards the outer fjord. Within the inner fjord environment, the logarithmic model clearly represents the variations of log(Mg/Al) in function of distance (Fig. 5.3b).
Since log(litho-Si/Al) is correlated with the mean IRD-free grain size (r = 0.928; p < 0.001), the inverse relation between log(litho-Si/Al) and distance is controlled by the decrease in mean IRD-free grain size with distance from the Baker river (Fig. 5.3b).
Within the Martinez Channel, five samples at different distances from the Baker river have similar log(K/Al) values (-0.54 to -0.55) (Fig. 5.3b). This observation indicates that log(K/Al) cannot be used to reconstruct terrestrial sediment supply.
Within the Martinez Channel, log(Ti/Al) has a small decreasing trend towards Golfo de Penas (Fig. 4.7c). However, log(Ti/Al) is not correlated with distance (r = -0.590; p =0.072). Within the Baker Channel, log(Ti/Al) increases towards Golfo de Penas (Fig. 4.7c) and is significantly positively correlated with distance (Fig. 5.3a). Such a difference in spatial variability between different channels confirms that the ability of inorganic geochemical properties to reconstruct terrestrial sediment supply must be evaluated within each channel individually.
52
Fig. 5.3. Spatial variability of the bulk inorganic geochemistry. A) Baker Channel, and B) Martinez Channel.
53
Canal Messier and Canal Wide Within the channels west of the SPI, six elemental log-ratios have a significant correlation (absolute value of r > 0.740 and p ≤ 0.004) with distance: log(K/Al), log(Ba/Al), log(Sr/Al), log(Si/Al), log(litho- Si/Al) and log(Na/Al) (Fig. 5.4).
The correlations of log(K/Al), log(Ba/Al) and log(Sr/Al) with distance are mainly due to the significant difference between the proximal and distal samples (Fig. 5.4). For Ba and Sr, this is likely due to the presence of carbonate-rich samples in the distal environments (Fig. 4.4b). Here, Ba and Sr can substitute Ca in the carbonates. These elemental log-ratios remain stable throughout Canal Messier and Canal Wide (Fig. 4.7 j, k & l). As a consequence, these elemental log-ratios are not fit to reconstruct terrestrial sediment supply.
Again, a difference in spatial variability of one certain elemental log- ratio between different channels can be observed. Log(K/Al) has a negative correlation with distance in the study area west of the SPI (Fig. 5.4), while a positive correlation is observed in the Martinez Channel (Fig. 5.3). Log(Ti/Al) increases towards Golfo de Penas throughout Canal Messier, while it decreases towards the open ocean in Canal Wide (Fig. 4.7c).
West of the SPI, log(litho-Si/Al) has a positive correlation with distance (Fig. 5.4). This correlation is mainly due to the relative high values in the distal samples and increase towards Golfo de Penas in Canal Messier (Fig. 4.7f). In contrast, log(litho-Si/Al) decreases towards the open ocean in Canal Wide (Fig. 4.7f). West of the SPI, log(litho-Si/Al) reflects the variations in mean IRD-free grain size (r = 0.940; p < 0.001) and IRD content (r = 0.817; p = 0.001).
These results indicate the complexity of terrestrial sediment supply towards the channels and fjords west of the SPI. The spatial variability of elemental log-ratios is significantly different between channels with different calving glaciers as sources of terrestrial material. As a consequence, no elemental log-ratios can be identified to reconstruct terrestrial sediment supply towards the entire area west of the SPI. In addition, the sampling density in this study area is too low to evaluate the ability of elemental log-ratios as a proxy of terrestrial sediment supply within every channel individually (Fig. 5.5). Throughout Canal Messier and Canal Wide, the results of log(Fe/Al), log(Ti/Al), log(Zr/Al) and log(Mg/Al) show increasing/decreasing trends towards the open ocean (Fig. 4.7 b, c, d & h). Although the absence of any correlation with distance, these results indicate the potential of elemental log-ratios as proxies of terrestrial sediment supply towards the channels and fjords west of the SPI individually. In order to unravel the complexity of terrestrial Fig. 5.4. Spatial variability of the bulk sediment supply towards the fjords west of the SPI, future inorganic geochemistry in the channels and fjords west of the SPI. research should focus on one specific channel/fjord and high sampling density.
54
The PCA biplot of the study area west of the SPI indicates a significant relation between log(Na/Al) and distance (Fig. 4.8d & 5.4). Throughout Canal Messier and Canal Wide, log(Na/Al) increases towards the open ocean (Fig. 4.7i & 5.4). This observation is confirmed by the high values of log(Na/Al) in the distal samples (Fig. 4.7i & 5.4). The positive relation between log(Na/Al) and distance is confirmed by the correlation matrix of the study area west of the SPI (r = 0.743; p = 0.004). This increase in log(Na/Al) reflects the increasing salinity towards the open ocean. For each sample, log(Na/Al) represents the amount of saline water in the sample before freeze drying it.
Inorganic geochemical proxies of terrestrial sediment supply Log(Fe/Al) and log(Zr/Al) are correlated with distance and have similar trends throughout the Baker and Martinez Channels (Fig. 5.3). This indicates that these Al-based elemental log-ratios can be used as proxies to estimate past changes in terrestrial sediment supply throughout the BMFC (Fig. 5.5). Log(Ti/Al) can be used as an additional inorganic geochemical proxy throughout the Baker Channel, but not throughout the Martinez Channel (Fig. 5.5). In the fjords north of the NPI, the use of Fe/Al, Ti/Al and Zr/Al, all lithophile elements, to estimate past changes in terrestrial sediment supply has been suggested by Bertrand et al. (2012b).
Throughout both the Baker and Martinez Channels, log(Mg/Al) increases with distance from the river mouths (Fig. 5.3). However, this increasing trend is slightly different between both channels (Fig. 5.3). Within the Baker Channel, the relation between log(Mg/Al) and distance is best expressed with a positive linear model, while in the Martinez Channel by a positive logarithmic model (Fig. 5.3). This difference indicates care should be taken when interpreting the inorganic geochemical proxy record within the BMFC. Consequently, it is recommended to interpret the records of the Baker and Martinez Channels separately. Notwithstanding, log(Mg/Al) can be used as a proxy to estimate past changes in terrestrial sediment supply throughout the Baker and Martinez Channels (Fig. 5.5).
In order to use these elemental log-ratios as proxies of terrestrial sediment supply, their provenance must be identified based on a mineralogical analysis of the surface sediment samples.
In general, iron is associated with relatively dense mafic minerals (amphibole, pyroxene and olivine), although it also occurs in minor proportions in a large series of minerals, including clay minerals (Monroe & Wicander, 2009). Within the fjords north of the NPI, the distribution of Fe is driven by the concentration of mafic minerals in the mid-fjords, but Fe is also concentrated in the silt fraction of the sediment (Bertrand et al., 2012b). Data from Bertrand et al. (2012b) demonstrate that Fe and Mg in the Chilean fjords are associated with sediment particles of low to intermediate grain size and intermediate density. Within the Martinez Channel, the distribution of Fe and Mg is controlled by grain size. Here, Fe and Mg are associated with fine grained sediments, since both Al-based elemental log-ratios are inversely correlated with the mean IRD-free grain size (Fe: r = -0.964; p < 0.001; Mg: r = -0.847; p = 0.002). Within the Baker Channel, log(Fe/Al) and log(Mg/Al) are not correlated with the mean IRD-free grain size (Appendix 8.4), their provenance is discussed is section 5.3. The low concentrations of Fe and Mg in some of the most proximal sites most likely reflect dilution by coarser and/or denser minerals such as quartz (Bertrand et al., 2012b). The occurrence of Fe in open marine samples is likely due to its presence in very fine-grained mafic minerals, in plagioclase and in most clay minerals (Bertrand et al., 2012b).
Within the Martinez Channel, log(Zr/Al) is correlated with the mean IRD-free grain size (r = 0.983; p < 0.001). Only a moderate correlation between log(Zr/Al) and the mean IRD-free grain size is observed in the Baker Channel (r = 0.651; p = 0.022). Zirconium is most often associated with zircon, which is a typical accessory mineral of the North Patagonian Batholith (Bertrand et al., 2012b). Zircon is a very dense and refractory mineral, which results in Zr being concentrated in the densest and coarsest mineralogical fraction of sediments (Bertrand et al., 2012b).
55
Fig. 5.5. Schematic representation of the two main study areas: 1)Baker Martinez Fjord complex, and 2) Channels and fjords west of the SPI (dashed red rectangles), and the spatial distribution of the bulk organic (δ13C and δ15N) and inorganic geochemical proxies (log(Fe/Al), log(Zr/Al), log(Ti/Al), log(Mg/Al) and log(litho-Si/Al)) of terrestrial sediment supply, in relation to distance (km) towards the nearest river mouth or glacier front. This illustration demonstrates that the signal of some Al-based elemental log-ratios (e.g. log(Mg/Al) and log(Ti/Al)) might differ between channels with different sources of terrestrial material, i.e., calving glaciers or rivers. As a consequence, the study area of the BMFC is subdivided into the Baker and Martinez Channels (dashed green rectangle). Sample locations (Cimar 20 and Copas 2014 research cruises) illustrate the low sampling density in the channels and fjords west of the SPI. As a result, no elemental log- ratios can be identified to reconstruct terrestrial sediment supply towards the entire area west of the SPI, or towards Canal Messier and Canal Wide individually.
56
Within the Baker Channel, log(Ti/Al) is not correlated with the mean IRD-free grain size (r = -0.232; p = 0.469). However a moderate correlation with the lithogenic fraction of the sediment is observed (r = -0.629; p = 0.028). Titanium occurs in most mafic minerals (amphibole, pyroxene and olivine) and in ilmenite, in association with iron, and it is a main constituent of less frequent and relatively dense iron-free minerals such as rutile and titanite (McLennan et al., 2003; Nesbitt & Young, 1996; Verhoogen, 1962). Ti is therefore associated with minerals of refractoriness (i.e., grain size) and density similar or higher than Fe-bearing minerals, but lower than Zr-bearing minerals (Bertrand et al., 2012b). As a consequence, Ti is distributed similar to the distribution of Fe, although Ti is transported on smaller distances than the average Fe-bearing minerals (Bertrand et al., 2012b). Ti is associated with minerals that are transported over longer distances than zircon (Bertrand et al., 2012b).
Throughout the Martinez Channel, the log(litho-Si/Al) signal is controlled by the mean IRD-free grain size (r = 0.928; p < 0.001). As a result, log(litho-Si/A) can be used as a proxy to estimate past changes in terrestrial sediment supply throughout the Martinez Channel (Fig. 5.5).
5.2. Glaciated fjords versus non glaciated fjords
A near absence of IRD can be observed within the Baker and Martinez Channels (Fig. 4.3b), which are channels containing no calving glaciers. The BMFC receives small amounts of glacial meltwater from the Jorge Montt glacier (Aiken, 2012), which is a calving glacier. Jorge Montt glacier is connected to the BMFC by the Baker Channel and Jorge Montt Fjord (Fig. 1.1). Within Jorge Montt Fjord, icebergs and IRD have been observed (De Wilde, 2016; Rivera et al., 2012). Probably, IRD does not reach the Baker Channel due to the presence of two shallow sills within Jorge Montt fjord (Moffat, 2014; Rivera et al., 2012) and the decreasing amount of IRD with distance from the glacier (De Wilde, 2016).
The presence of a small amount of IRD in ST3 (Fig. 4.3b) can be explained by the winnowing of fine grains and sediment re-suspension due to strong currents (Faust et al., 2014; Munoz & Wellner, 2016). The IRD content of ST96 and ST97 (Fig. 4.3b) has no glacial source since no calving glaciers are present within their watershed. Their IRD content can be explained by local soil slide events or currents reworking the fjord-floor deposits.
Within the second study area, west of the SPI, significant amounts of IRD were observed (Fig. 4.3b). This is due to the presence of calving glaciers within these fjords. Samples ST80 and ST86 have extreme amounts of IRD compared to the other samples (Fig. 4.3b). Probably, icebergs reached further towards the ocean during the Little Ice Age, bringing IRD to the coastal area. The presence of strong currents in this coastal area can prevent the burial of the IRD, which explains this extreme IRD observation.
An important difference between fjords with calving glaciers and fjords without calving glaciers is the presence and absence of icebergs, respectively. Icebergs result from the calving activity of calving glaciers and are the primary transport agent for IRD (Kuijpers et al., 2016). Our IRD results indicate that IRD can be used as a proxy to differentiate glaciated fjords from non glaciated ones (Fig. 5.6).
57
Fig. 5.6.Differentiating glaciated fjords, i.e., fjords with calving glaciers (study areas (2) channels and fjords west of the SPI, and (3) Jorge Montt Fjord, dashed red rectangles), from non glaciated fjords, i.e., fjords connected to rivers (study area (1) Baker Martinez Fjord Complex, dashed red rectangle), based on the presence of ice rafted debris. In all study areas, the sources of terrestrial sediment (rivers and glaciers) are presented with a number, which refers to the type of sediment source on the left (calving glacier) and right (river).
58
5.3. Provenance of terrestrial sediment supply
Figure 2.1 shows the difference in bedrock lithology between the NPI and SPI. The NPI lies above the Northern Patagonian Batholith which is predominantly composed of granodiorites and tonalities (Pankhurst et al., 1999). The SPI is located above the Eastern Andes metamorphic complex, mainly containing polydeformed turbidite successions (Allamand et al., 2008). This difference in bedrock lithology between the NPI and SPI is reflected in the mass-specific magnetic susceptibility data from the BMFC and channels west of the SPI. High mass-specific MS is observed in the Martinez Channel and low mass-specific MS in the channels and fjords west of the SPI (Fig. 4.1). Within the Baker Channel, low mass-specific MS values, approaching those from west of the SPI, are observed near the outlet of the Pascua river (Fig. 4.1). Throughout the Baker Channel, mass-specific MS increases towards Golfo de Penas, approaching the maximum values observed in the Martinez Channel (Fig. 4.1). The source of terrestrial sediment, i.e., the NPI or SPI, differs between both study areas, and between the Baker and Martinez Channels. These observations indicate the control of terrestrial sediment provenance on mass-specific MS. Therefore, the mass-specific MS results will be discussed within each fjord environment separately.
Figure 5.7a visualizes the possible controls on the mass-specific MS signal throughout the Martinez Channel. Here, both log(Ti/Al) and log(Fe/Al) are high and change little with distance from the Baker river (Fig. 5.7a). From figure 5.7a, it can be inferred that the mass-specific MS signal throughout the Martinez Channel is not related to log(Ti/Al) and log(Fe/Al) of the sediment. Within the Martinez Channel, the mass- specific MS is correlated with the mean IRD-free grain size (r = 0.959; p <0.001), log(litho-Si/Al) (r = 0.948; p < 0.001) and with the lithogenic fraction of the sediment (r = 0.750; p = 0.020). Together with Figure 5.7a, these correlations indicate that mass-specific MS is driven by the mean IRD-free grain size, within the Martinez Channel. These correlations are in agreement with the observations regarding MS and grain size by Bertrand et al. (2012b) within the fjords north of the NPI. Since both the fjords north of the NPI and the Martinez Channel receive large amounts of terrestrial sediment from the NPI, the same assumptions regarding the relation between MS and the combination of grain size and the lithogenic fraction of the sediments by Bertrand et al. (2012b) can be made for the Martinez Channel: 1) the MS signal is diluted by non-lithogenic particles, 2) enrichment of heavy ferromagnetic minerals in the coarse fraction of the sediment during sedimentary sorting. In addition, it has been observed that MS increases with grain size within the sediments of the Baker river (Buydens, 2015). However, Fe is associated with fine grained sediments within the Martinez Channel, since log(Fe/Al) is inversely correlated with the mean IRD-free grain size (r = -0.964; p < 0.001). This does not reverse previous assumptions, since the Fe content remains high throughout the Martinez Channel (Fig. 5.7a).
Figure 5.7b visualizes the possible controls on the mass-specific MS signal throughout the Baker Channel. Here, mass-specific MS is not related to the mean IRD-free grain size (Fig. 5.7b). This is confirmed by statistical analysis, no correlation is present between mass-specific MS and mean IRD-free grain size (r = 0.295; p = 0.352). Figure 5.7b indicates the relation between log(Ti/Al) and log(Fe/Al) and the mass-specific MS signal throughout the Baker Channel. Here, both log(Ti/Al) and log(Fe/Al) are correlated with mass- specific MS (log(Ti/Al): r = 0.701; p = 0.011; log(Fe/Al) r = 0.886; p < 0.001). These correlations indicate that the origin of the MS signal within the Baker Channel is driven by Fe- and Ti-bearing ferro- and/or paramagnetic minerals. However, multiple observations complicate the reasoning that the mass-specific MS signal originates from the lithogenic fraction of the sediment which is transported from the SPI towards the Baker Channel by the Pascua river and Jorge Montt glacier (Fig. 5.7b): 1) mass-specific MS, log(Ti/Al) and log(Fe/Al) are not correlated with the mean IRD-free grain size (Appendix 8.4) and 2) mass-specific MS is inversely correlated with the lithogenic fraction of the sediment (r = -0.978; p < 0.001).
59
Fig. 5.7. Spatial variability of mass-specific magnetic susceptibility, and of the physical and bulk inorganic geochemical properties of the sediments potentially controlling the mass-specific MS signal. A) Martinez Channel, B) Baker Channel, and C) Channels and fjords west of the SPI. Within the mass-specific MS plot of the channels west of the SPI (C), one outlier (ST80; 4,265 x 10-6 m3/kg) is not included for plotting purposes.
60
Figure 5.7b indicates that the samples proximal to the Pascua river have low mass-specific MS and low Fe and Ti contents. Throughout the Baker Channel, these three variables increase, with distance from the Pascua river, towards the maximum values observed within the Martinez Channel (Fig. 5.7 a & b). These observations suggest that the mass-specific MS signal within the Baker Channel originates from the supply of Fe- and Ti-bearing ferro- and/or paramagnetic minerals from the NPI throughout the Martinez and Troya Channels (Fig. 5.8).
Figure 5.7c visualizes the possible controls on the mass-specific MS signal throughout the channels west of the SPI. Except for two outliers, the Ti and Fe content remain relatively low and constant with distance from the glacier fronts (Fig. 5.7c). Here, log(Ti/Al) and log(Fe/Al) are not correlated with mass-specific MS (Appendix 8.4). Sample ST80 has an extremely high mass-specific MS (4.265 kg/m3) (Fig. 4.1) and is not included in the mass-specific MS plot of figure 5.7c. Figure 5.7c indicates that the mass-specific MS remains relatively low and constant throughout the channels west of the SPI. Although, a small increase towards the open ocean can be observed (Fig. 5.7c). This small mass-specific MS increase could be controlled by the mean IRD-free grain size (Fig. 5.7c), however these variables are not correlated (r = 0.430; p = 0.142). A moderate correlation between IRD and mass-specific MS (r = 0.644; p = 0.018) indicates a small control of ice rafted debris content on MS. Therefore, it can be assumed that the increase in mass-specific MS is due to the enrichment of heavy ferromagnetic minerals in the ice rafted debris, i.e., the coarse fraction of the sediment (Bertrand et al., 2012b).
Figure 5.8 summarizes the control of terrestrial sediment provenance on the mass-specific MS signal in both study areas. The Martinez Channel only receives glacial meltwater and terrestrial sediment from the NPI. Here, the mass-specific MS remains high (Fig. 5.8) and only decreases due to decreasing grain size of the lithogenic fraction of the sediment (Fig. 5.7a). The channels and fjords west of the SPI only receive glacial meltwater and terrestrial sediment from the SPI. Here, the mass-specific MS remains low (Fig. 5.8), the small increase is due to the increase in IRD (Fig. 5.7c). Within the Baker Channel, the proximal samples have a low mass-specific MS and low Ti and Fe contents, approaching the minimal values observed west of the SPI (Fig. 5.7). Within the distal samples located in the Baker Channel, these variables approach the maximum values observed within the Martinez Channel (Fig. 5.7). Within the Baker Channel, mass-specific MS is not correlated with grain size or IRD content (Appendix 8.4). Here, mass-specific MS is related to the log(Fe/Al) and log(Ti/Al) (Fig. 5.7b). However, the log(Fe/Al) and log(Ti/Al) is relatively low in the lithogenic fraction of the sediments coming from the SPI (Fig. 5.7c). These results indicate a significant input of glacial meltwater and terrestrial sediments from the NPI into the Baker Channel, throughout the Martinez and Troya Channel (Fig. 5.7b & 5.8). The terrestrial sediments from the NPI contain Fe- and Ti- bearing ferro- and/or paramagnetic minerals, which induce the mass-specific MS signal throughout the Baker Channel (Fig. 5.7b & 5.8).
Figure 5.8 indicates that the terrestrial sediment from the NPI can be recognized by high mass-specific MS, and those from the SPI by low mass-specific MS. Within the Baker Channel, terrestrial sediments from the NPI and SPI mix, resulting in an increase of mass-specific MS with distance from the Pascua river and Jorge Montt glacier (Fig. 5.8). These results suggest that mass-specific MS can be used as a proxy to differentiate terrestrial sediment supply from the NPI and SPI into the BMFC (Fig. 5.8).
61
Fig. 5.8 Schematic representation of sediment provenance controlling the mass-specific magnetic susceptibility signal. High mass-specific MS of the Patagonian Batholith (red) controls the mass-specific MS signal of the lithogenic fraction of the sediments within the Martinez Channel. Low mass-specific MS of the Eastern Andes Metamorphic Complex (green) controls the mass-specific MS signal of the lithogenic fraction of the sediments within the channels and fjords west of the SPI. Throughout the Baker Channel, the lithogenic particles from the Patagonian Batholith (red) and Eastern Andes Metamorphic Complex (green) mix and result in an increasing mass-specific MS towards Golfo de Penas. Geological information from SERNAGEOMIN (2003).
62
5.4. Future research
The results of this study could help the interpretation of physical, sedimentological and geochemical data from fjord sediment cores in the scope of glacier variability. This study identified bulk organic geochemical proxies (δ13C and δ15N) of terrestrial sediment supply towards the fjords of the BMFC and west of the SPI (Fig. 5.5). In the case of glaciated fjords, terrestrial sediment supply can serve as an indicator for glacier variability. Bulk inorganic geochemical proxies (log(Fe/Al), log(Zr/Al), log(Ti/Al), log(Mg/Al) and log(litho- Si/Al)) of terrestrial sediment supply were identified in the BMFC, but not in the channels and fjords west of the SPI (Fig. 5.5). The BMFC has multiple sources of terrestrial sediment supply, i.e., multiple rivers and Jorge Montt glacier (Aiken, 2012). Therefore, terrestrial sediment supply cannot be used as an indicator for glacier variability within the Baker and Martinez Channels. The presence or absence of IRD in sediment cores obtained within our study areas can indicate whether the investigated fjord environment was influenced by calving glaciers or not. Within sediment cores obtained in the BMFC, mass-specific MS can differentiate the provenance of terrestrial sediments, i.e., the Northern and Southern Patagonian Icefields (Fig. 5.8).
Future studies should focus on one specific channel/fjord, e.g. Jorge Montt Fjord, so the sedimentological record mainly represents the terrestrial sediment supply from one glacier or river. The sampling density of the surface sediment should be high enough to construct a significant correlation between the proxy of interest and the reference, e.g. distance, to which the relation with terrestrial sediment supply is evaluated. Currently, this is being done in the framework of the HYDROPROX project at Ghent University.
Within the BMFC, care should be taken when interpreting the sedimentological data in the scope of palaeoclimatological investigations since these sediments are possibly disturbed by (mega)turbidites and earthquakes (Piret, 2016). While the seismic influence on the studied surface sediment samples might be negligible (section 2.2 ‘Tectonic Setting’), seismic disturbance on sediment cores might be considerable. Multiple authors have observed that the sediments within Patagonian fjords contain earthquake-triggered deposits (Chapron et al., 2006; Sepúlveda et al., 2010; St-onge et al., 2012; Van Daele et al., 2013). This means that future palaeoclimatological research should carefully consider their core locations, in order to avoid seismic related event deposits.
6. Conclusion
This master dissertation investigated the spatial variability of physical properties, sediment composition, and bulk organic and inorganic geochemistry within fjord sediments along the Southern Patagonian Icefield. This with the aim to evaluate the ability of these fjord sediment properties to estimate past changes in terrestrial sediment supply into the Patagonian fjords. In the case of glaciated fjords, terrestrial sediment supply can serve as an indicator for glacier variability. In addition, these fjord sediments were investigated towards their ability to distinguish terrestrial sediment supply coming from the NPI and SPI, and to differentiate between fjords that are influenced by calving glaciers and fjords that are not. The results of this study could help the interpretation of physical, sedimentological and geochemical data from fjord sediment cores in the scope of glacier variability. In order to accurately reconstruct past glacier variability, multi-proxy based methods are inevitable.
63
This study was conducted in two study areas: 1) the Baker Martinez Fjord Complex, comprising the Baker and Martinez Channels, and 2) the channels and fjords west of the SPI, with a focus on Canal Messier and Canal Wide. Important to notice, is the pioneering aspect of this study. Especially in the channels west of the SPI, the sampling density of the surface sediments was low. This resulted in the difficulty of constructing significant correlations between the investigated properties and the reference variable distance (km), to which the relation with terrestrial sediment supply was evaluated. The results indicate the potential use of the investigated proxies. In order to use these proxies, future research should focus on one specific channel/fjord and the sampling density must be high enough to construct significant correlations. This means that the following conclusions are only preliminary:
1) The carbon and nitrogen isotopic composition of fjord sediments can be used as proxies to estimate past changes in the terrestrial sediment supply into the Baker Martinez Fjord Complex and fjords west of the SPI. More depleted δ13C (‰) and δ15N (‰) values are found near glacier fronts and river outlets, while less depleted δ13C (‰) and δ15N (‰) values represent the more marine environments.
2) Over both study areas, log(litho-Si/Al) is significantly positively correlated with the mean IRD- free grain size (µm) and ice rafted debris (> 150 µm, wt%) content. This indicates that log(litho- Si/Al) can be used as a proxy to estimate grain size variations. However, log(litho-Si/Al) cannot be used to distinguish ice rafted debris from bulk grain size. This relation of log(litho-Si/Al) with grain size is interpreted as quartz being concentrated in the coarse fraction of the sediment (Bertrand et al., 2012b). However, this hypothesis should be investigated by a mineralogical analysis.
3) The use of Al-based elemental log-ratios is best expressed in studies confined to one certain channel, where high sampling density results in a significant correlation with the chosen reference variable, e.g., distance (km). This conclusion is inevitable since the spatial variability of elemental log-ratios may drastically vary between fjords with different terrestrial sources, i.e., calving glaciers or rivers. This is illustrated by the results of this study. The following log-ratios were identified as potential proxies to estimate past changes in terrestrial sediment supply into the Patagonian fjords:
Within the Baker Channel: log(Fe/Al), log(Zr/Al), log(Ti/Al) and log(Mg/Al).
Within the Martinez Channel: log(Fe/Al), log(Zr/Al), log(litho-SI/Al) and log(Mg/Al).
Within the channels and fjords west of the SPI, no Al-based elemental log-ratios were identified to reconstruct terrestrial sediment supply towards the entire area west of the SPI. However, results of log(Fe/Al), log(Ti/Al), log(Zr/Al) and log(Mg/Al) indicate the potential of Al-based elemental log-ratios as proxies of terrestrial sediment supply in the channels and fjords west of the SPI, individually. In order to unravel the complexity of terrestrial sediment supply towards the fjords west of the SPI, future research should focus on one specific channel/fjord and high sampling density.
64
4) Icebergs result from the calving activity of calving glaciers and are the primary transport agent for IRD (Kuijpers et al., 2016). IRD is almost absent within the Baker and Martinez Channels, while significant amounts of IRD were observed within the channels west of the SPI and in Jorge Montt Fjord (De Wilde, 2016). The Baker and Martinez Channels are not influenced by calving glaciers, while the channels west of the SPI and Jorge Montt Fjord are. Our results indicate that IRD > 150 µm (wt%) can be used as a proxy to differentiate fjords which are influenced by calving glaciers from fjords which are not.
5) Throughout the BMFC, mass-specific MS can be used to differentiate the provenance of terrestrial sediment, i.e., the NPI and SPI. Terrestrial sediments from the NPI have high mass- specific MS, while those from the SPI have low mass-specific MS. The Martinez Channel receives terrestrial sediment from the NPI, which is characterized by high mass-specific MS. Throughout this channel, mass-specific MS decreases with decreasing grain size. The channels and fjords west of the SPI receive terrestrial sediment from the SPI, which is characterized by low mass-specific MS. Here, the mass-specific MS increases with increasing IRD content. Within the Baker Channel, terrestrial material from the NPI and SPI mix and the mass-specific MS signal is controlled by Fe- and Ti-bearing ferro- and/or paramagnetic minerals. Here, mass-specific magnetic susceptibility can be used to distinguish the provenance of terrestrial sediment, i.e., the Northern and Southern Patagonian Icefields.
65
7. References
Ackert, R. P., Becker, R. A., Singer, B. S., Kurz, M. D., Caffee, M. W., & Mickelson, D. M. (2008). Patagonian glacier response during the Late Glacial–Holocene transition. Science, 321(5887), 392– 396. Agurto-Detzel, H., Rietbrock, A., Bataille, K., Miller, M., Iwamori, H., & Priestley, K. (2014). Seismicity distribution in the vicinity of the Chile Triple Junction, Aysén Region, southern Chile. Journal of South American Earth Sciences, 51, 1–11. Aiken, C. M. (2012). Seasonal thermal structure and exchange in Baker Channel, Chile. Dynamics of Atmospheres and Oceans, 58, 1–19. Allamand, F. E. I. H., Calderón, M., & Faúndez, V. (2008). The metamorphic complexes of the Patagonian and Fuegian Andes. Geologica Acta, 6(1), 43–53. Aniya, M. (1995). Holocene Glacial Chronology in Patagonia : Tyndall and Upsala Glaciers. Arctic and Alpine Research, 27(4), 311–322. Aniya, M. (1996). Holocene variations of Ameghino Glacier, southern Patagonia. The Holocene, 6, 247– 252. Aniya, M. (1999). Recent Glacier Variations of the Hielos Patagónicos, South America, and Their Contribution to Sea-Level Change. Arctic, Antarctic, and Alpine Research, 31(2), 165–173. Aniya, M. (2013). Holocene glaciations of Hielo Patagónico (Patagonia Icefield), South America: A brief review. Geochemical Journal, 47, 97–105. Aniya, M., Skvarca, P., Casassa, G., Sato, H., & Naruse, R. (1997). Recent Glacier Variations in the Southern Patagonia Icefield, South America. Arctic and Alpine Research, 29(1), 1–12. 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 fjords (41–55°S) as a response to water column productivity. Continental Shelf Research, 31(3), 340–355. Bennett, M. M., & Glasser, N. F. (Eds.). (2011). Glacial geology: ice sheets and landforms. John Wiley & Sons. Bertrand, S., Hughen, K. A., Lamy, F., Stuut, J. B., Torrejon, F., & Lange, C. B. (2012a). Precipitation as the main driver of Neoglacial fluctuations of Gualas glacier, Northern Patagonian Icefield. Climate of the Past, 8(2), 519–534. Bertrand, S., Hughen, K., 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, 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. Bertrand, S., Sterken, M., Vargas-Ramirez, L., De Batist, M., Vyverman, W., Lepoint, G., & Fagel, N. (2010). Bulk organic geochemistry of sediments from Puyehue Lake and its watershed (Chile, 40°S): Implications for paleoenvironmental reconstructions. Palaeogeography, Palaeoclimatology, Palaeoecology, 294(1–2), 56–71. Boldt, K. V. (2014). Fjord sedimentation during the rapid retreat of tidewater glaciers: observations and modeling. PhD thesis in Geological Oceanography, University of Washington. Buydens, L. (2015). De relatie tussen magnetische susceptibiliteit en korrelgrootte in Patagonische riviersedimenten. Ghent University. Cande, S., & Leslie, R. B. (1986). Late Cenozoic tectonics of the Southern Chile Trench. Journal of Geophysical Research: Solid Earth, 91(B1), 91, 471–496. Caniupán, M., Lamy, F., Lange, C. B., Kaiser, J., Arz, H., Kilian, R., … Tiedemann, R. (2011). Millennial- scale sea surface temperature and Patagonian Ice Sheet changes off southernmost Chile (53°S) over the past 60 kyr. Paleoceanography, 26, 1–10. Carter, S. J., & Colman, S. M. (1994). Biogenic Silica in Lake Baikal Sediments: Results From 1990-1992 American Cores. Journal of Great Lakes Research, 20(4), 751–760. Casanova, M., Salazar, O., Seguel, O., & Luzio, W. (2013). The Soils of Chile. In World Soils Book Series, p. 20. Cembrano, J., Hervé, F., & Lavenu, A. (1996). The Liquiñe Ofqui fault zone: a long-lived intra-arc fault system in southern Chile. Tectonophysics, 259(1–3), 55–66.
66
Cembrano, J., Schermer, E., Lavenu, A., & Sanhueza, A. (2000). Contrasting nature of deformation along an intra-arc shear zone, the Liquiñe–Ofqui fault zone, southern Chilean Andes. Tectonophysics, 319(2), 129–149. Chapron, E., Ariztegui, D., Mulsow, S., Villarosa, G., Crivelli, E., Pino, M., … Juvignie, E. (2006). Impact of the 1960 major subduction earthquake in Northern Patagonia (Chile, Argentina). Quaternary International, 158, 58–71. Chu, V. W. (2014). Greenland ice sheet hydrology: A review. Progress in Physical Geography, 38(1), 19– 54. Conley, D. J., & Schelske, C. L. (2001). Biogenic silica. In Tracking Environmental Change Using Lake Sediments. Volume 3: Terrestrial, Algal, and Siliceous Indicators (eds. Smol, J.P., Birks, H.J.B. & Last, W.M.) (pp. 281–293). Dordrecht: Kluwer Academic Publishers. Covelo, E. F., Vega, F. A., & Andrade, M. L. (2008). Sorption and desorption of Cd, Cr, Cu, Ni, Pb and Zn by a Fibric Histosol and its organo-mineral fraction. Journal of Hazardous Materials, 159(2), 342–347. Davies, B. J., & Glasser, N. F. (2012). Accelerating shrinkage of Patagonian glaciers from the Little Ice Age (AD 1870) to 2011. Journal of Glaciology, 58(212), 1063–1084. Davila, P., Figueroa, D., & Muller, E. (2002). Freshwater input into the coastal ocean and its relation with the salinity distribution off austral Chile (35–55°S). Continental Shelf Research, 22, 521–534. Dearing, J. (1994). Environmental magnetic susceptibility. Using the Bartington MS2 System. Kenilworth, Chi Publ. De Wilde, E. (2016). Fjord sedimentary record of the recent and rapid retreat of Jorge Montt glacier, Chilean Patagonia. Ghent University. Diraison, M., Cobbold, P. R., Rossello, E. A., & Amos, A. J. (1998). Neogene dextral transpression due to oblique convergence across the Andes of northwestern Patagonia, Argentina. Journal of South American Earth Sciences, 11(6), 519–532. Dussaillant J, A., Buytaert, W., Meier, C., & Espinoza, F. (2012). Hydrological regime of remote catchments with extreme gradients under accelerated change: the Baker basin in Patagonia. Hydrological Sciences Journal, 57(8), 1530–1542. Faust, J. C., Knies, J., Slagstad, T., Vogt, C., Milzer, G., & Giraudeau, J. (2014). Geochemical composition of Trondheimsfjord surface sediments: Sources and spatial variability of marine and terrigenous components. Continental Shelf Research, 88, 61–71. Fiers, G. (2016). Postglacial evolution of biogenic silica productivity in Lago Castor, northern Chilean Patagonia. Ghent University. García, J. L., Kaplan, M. R., Hall, B. L., Schaefer, J. M., Vega, R. M., Schwartz, R., & Finkel, R. (2012). Glacier expansion in southern Patagonia throughout the Antarctic cold reversal. Geology, 40(9), 859– 862. Garreaud, R., Lopez, P., Minvielle, M., & Rojas, M. (2013). Large-Scale Control on the Patagonian Climate. Journal of Climate, 26(1), 215–230. Glasser, N. F., & Ghiglione, M. C. (2009). Structural, tectonic and glaciological controls on the evolution of fjord landscapes. Geomorphology, 105(3–4), 291–302. Glasser, N. F., Harrison, S., Schnabel, C., Fabel, D., & Jansson, K. N. (2012). Younger Dryas and early Holocene age glacier advances in Patagonia. Quaternary Science Reviews, 58, 7–17. Glasser, N. F., Harrison, S., Winchester, V., & Aniya, M. (2004). Late Pleistocene and Holocene palaeoclimate and glacier fluctuations in Patagonia. Global and Planetary Change, 43, 79–101. Glasser, N. F., Jansson, K. N., Duller, G. A. T., Singaray, J., Holloway, M., & Harrison, S. (2016). Glacial lake drainage in Patagonia (13-8 kyr) and response of the adjacent Pacific Ocean. Scientific Reports, 6. Glasser, N. F., Jansson, K. N., Harrison, S., & Kleman, J. (2008). The glacial geomorphology and Pleistocene history of South America between 38 S and 56 S. Quaternary Science Reviews, 27(3), 365–390. Glasser, N., & Jansson, K. (2008). The glacial map of southern South America. Journal of Maps, 4(1), 175– 196. Govindaraju, K., Potts, P. J., Webb, P. C., & Watson, J. S. (1994). 1994 report on Whin sill dolerite WS-E from England and Pitscurrie microgabbro PM-S from Scotland: assessment by one hundred and four international laboratories. Geostandards and Geoanalytical Research, 18(2), 211–300. Granon, E. (2015). Post-glacial evolution of Lago Castor (Chile, 46°S): a bulk organic geochemical
67
perspective. Ghent University. Gut, B. (2008). Geology, climate, and soils of Patagonia. In Trees in Patagonia (pp. 9–18). Basel: Springer. Harrison, S., & Glasser, N. F. (2011). The Pleistocene Glaciations of Chile. Developments in Quaternary Science, 15, 739–756. Harrison, S., & Winchester, V. (2000). Nineteenth- and Twentieth-Century Glacier Fluctuations and Climatic Implications in the Arco and Colonia Valleys, Hielo Patagonico Norte, Chile. Arctic, Antarctic, and Alpine Research, 32(1), 55–63. Heiri, O., Lotter, A. F., & Lemcke, G. (2001). Loss on Ignition as a Method for Estimating Organic and Carbonate Content in Sediments : Reproducibility and Comparability of Results. Journal of Paleolimnology, 25(1), 101–110. Hervé, F., Pankhurst, R. J., Fanning, C. M., Calderon, M., & Yaxley, G. M. (2007). The South Patagonian batholith: 150my of granite magmatism on a plate margin. Lithos, 97, 373–394. 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. Howard, P. J. A., & Howard, D. M. (1990). Use of organic carbon and loss-on-ignition to estimate soil organic matter in different soil types and horizons. Biology and Fertility of Soils, 9(4), 306–310. Kaplan, M. R., Ackert, R. P., Singer, B. S., Douglass, D. C., Kurz, M. D., Kaplan, M. R., … Kurz, M. D. (2004). Cosmogenic nuclide chronology of millennial-scale glacial advances during O-isotope stage 2 in Patagonia. Geological Society of America Bulletin, 116(3–4), 308–321. Kaplan, M. R., Douglass, D. C., Singer, B. S., Ackert, R. P., & Caffee, M. W. (2005). Cosmogenic nuclide chronology of pre-last glacial maximum moraines at Lago Buenos Aires, 46°S, Argentina. Quaternary Research, 63(3), 301–315. Kilian, R., Hohner, M., Biester, H., Wallrabe-Adams, H. J., & Stern, C. R. (2003). Holocene peat and lake sediment tephra record from the southernmost Chilean Andes (53-55 S). Revista Geologica de Chile, 30(1), 23–37. Kuijpers, A., Knutz, P., & Moros, M. (2016). ICE-RAFTED DEBRIS. In Encyclopedia of Marine Geosciences (eds. Harff, J., Meschede, M., Petersen, J. & Thiede, J.) (pp. 359–362). Dordrecht: Springer Netherlands. Lagabrielle, Y., Guivel, C., Maury, R. C., Bourgois, J., Fourcade, S., & Martin, H. (2000). Magmatic-tectonic effects of high thermal regime at the siteof active ridge subduction: The Chile Triple Junction model. Tectonophysics, 326(3–4), 255–268. Lamb, A. L., Wilson, G. P., & Leng, M. J. (2006). A review of coastal palaeoclimate and relative sea-level reconstructions using δ13C and C/N ratios in organic material. Earth-Science Reviews, 75(1), 29–57. Last, W. M., & Smol, J. P. (2002). Tracking Environmental Change Using Lake Sediments. Volume 2: Physical and Geochemical Methods. Book (Vol. 2). Springer Science & Business Media. Lopez, P., Chevallier, P., Favier, V., Pouyaud, B., Ordenes, F., & Oerlemans, J. (2010). A regional view of fluctuations in glacier length in southern South America. Global and Planetary Change, 71(1–2), 85– 108. Masiokas, M. H., Rivera, A., Espizua, L. E., Villalba, R., Delgado, S., & Carlos, J. (2009). Glacier fluctuations in extratropical South America during the past 1000 years. Palaeogeography, Palaeoclimatology, Palaeoecology, 281(3–4), 242–268. McLennan, S. M., Bock, B., Hemming, S. R., Hurowitz, J. A., Lev, S. M., & McDaniel, D. K. (2003). The roles of provenance and sedimentary processes in the geochemistry of sedimentary rocks. In D. R. Lentz (Ed.), Geochemistry of Sediments and Sedimentary Rocks: Evolution Considerations to Mineral Deposit-Forming Environments. (pp. 7–38). Geological Association of Canada, GeoText 4, St. John’s, Newfoundland. Melnick, D., Bookhagen, B., Strecker, M. R., & Echtler, H. P. (2009). Segmentation of megathrust rupture zones from fore-arc deformation patterns over hundreds to millions of years, Arauco peninsula, Chile. Journal of Geophysical Research: Solid Earth, 114(B1), 1–23. Menounos, B., Clague, J. J., Osborn, G., Davis, P. T., Ponce, F., Goehring, B., … Marr, R. (2013). Latest Pleistocene and Holocene glacier fluctuations in southernmost Tierra del Fuego, Argentina. Quaternary Science Reviews, 77, 70–79. Mercer, J. H. (1970). Variations of some Patagonian glaciers since the Late-Glacial. American Journal of Science, 269, 1–25.
68
Mercer, J. H. (1976). Glacial history of southernmost South America. Quaternary Research, 6, 125–166. Mercer, J. H. (1982). Holocene glacier variations in southern Patagonia. Striae, 18, 35–40. Meyers, P. A., & Teranes, J. L. (2001). Sediment Organic matter. In Tracking Environmental Change Using Lake Sediments Volume 2: Physical and Geochemical Methods (eds. Last, W.M. & Smol, J.P.) (pp. 239–269). Dordrecht: Kluwer Academic Publishers. Moffat, C. (2014). Wind-driven modulation of warm water supply to a proglacial fjord, Jorge Montt Glacier, Patagonia. Geophysical Research Letters, 41(11), 3943–3950. Monroe, J. S., & Wicander, R. (2009). The Changing Earth: Exploring Geology and Evolution. Brooks Cold, Belmont, CA, USA. Moreno, P. I., Kaplan, M. R., François, J. P., Moy, C. M., Stern, C. R., & Kubik, P. W. (2009). Renewed glacial activity during the Antarctic cold reversal and persistence of cold conditions until 11.5 ka in southwestern Patagonia. Geology, 37(4), 375–378. Moreno, T., & Gibbons, W. (2007). The geology of Chile. Geological Society of London. Mortlock, R. A., & Froelich, P. N. (1989). A simple method for the rapid determination of biogenic opal in pelagic marine sediments. Oceanographic Research Papers, 36(9), 1415–1426. Mugford, R. I., & Dowdeswell, J. A. (2011). Modeling glacial meltwater plume dynamics and sedimentation in high ‐ latitude fjords. Journal of Geophysical Research, 116, 1–20. Mulder, T., Syvitski, J. P. M., Migeon, S., Faugeres, J. C., & Savoye, B. (2003). Marine hyperpycnal flows: Initiation, behaviorand related deposits. A review. Marine and Petroleum Geology, 20(6–8), 861–882. Mulitza, S., Prange, M., Stuut, J. B., Zabel, M., Von Dobeneck, T., Itambi, A. C., … Wefer, G. (2008). Sahel megadroughts triggered by glacial slowdowns of Atlantic meridional overturning. Paleoceanography, 23(4), 1–11. Munoz, Y. P., & Wellner, J. S. (2016). Local controls on sediment accumulation and distribution in a fjord in the West Antarctic Peninsula: implications for palaeoenvironmental interpretations. Polar Research, 35(1), 25284. Murdie, R. E., Prior, D. J., Styles, P., Flint, S. S., Pearce, R. G., & Agar, S. M. (1993). Seismic responses to ridge-transform subduction:Chile triple junction. Geology, 21(12), 1095–1098. Murray, R. W., Miller, D. J., & Kryc, K. A. (2000). Analysis of major and trace elements in rocks, sediments, and interstitial waters by inductively coupled plasma–atomic emission spectrometry (ICP-AES). ODP Technical Note 29. Nelson, E., Forsythe, R., & Arit, I. (1994). Ridge collision tectonics in terrane development. Journal of South American Earth Sciences, 7(3–4), 271–278. Nesbitt, H. W., & Young, G. M. (1996). Petrogenisis of sediments in the absence of chemical weathering: effects of abrasion and sorting on bulk composition and mineralogy. Sedimentology, 43, 341–358. Ohlendorf, C., & Sturm, M. (2008). A modified method for biogenic silica determination. Journal of Paleolimnology, 39(1), 137–142. Pankhurst, R. J., Weaver, S. D., Hervé, F., & Larrondo, P. (1999). Mesozoic – Cenozoic Evolution of the North Patagonian Batholith in Aysén, Southern Chile. Journal of the Geological Society, London, 156, 673–694. Pantoja, S., Iriarte, L., & Daneri, G. (2011). Oceanography of the Chilean Patagonia. Continental Shelf Research, 31(3), 149–153. Pardo‐Casas, F., & Molnar, P. (1987). Relative motion of the Nazca (Farallon) and South American plates since Late Cretaceous time. Tectonics, 6(3), 233–248. Parplies, J., Lücke, A., Vos, H., Mingram, J., Stebich, M., Radtke, U., … Schleser, G. H. (2008). Late glacial environment and climate development in northeastern China derived from geochemical and isotopic investigations of the varved sediment record from Lake Sihailongwan (Jilin Province). Journal of Paleolimnology, 40(1), 471–487. 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(1), 65–72. Peroni, J. I., Tassone, A. A., Menichetti, M., & Cerredo, M. E. (2009). Geophysical modeling and structure of Ushuaia Pluton, Fuegian Andes, Argentina. Tectonophysics, 476(3), 436–449. Piret, L. (2016). Triggering mechanism and depositional processes of a late Holocene megaturbidite in Baker fjord, Chilean Patagonia. Ghent University. Rebolledo, L., Lange, C. B., Bertrand, S., Munoz, P., Salamanca, M., Lazo, P., … Dezileau, L. (2015). Late Holocene precipitation variability recorded in the sediments of Reloncaví Fjord (41 S, 72 W), Chile.
69
Quaternary Research, 84(1), 21–36. Rebolledo, L., Lange, C., Bertrand, S., Quiroga, E., & Silva, N. (in prep.). Geochemical characterization of the surface sediments of the Martinez/Baker fjord system. In Preparation. Rignot, E., Rivera, A., & Casassa, G. (2003). Contribution of the Patagonia Icefields of South America to sea level rise. Science, 302(5644), 434–437. Rivera, A., Benham, T., Casassa, G., & Bamber, J. (2007). Ice elevation and areal changes of glaciers from the Northern Patagonia Ice field, Chile. Global Planetary Change, 59, 126–137. 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(2), 403–414. Robinson, R. S., Kienast, M., Albuquerque, A. L., Altabet, M., Contreras, S., Holz, R. D. P., … Schneider- mor, A. (2012). A review of nitrogen isotopic alteration in marine sediments. Paleoceanography, 27(4). Rosling, A., Landeweert, R., Lindahl, B. D., Larsson, K. H., Kuyper, T. W., Taylor, A. F. S., & Finlay, R. D. (2003). Vertical distribution of ectomycorrhizal fungal taxa in a podzol soil profile. New Phytologist, 159(3), 775–783. Ryzak, M., & Bieganowski, A. (2011). Methodological aspects of determining soil particle-size distribution using the laser diffraction method. Journal of Plant Nutrition and Soil Science, 174(4), 624–633. Sagredo, E. A., & Lowell, T. V. (2012). Climatology of Andean glaciers : A framework to understand glacier response to climate change. Global and Planetary Change, 86–87, 101–109. Sandgren, P., & Snowball, I. (2001). Application of mineral magnetic techniques to paleolimnology. In Tracking Environmental Change Using Lake Sediments Volume 2: Physical and Geochemical Methods (eds. Last, W.M. & Smol, J.P.) (pp. 217–237). Dordrecht: Kluwer Academic Publishers. Sepúlveda, J., Pantoja, S., & Hughen, K. A. (2011). Sources and distribution of organic matter in northern Patagonia fjords, Chile (~44–47°S): A multi-tracer approach for carbon cycling assessment. Continental Shelf Research, 31(3), 315–329. Sepúlveda, S. A., Serey, A., Lara, M., Pavez, A., & Rebolledo, S. (2010). Landslides induced by the April 2007 Aysén Fjord earthquake, Chilean Patagonia. Landslides, 7, 483–492. SERNAGEOMIN. (2003). Mapa Geológico de Chile: versión digital. Servicio Nacional de Geología Y Minería, Publicación Geológica Digital, No. 4, (CD-ROM, versión1.0, 2003). Santiago. Sievers, H. A., & Silva, N. (2008). Water masses and circulation in austral Chilean channels and fjords. Progress in the Oceanographic Knowledge of Chilean Interior Waters, from Puerto Montt to Cape Horn. Comité Oceanográfico Nacional-Pontificia Universidad Católica de Valparaíso, Valparaíso, 53– 58. Silva, N., & Prego, R. (2002). Carbon and Nitrogen Spatial Segregation and Stoichiometry in the Surface Sediments of Southern Chilean Inlets (41°–56°S). Estuarine, Coastal and Shelf Science, 55(5), 763– 775. Silva, N., Vargas, C. A., & 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), 330–339. St-onge, G., Chapron, E., Mulsow, S., Salas, M., Viel, M., Debret, M., … Locat, J. (2012). Comparison of earthquake-triggered turbidites from the Saguenay (Eastern Canada) and Reloncavi (Chilean margin) Fjords: Implications for paleoseismicity and sedimentology. Sedimentary Geology, 243–244, 89–107. Stern, C. R. (2004). Active Andean volcanism: its geological and tectonic setting. Revista Geologica de Chile, 31(2), 161–206. Stern, C. R. (2008). Holocene tephrochronology record of large explosive eruptions in the southernmost Patagonian Andes. Bulletin of Volcanology, 70(4), 435–454. Strelin, J. A., Denton, G. H., Vandergoes, M. J., Ninnemann, U. S., & Putnam, A. E. (2011). Radiocarbon chronology of the late-glacial Puerto Bandera moraines, Southern Patagonian Icefield, Argentina. Quaternary Science Reviews, 30(19), 2551–2569. Sugden, D. E., Bentley, M. J., Fogwill, C. J., Hulton, N. R. J., Mcculloch, R. D., & Purves, R. S. (2005). Late-Glacial Glacier Events in Southernmost South America: A Blend of “Northern” and “Southern” Hemispheric Climatic Signals? Geografiska Annaler: Series A, Physical Geography, 87(2), 273–288. Syvitski, J. P., Burrell, D. C., & Skei, J. M. (1987). Fjords: processes and products. Springer-Verlag New York Inc. Syvitski, J. P., & Shaw, J. (1995). Sedimentology and geomorphology of fjords. Developments in
70
Sedimentology, 53, 113–178. Talbot, M. R., & Laerdal, T. (2000). The Late Pleistocene-Holocene palaeolimnology of Lake Victoria, East Africa, based upon elemental and isotopic analyses of sedimentary organic matter. Journal of Paleolimnology, 23(2), 141–164. Teranes, J. L., & Bernasconi, S. M. (2000). The record of nitrate utilization and productivity limitation provided by δ15N values in lake organic matter—A study of sediment trap and core sediments from Baldeggersee, Switzerland. Limnology and Oceanography, 45(4), 801–813. Thomson, S. N. (2002). Late Cenozoic geomorphic and tectonic evolution of the Patagonian Andes between latitudes 42 S and 46 S: An appraisal based on fission-track results from the transpressional intra-arc Liquine-Ofqui fault zone. Geological Society of America Bulletin, 114(9), 1159–1173. Tribovillard, N., Algeo, T. J., Lyons, T., & Riboulleau, A. (2006). Trace metals as paleoredox and paleoproductivity proxies: An update. Chemical Geology, 232(1), 12–32. Turner, K. J., Fogwill, C. J., McCulloch, R. D., & Sugden, D. E. (2005). Deglaciation of the Eastern Flank of the North Patagonian Icefield and Associated Continental-Scale Lake Diversions. Geografiska Annaler: Series A, Physical Geography, 87(2), 363–374. Valdenegro, A., & Silva, N. (2003). Caracterización oceanográfica física y química de la zona de canales y fiordos australes de Chile entre el estrecho de Magallanes y cabo de Hornos (Cimar 3 Fiordos). Ciencia Y Tecnología Del Mar, 26(2), 19–60. Van Daele, M., Versteeg, W., Pino, M., Urrutia, R., & Batist, M. De. (2013). Widespread deformation of basin-plain sediments in Aysén fjord (Chile) due to impact by earthquake-triggered, onshore- generated mass movements. Marine Geology, 337, 67–79. Vandekerkhove, E. (2014). The volcanic ash soils of Northern Chilean Patagonia (44°–48°S): Distribution, weathering and influence on river chemistry. Ghent University. Vandekerkhove, E., Bertrand, S., Reid, B., Bartels, A., & Charlier, B. (2016). Sources of dissolved silica to the fjords of northern Patagonia (44-48°S): The importance of volcanic ash soil distribution and weathering. Earth Surface Processes and Landforms, 41(4), 499–512. Vargas, C. A., Martinez, R. A., San Martin, 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 Shelf Research, 31(3), 187–201. Verardo, D. J., Froelich, P. N., & McIntyre, A. (1990). Determination of organic carbon and nitrogen in marine sediments using the Carlo Erba NA-1500 Analyzer. Oceanographic Research Papers, 37(1), 157–165. Verhoogen, J. (1962). Distribution of titanium between silicates and oxides in igneous rocks. American Journal of Science, 260, 211–220. Warren, C., & Aniya, M. (1999). The calving glaciers of southern South America. Global and Planetary Change, 22(1), 59–77. Warren, C., & Sugden, D. (1993). The Patagonian Icefields: A Glaciological Review. Arctic and Alpine Research, 25, 316–331. Weltje, G. J. (2012). Quantitative models of sediment generation and provenance: state of the art and future developments. Sedimentary Geology, 280, 4–20. Weltje, G. J., & Tjallingii, R. (2008). Calibration of XRF core scanners for quantitative geochemical logging of sediment cores: theory and application. Earth and Planetary Science Letters, 274(3), 423–438.
71
8. Appendix
8.1 Data
Table 8.1.1. Number, research cruise, sample label, location (Lat(S) = latitude south, and Lon(w) = longitude west), depth of the surface salinity measurements (m), surface salinity (psu) and distance (km) of the sampling stations. Cimar 20 samples are indicated with prefix ‘ST’, Copas 2014 samples with prefix ‘E’ (except for sample ‘Isla Irene’) and Gutierrez 2013 samples with suffix ‘G13’. (*) Distance (km) towards the nearest river mouth or glacier front (Appendix 8.2). Cruise Sample Lat (S) Lon (W) Depth salinity measurent (m) Surface Salinity (psu) Distance (km) * 1 Cimar 20 2014 ST3 -47.74 -74.76 1.00 30.779 131.13 2 Cimar 20 2014 ST4 -47.79 -74.53 0.00 17.000 113.79 3 Cimar 20 2014 ST5 -47.91 -74.48 2.00 29.301 98.66 4 Cimar 20 2014 ST5P -47.89 -74.55 1.00 23.092 104.30 5 Cimar 20 2014 ST6 -47.98 -74.27 1.00 28.378 81.18 6 Cimar 20 2014 ST7 -47.96 -74.02 1.00 29.113 61.92 7 Cimar 20 2014 ST8 -48.00 -73.78 1.00 13.739 43.56 8 Cimar 20 2014 ST9 -48.03 -73.59 1.00 12.092 28.86 9 Cimar 20 2014 ST10 -48.17 -73.36 1.00 6.360 4.58 10 Cimar 20 2014 ST11 -47.88 -73.74 1.00 13.877 16.45 11 Cimar 20 2014 ST12 -47.79 -73.64 1.00 1.157 4.09 12 Cimar 20 2014 ST14 -47.77 -73.69 1.00 9.757 8.87 13 Cimar 20 2014 ST18 -48.50 -74.50 1.00 24.860 31.83 14 Cimar 20 2014 ST19 -48.65 -74.38 1.00 23.786 12.99 15 Cimar 20 2014 ST20 -48.89 -74.43 1.00 22.132 33.90 16 Cimar 20 2014 ST21B -48.72 -74.05 1.00 19.822 6.55 17 Cimar 20 2014 ST23 -49.19 -74.38 1.00 27.100 69.60 18 Cimar 20 2014 ST25 -49.56 -74.23 1.00 27.391 41.75 19 Cimar 20 2014 ST28 -49.27 -74.06 1.00 28.526 5.16 20 Cimar 20 2014 ST30 -49.53 -73.98 1.00 26.416 23.65 21 Cimar 20 2014 ST31 -49.80 -74.37 1.00 26.863 69.65 22 Cimar 20 2014 ST80 -49.96 -74.96 1.00 29.240 140.11 23 Cimar 20 2014 ST83 -49.81 -75.16 1.00 30.081 162.93 24 Cimar 20 2014 ST86 -49.29 -75.52 1.00 28.424 226.91 25 Cimar 20 2014 ST91 -48.06 -75.24 1.00 28.800 130.70 26 Cimar 20 2014 ST93 -47.37 -74.65 1.00 31.393 173.45 27 Cimar 20 2014 ST96 -47.77 -74.23 1.00 26.110 52.17 28 Cimar 20 2014 ST97 -47.77 -74.02 1.00 9.072 36.97 29 Cimar 20 2014 ST98 -47.82 -73.84 1.00 10.267 20.10 30 Copas October 2014 E1 Draga -47.79 -73.61 1.00 1.0527 2.21 31 Copas October 2014 E6 Draga -47.80 -74.60 1.00 29.2205 118.48 32 Copas October 2014 E4 -47.77 -74.23 1.00 14.2623 52.55 33 Copas October 2014 Isla Irene -47.81 -74.06 34.84 34 Copas October 2014 E13 -48.19 -73.34 1.00 3.1207 2.58 35 Copas October 2014 E8 -47.95 -74.13 1.00 28.9897 70.37 36 Copas October 2014 E3 -47.81 -73.96 1.00 7.5588 29.87 37 Gutierrez 2013 T1-G13 -47.80 -73.53 1.96 38 Gutierrez 2013 T2-G13 -47.81 -73.54 1.06 39 Gutierrez 2013 T3-G13 -47.81 -73.54 1.01 40 Gutierrez 2013 E1-G13 -48.30 -73.46 2.97 41 Gutierrez 2013 E2-G13 -48.29 -73.46 3.43 42 Gutierrez 2013 E4-G13 -48.24 -73.47 11.05 43 Gutierrez 2013 E5-G13 -48.19 -73.49 17.28 44 Gutierrez 2013 E3-G13 -48.19 -73.49 17.34 45 Gutierrez 2013 E6-G13 -48.16 -73.46 12.52 46 Gutierrez 2013 E1B-G13 -48.28 -73.44 5.00 47 Gutierrez 2013 E4B-G13 -48.23 -73.47 11.66 48 Gutierrez 2013 E2B-G13 -48.19 -73.51 17.54 49 Gutierrez 2013 E5B-G13 -48.19 -73.49 17.66 50 Gutierrez 2013 E3B-G13 -48.07 -73.54 22.46
72
Table 8.1.2. Physical data. MS = Magnetic Susceptibility and IRD = Ice Rafted Debris. Sample Mass-Specific MS (10-6 m3/kg) Mean IRD-Free Grain Size (µm) Grains larger than 150 µm (%) IRD >63 µm (wt%) IRD >150 µm (wt%) ST3 916 11.70 27.04 5.17 1.48 ST4 684 6.95 0.00 1.88 0.42 ST5 533 6.61 0.04 1.40 0.18 ST5P 493 5.82 0.11 0.47 0.11 ST6 527 5.25 0.00 0.43 0.07 ST7 503 4.92 0.00 0.67 0.25 ST8 401 5.70 0.00 ST9 260 5.30 0.00 0.04 0.02 ST10 257 17.10 1.84 3.34 0.86 ST11 434 5.32 0.00 ST12 849 14.00 1.30 0.66 0.02 ST14 603 6.81 2.01 0.04 0.00 ST18 551 12.20 28.87 6.11 1.31 ST19 344 6.17 0.00 14.04 10.26 ST20 375 6.03 7.78 17.00 5.01 ST21B 416 6.54 0.00 6.05 1.72 ST23 323 8.26 3.96 13.09 2.97 ST25 379 7.66 0.00 1.60 0.36 ST28 513 13.60 0.53 1.68 0.37 ST30 349 11.80 11.36 27.29 24.75 ST31 326 7.98 0.00 4.29 1.75 ST80 4 265 59.00 58.98 68.15 56.11 ST83 531 23.40 1.10 6.67 2.06 ST86 585 106.00 62.33 80.31 57.77 ST91 649 15.70 32.67 15.81 2.43 ST93 2 582 42.20 0.68 16.19 0.03 ST96 436 4.05 10.79 11.84 2.54 ST97 424 4.36 0.00 7.57 1.29 ST98 564 6.15 7.44 0.69 0.15 E1 Draga 990 14.70 15.62 E6 Draga 1 590 20.50 29.71 E4 356 4.21 0.00 Isla Irene 351 4.06 19.01 E13 249 23.30 15.48 E8 459 7.12 12.12 E3 332 5.32 10.17
73
Table 8.1.3. Sediment composition data: Loss On Ignition (LOI), Biogenic Opal content (Bio-Opal, wt%), Carbonate content
(CaCO3, wt%), Organic Matter content (OM, wt%) and the lithogenic fraction of the sediment (wt%). s.d. = standard deviation.
Sample LOI550 (%) LOI950 (%) Bio-Opal (wt%) CaCO3 (wt%) OM (wt%) Lithogenic Fraction (wt%) ST3 8.82 2.58 7.45 2.42 5.30 84.82 ST4 7.88 5.23 6.74 1.14 3.78 88.34 ST5 7.59 5.66 6.05 0.60 3.07 90.28 ST5P 6.39 2.18 6.36 0.54 2.50 90.60 ST6 5.38 1.58 5.19 0.26 1.90 92.65 ST7 4.46 1.48 4.35 0.09 1.18 94.39 ST8 4.37 1.73 3.59 0.05 0.93 95.42 ST9 4.04 1.29 3.57 0.02 1.01 95.40 ST10 4.23 1.62 4.04 0.02 0.98 94.96 ST11 4.53 1.65 3.30 0.03 0.84 95.83 ST12 4.16 1.37 2.33 0.02 1.11 96.54 ST14 3.33 1.30 3.37 0.01 0.61 96.02 ST18 6.43 2.51 8.04 0.09 2.18 89.69 ST19 4.31 2.21 8.34 0.02 1.33 90.31 ST20 6.00 1.60 8.69 0.07 1.94 89.29 ST21B 4.32 1.51 7.99 0.01 1.42 90.58 ST23 9.60 3.93 9.71 2.25 4.89 83.15 ST25 3.64 1.76 4.90 0.11 1.14 93.86 ST28 2.60 0.92 4.16 0.13 0.54 95.18 ST30 3.82 1.22 3.51 0.38 1.10 95.00 ST31 5.67 2.67 5.64 0.28 1.78 92.30 ST80 7.44 3.82 3.20 13.16 3.49 80.15 ST83 20.69 13.48 6.58 34.04 13.44 45.94 ST86 5.14 2.93 0.57 10.29 1.76 87.39 ST91 15.40 6.86 8.13 16.97 9.47 65.43 ST93 3.42 0.76 2.99 0.42 1.55 95.05 ST96 5.80 2.08 5.66 0.23 1.76 92.35 ST97 4.92 2.60 3.88 0.04 1.09 94.99 ST98 4.35 2.00 3.46 0.01 0.75 95.77 E1 Draga 3.93 0.88 1.49 0.04 1.31 97.16 E6 Draga 10.58 4.29 5.27 16.87 5.85 72.01 E4 4.51 0.11 1.18 94.20 Isla Irene 4.61 0.12 1.02 94.25 E13 5.05 1.99 3.99 0.03 1.72 94.26 E8 5.89 0.23 0.71 93.17 E3 4.08 0.37 T1-G13 2.47 T2-G13 2.35 T3-G13 1.38 E1-G13 1.78 E2-G13 1.25 E4-G13 1.32 E5-G13 1.40 E3-G13 1.54 E6-G13 5.70 E1B-G13 1.10 E4B-G13 1.07 E2B-G13 2.77 E5B-G13 3.66 E3B-G13 5.76
s.d. 0.15 0.03
74
13 Table 8.1.4. Bulk organic geochemical data: Total Organic Carbon content (Corg, wt%), Carbon Isotopic Composition (δ C, ‰), Total Nitrogen content (N, wt%), Nitrogen Isotopic Composition (δ15N, ‰), Atomic N/C ratio, Fraction of Terrestrial
Organic Carbon (FT, %) and Fraction of Marine Organic Carbon (FM, %). s.d. = standard deviation. 13 15 Sample Corg (wt%) δ C (‰) N (wt%) δ N (‰) Atomic N/C FT (%) FM (%) ST3 2.41 -19.92 0.30 9.26 0.14 0.01 0.99 ST4 1.72 -21.14 0.21 8.86 0.14 0.16 0.84 ST5 1.40 -21.51 0.18 8.59 0.15 0.21 0.79 ST5P 1.14 -22.04 0.14 8.33 0.15 0.28 0.72 ST6 0.86 -22.15 0.11 7.88 0.15 0.29 0.71 ST7 0.54 -22.75 0.07 6.94 0.16 0.37 0.63 ST8 0.42 -24.22 0.06 5.49 0.16 0.55 0.45 ST9 0.46 -24.90 0.06 4.38 0.16 0.64 0.36 ST10 0.45 -25.75 0.06 4.08 0.16 0.75 0.25 ST11 0.38 -25.40 0.04 4.40 0.12 0.70 0.30 ST12 0.50 -26.57 0.04 2.33 0.09 0.85 0.15 ST14 0.28 -25.88 0.03 3.93 0.11 0.77 0.23 ST18 0.99 -22.14 0.12 8.13 0.14 0.29 0.71 ST19 0.61 -23.72 0.08 6.22 0.15 0.49 0.51 ST20 0.88 -22.86 0.11 7.18 0.14 0.38 0.62 ST21B 0.64 -25.37 0.07 5.04 0.12 0.70 0.30 ST23 2.22 -20.49 0.27 8.66 0.14 0.08 0.92 ST25 0.52 -22.38 0.08 5.97 0.17 0.32 0.68 ST28 0.24 -24.39 0.04 3.37 0.18 0.58 0.42 ST30 0.50 -23.37 0.07 5.84 0.16 0.45 0.55 ST31 0.81 -21.44 0.12 7.47 0.17 0.20 0.80 ST80 2.98 -5.34 0.11 9.65 ------ST83 6.51 -11.38 0.52 9.99 ------ST86 1.77 -3.67 0.06 9.29 ------ST91 3.55 -14.51 0.34 9.53 ------ST93 0.70 -20.27 0.09 8.64 0.14 0.05 0.95 ST96 0.80 -22.80 0.09 7.78 0.14 0.37 0.63 ST97 0.50 -24.50 0.05 6.33 0.12 0.59 0.41 ST98 0.34 -25.50 0.03 4.24 0.12 0.72 0.28 E1 Draga 0.60 -26.30 0.06 1.31 0.12 0.82 0.18 E6 Draga 2.72 -10.29 0.15 8.79 ------E4 0.54 -22.41 0.06 11.96 0.14 0.32 0.68 Isla Irene 0.46 -23.41 0.05 7.10 0.13 0.45 0.55 E13 0.78 -27.10 0.08 3.01 0.12 0.92 0.08 E8 0.32 -23.00 0.03 7.05 0.12 0.40 0.60 E3 0.17 -24.86 0.00 2.82 0.02 0.64 0.36
s.d. 0.56 0.10 0.09 0.08
75
Table 8.1.5. Bulk inorganic geochemical data. Elemental concentration in the lithogenic fraction of the sediment, expressed as weight percentage (wt%) for the major elements and in parts per million (ppm) for the minor elements. rsd (%) = relative standard deviation (%). Sample Al (wt%) Ca (wt%) Fe (wt%) K (wt%) Mg (wt%) Mn (wt%) Na (wt%) P (wt%) Si (wt%) Ti (wt%) litho-Si (wt%) Ba (ppm) Sr (ppm) Zr (ppm) ST3 8.45 3.39 4.92 2.36 2.17 0.10 3.93 0.13 25.40 0.51 22.29 585 316 59 ST4 8.86 2.65 5.31 2.44 2.37 0.22 5.28 0.13 24.21 0.51 21.40 630 287 75 ST5 9.03 2.21 5.24 2.49 2.31 0.20 5.65 0.11 23.67 0.50 21.14 640 258 68 ST5P 10.18 2.04 6.03 2.97 2.21 0.14 3.59 0.12 25.54 0.54 22.89 742 241 88 ST6 10.66 1.84 6.03 3.05 2.14 0.12 3.16 0.10 26.15 0.55 23.99 768 230 93 ST7 10.73 1.61 6.05 3.07 2.07 0.16 2.93 0.10 25.96 0.56 24.15 798 212 95 ST8 10.91 1.41 5.84 3.02 1.98 0.15 3.15 0.09 26.80 0.55 25.30 809 197 117 ST9 11.65 1.04 5.84 3.15 1.74 0.08 2.29 0.08 26.82 0.52 25.33 866 170 136 ST10 10.19 0.99 4.72 2.61 1.41 0.12 2.85 0.08 28.17 0.45 26.49 732 186 207 ST11 10.29 1.96 6.28 2.93 2.26 0.24 3.43 0.10 26.00 0.60 24.63 739 242 81 ST12 9.84 2.03 5.60 2.66 1.95 0.16 3.28 0.10 27.63 0.60 26.66 723 248 154 ST14 10.14 2.40 6.13 2.80 2.32 0.15 3.46 0.12 27.96 0.65 26.55 654 273 93 ST18 9.76 1.89 5.10 2.47 2.02 0.10 3.85 0.09 26.74 0.50 23.39 686 272 106 ST19 10.91 1.74 5.57 2.72 1.96 0.11 3.17 0.10 27.30 0.50 23.83 770 275 112 ST20 10.71 1.75 5.63 2.71 2.04 0.44 3.27 0.10 26.28 0.49 22.66 775 284 105 ST21B 11.00 1.74 5.61 2.76 1.93 0.13 2.61 0.09 27.11 0.50 23.78 789 278 112 ST23 9.14 2.49 4.59 2.34 1.87 0.11 3.84 0.09 22.55 0.41 18.50 638 261 92 ST25 10.20 1.15 4.72 2.62 1.56 0.17 2.72 0.09 29.10 0.49 27.06 725 216 161 ST28 9.49 1.55 4.16 2.23 1.37 0.07 2.40 0.08 31.73 0.49 30.00 632 250 202 ST30 10.22 1.42 4.73 2.63 1.58 0.08 2.76 0.08 29.87 0.51 28.41 714 228 177 ST31 9.88 1.33 4.69 2.65 1.75 0.23 3.89 0.09 28.05 0.46 25.70 703 220 134 ST80 6.02 10.22 4.20 1.22 1.17 0.05 2.46 0.07 25.52 0.26 24.19 221 697 104 ST83 3.54 19.01 1.72 0.80 1.48 0.04 2.88 0.11 12.73 0.18 9.99 191 1665 45 ST86 5.53 7.10 1.04 0.88 0.55 0.03 2.34 0.05 34.76 0.13 34.52 247 646 209 ST91 6.23 9.34 3.18 1.44 1.66 0.08 4.19 0.10 20.73 0.32 17.34 387 837 68 ST93 8.01 3.24 3.58 1.38 1.35 0.07 3.31 0.10 34.52 0.48 33.28 394 353 315 ST96 10.26 2.21 6.64 2.97 2.47 0.29 3.92 0.14 25.98 0.60 23.62 731 267 63 ST97 10.00 2.05 6.48 2.88 2.41 0.27 3.99 0.12 25.03 0.60 23.42 705 251 63 ST98 10.14 2.11 6.32 2.88 2.37 0.16 3.82 0.11 26.01 0.61 24.57 716 253 82 E1 Draga 9.70 2.36 5.34 2.63 1.83 0.13 2.89 0.10 28.74 0.59 28.12 715 264 167 E6 Draga 6.79 9.31 5.22 1.85 1.58 0.08 3.08 0.10 20.87 0.39 18.68 426 528 63 E4 8.69 1.98 5.72 2.46 2.50 0.59 5.95 0.14 22.04 0.52 20.16 619 258 51 Isla Irene 9.05 1.97 5.86 2.56 2.42 0.53 5.39 0.12 22.70 0.54 20.78 640 255 51 E13 9.58 1.00 4.41 2.45 1.40 0.12 3.20 0.08 28.96 0.44 27.29 682 185 223 E8 9.88 1.90 5.96 2.76 2.21 0.20 4.11 0.13 24.25 0.54 21.80 716 247 86 E3 9.74 1.98 6.09 2.76 2.35 0.21 4.42 0.11 24.52 0.58 22.82 681 246 66
rsd (%) 0.43 0.37 0.43 0.53 0.43 0.41 0.56 0.64 0.44 0.42 0.48 0.45 0.61
76
Table 8.1.6. Bulk inorganic geochemical data: Al-based elemental log-ratios. Sample log (Ca/Al) log (Fe/Al) log (K /Al) log (Mg/Al) log (Mn/Al) log (Na/Al) log (P/Al) log (Si /Al) log (Ti/Al) log (litho-Si /Al) log (Ba /Al) log (Sr/Al) log (Zr /Al) ST3 -0.40 -0.24 -0.55 -0.59 -1.95 -0.33 -1.82 0.48 -1.22 0.42 1.84 1.57 0.85 ST4 -0.52 -0.22 -0.56 -0.57 -1.60 -0.23 -1.84 0.44 -1.24 0.38 1.85 1.51 0.93 ST5 -0.61 -0.24 -0.56 -0.59 -1.64 -0.20 -1.90 0.42 -1.26 0.37 1.85 1.46 0.88 ST5P -0.70 -0.23 -0.54 -0.66 -1.87 -0.45 -1.93 0.40 -1.27 0.35 1.86 1.37 0.94 ST6 -0.76 -0.25 -0.54 -0.70 -1.96 -0.53 -2.01 0.39 -1.28 0.35 1.86 1.33 0.94 ST7 -0.82 -0.25 -0.54 -0.72 -1.82 -0.56 -2.04 0.38 -1.28 0.35 1.87 1.30 0.95 ST8 -0.89 -0.27 -0.56 -0.74 -1.87 -0.54 -2.08 0.39 -1.29 0.37 1.87 1.26 1.03 ST9 -1.05 -0.30 -0.57 -0.82 -2.15 -0.71 -2.17 0.36 -1.35 0.34 1.87 1.16 1.07 ST10 -1.01 -0.33 -0.59 -0.86 -1.93 -0.55 -2.10 0.44 -1.36 0.42 1.86 1.26 1.31 ST11 -0.72 -0.21 -0.55 -0.66 -1.62 -0.48 -2.01 0.40 -1.23 0.38 1.86 1.37 0.90 ST12 -0.69 -0.25 -0.57 -0.70 -1.80 -0.48 -1.98 0.45 -1.21 0.43 1.87 1.40 1.20 ST14 -0.63 -0.22 -0.56 -0.64 -1.83 -0.47 -1.93 0.44 -1.19 0.42 1.81 1.43 0.96 ST18 -0.71 -0.28 -0.60 -0.68 -1.99 -0.40 -2.03 0.44 -1.29 0.38 1.85 1.44 1.03 ST19 -0.80 -0.29 -0.60 -0.75 -1.99 -0.54 -2.03 0.40 -1.34 0.34 1.85 1.40 1.01 ST20 -0.79 -0.28 -0.60 -0.72 -1.39 -0.52 -2.02 0.39 -1.34 0.33 1.86 1.42 0.99 ST21B -0.80 -0.29 -0.60 -0.75 -1.94 -0.62 -2.08 0.39 -1.35 0.33 1.86 1.40 1.01 ST23 -0.56 -0.30 -0.59 -0.69 -1.93 -0.38 -2.00 0.39 -1.35 0.31 1.84 1.46 1.00 ST25 -0.95 -0.33 -0.59 -0.82 -1.77 -0.57 -2.06 0.46 -1.32 0.42 1.85 1.33 1.20 ST28 -0.79 -0.36 -0.63 -0.84 -2.13 -0.60 -2.08 0.52 -1.29 0.50 1.82 1.42 1.33 ST30 -0.86 -0.33 -0.59 -0.81 -2.12 -0.57 -2.10 0.47 -1.30 0.44 1.84 1.35 1.24 ST31 -0.87 -0.32 -0.57 -0.75 -1.63 -0.40 -2.03 0.45 -1.33 0.42 1.85 1.35 1.13 ST80 0.23 -0.16 -0.69 -0.71 -2.08 -0.39 -1.96 0.63 -1.37 0.60 1.56 2.06 1.24 ST83 0.73 -0.31 -0.65 -0.38 -1.96 -0.09 -1.51 0.56 -1.31 0.45 1.73 2.67 1.11 ST86 0.11 -0.72 -0.80 -1.00 -2.32 -0.37 -2.06 0.80 -1.62 0.80 1.65 2.07 1.58 ST91 0.18 -0.29 -0.64 -0.58 -1.88 -0.17 -1.81 0.52 -1.29 0.44 1.79 2.13 1.04 ST93 -0.39 -0.35 -0.76 -0.77 -2.08 -0.38 -1.90 0.63 -1.22 0.62 1.69 1.64 1.59 ST96 -0.67 -0.19 -0.54 -0.62 -1.55 -0.42 -1.85 0.40 -1.23 0.36 1.85 1.42 0.79 ST97 -0.69 -0.19 -0.54 -0.62 -1.57 -0.40 -1.91 0.40 -1.22 0.37 1.85 1.40 0.80 ST98 -0.68 -0.21 -0.55 -0.63 -1.81 -0.42 -1.98 0.41 -1.22 0.38 1.85 1.40 0.91 E1 Draga -0.61 -0.26 -0.57 -0.73 -1.88 -0.53 -1.97 0.47 -1.21 0.46 1.87 1.43 1.24 E6 Draga 0.14 -0.11 -0.56 -0.63 -1.94 -0.34 -1.85 0.49 -1.24 0.44 1.80 1.89 0.97 E4 -0.64 -0.18 -0.55 -0.54 -1.17 -0.16 -1.80 0.40 -1.22 0.37 1.85 1.47 0.76 Isla Irene -0.66 -0.19 -0.55 -0.57 -1.23 -0.22 -1.88 0.40 -1.23 0.36 1.85 1.45 0.76 E13 -0.98 -0.34 -0.59 -0.83 -1.92 -0.48 -2.07 0.48 -1.34 0.45 1.85 1.29 1.37 E8 -0.72 -0.22 -0.55 -0.65 -1.69 -0.38 -1.89 0.39 -1.26 0.34 1.86 1.40 0.94 E3 -0.69 -0.20 -0.55 -0.62 -1.66 -0.34 -1.97 0.40 -1.23 0.37 1.84 1.40 0.83
77
8.2 Determination of the reference variable distance (km)
In order to evaluate the spatial variability of the investigated variables in function of terrestrial sediment supply, a reference variable was defined, i.e., distance (km). For this purpose, the distance (km) of each sample location towards the nearest river mouth or glacier front was determined with Google Earth, following the red pathway in figures 8.1 and 8.2.
Fig. 8.1. Distance (km) of each sample location towards the nearest river mouth or glacier front along the red pathway, within the Baker Martinez Fjord Complex. Names of the involved rivers (green). (Google Earth)
78
Fig. 8.2. Distance (km) of each sample location towards the nearest river mouth or glacier front along the red pathway, within the channels and fjords west of the SPI. Names of the involved rivers and glaciers (green). (Google Earth)
79
8.3 Organic matter content
For five samples (ST80, ST83, ST86, ST91 and E6 Draga) the bulk organic geochemical analysis at the Stable Isotope Facility (SIF, University of California) yielded incorrect results (explained in section 4.3 ‘Bulk Organic Geochemistry’). Probably, the preparation method was insufficient to fully decarbonate these samples. Since the organic matter content (OM, wt%) of these samples was analyzed by the loss on ignition method, their total organic carbon content (Corg, wt%) was calculated based on the correlation between LOI550 (%) and Corg (wt%) (Fig. 8.3.1 and Table 8.3.1). The organic matter content (OM, wt%) of the surface sediment samples was calculated by multiplying Corg (wt%) by 2.2 (Table 8.3.1) (Bertrand et al., 2012b).
Fig. 8.3.1. Organic carbon content (Corg , wt%) versus LOI550 (wt%).
Table 8.3.1.
Sample LOI550 (%) Corg (wt%) OM (wt%) ST80 20.69 6.11 13.44 ST83 5.14 0.80 1.76 ST86 15.40 4.30 9.47 ST91 10.58 2.66 5.85 E6 Draga 7.44 1.58 3.49
LOI550 (%) = 2.928398942 * Corg (wt%) + 2.796721834
80
8.4 Correlation matrices
1
.482*
0.208
0.215
0.001
0.000
0.001
0.000
0.046
0.087
0.000
0.000
0.000
0.000
0.416
0.140
0.098
0.354
0.354
0.172
0.145
0.268
0.033
0.356
0.960
0.009
0.888
0.425
0.139
0.010
0.011
0.000
0.050
0.329
0.157
0.241
0.779
0.049
-.335*
-.384*
-.422*
.770**
.733**
.693**
-0.289
-0.280
-0.172
-0.171
-0.024
-.510**
-.546**
-.680**
-.706**
-.787**
-.724**
log(Zr/Al)
1
.357*
.442*
.438*
.421*
.420*
0.208
0.215
0.000
0.001
0.122
0.000
0.000
0.000
0.153
0.033
0.000
0.281
0.000
0.000
0.003
0.003
0.378
0.013
0.003
0.000
0.000
0.000
0.705
0.065
0.022
0.000
0.011
0.000
0.012
.529**
.684**
.679**
.628**
.978**
.508**
.509**
.824**
.944**
.565**
.729**
-0.262
-0.243
-0.184
-0.164
-.733**
-.582**
-.957**
-.508**
-.887**
log(Sr/Al)
1
0.001
0.000
0.000
0.009
0.000
0.155
0.077
0.010
0.560
0.101
0.000
0.064
0.312
0.000
0.000
0.074
0.074
0.326
0.984
0.313
0.073
0.008
0.044
0.000
0.126
0.260
0.000
0.000
0.000
0.000
0.020
-.338*
-.392*
.430**
.426**
.817**
.760**
.439**
-0.242
-0.298
-0.326
-0.004
-0.187
-0.327
-.510**
-.733**
-.811**
-.859**
-.737**
-.581**
-.770**
-.835**
-.776**
-.702**
log(Ba/Al)
1
.354*
0.000
0.001
0.000
0.002
0.000
0.971
0.006
0.500
0.116
0.001
0.008
0.000
0.000
0.001
0.000
0.883
0.883
0.028
0.664
0.081
0.247
0.888
0.387
0.581
0.095
0.037
0.000
0.000
0.000
0.001
0.000
0.273
0.190
.770**
.529**
.972**
.515**
.723**
.932**
.528**
.588**
-0.028
-0.214
-0.026
-0.151
-.811**
-.498**
-.519**
-.433**
-.858**
-.636**
-.625**
-.557**
Si/Al)
log(litho-
1
.380*
0.001
0.122
0.009
0.002
0.001
0.022
0.278
0.186
0.002
0.000
0.000
0.000
0.305
0.130
0.257
0.924
0.924
0.018
0.009
0.879
0.029
0.925
0.340
0.166
0.703
0.185
0.884
0.025
0.000
0.000
0.979
0.005
0.039
0.126
-.345*
.430**
.495**
.631**
.636**
.800**
-0.262
-0.176
-0.018
-0.018
-0.066
-0.230
-0.264
-.546**
-.498**
-.527**
-.459**
-.697**
-.691**
log(Ti/Al)
1
0.000
0.000
0.000
0.000
0.001
0.376
0.152
0.149
0.245
0.001
0.067
0.000
0.000
0.000
0.000
0.427
0.148
0.427
0.481
0.131
0.718
0.422
0.150
0.037
0.076
0.299
0.002
0.023
0.000
0.000
0.001
0.000
0.064
0.317
-.354*
-.379*
.733**
.684**
.972**
.659**
.515**
.717**
.932**
.532**
.692**
-0.309
-0.148
-0.067
, correlation is significant at is , at correlation significant
-.859**
-.527**
-.532**
-.894**
-.628**
-.757**
log(Si/Al)
1
.380*
.333*
.429*
.428*
0.046
0.000
0.155
0.971
0.006
0.022
0.376
0.152
0.000
0.055
0.323
0.000
0.961
0.008
0.047
0.000
0.000
0.016
0.016
0.158
0.001
0.016
0.000
0.000
0.000
0.165
0.237
0.458
0.895
0.348
0.161
0.002
0.143
0.253
-.335*
-.429*
.679**
.842**
.849**
.700**
.548**
.722**
.642**
.496**
-0.242
-0.260
-0.149
-0.023
-.664**
-.688**
log(P/Al)
1
.395*
.455*
0.087
0.000
0.077
0.500
0.116
0.278
0.186
0.149
0.245
0.000
0.017
0.000
0.566
0.305
0.176
0.000
0.000
0.010
0.010
0.136
0.001
0.010
0.000
0.000
0.001
0.201
0.218
0.805
0.050
0.455
0.128
0.459
0.127
0.000
0.144
0.252
.628**
.842**
.721**
.649**
.456**
.580**
.612**
.539**
.559**
-0.289
-0.298
-0.099
-0.274
-.706**
-.456**
-.590**
log(Na/Al)
1
.395*
0.000
0.153
0.010
0.001
0.002
0.001
0.055
0.323
0.017
0.001
0.000
0.001
0.179
0.210
0.214
0.596
0.099
0.596
0.177
0.055
0.348
0.611
0.095
0.307
0.178
0.371
0.134
0.301
0.177
0.012
0.001
0.094
0.109
0.165
-.475*
.426**
.495**
.539**
.555**
.537**
-0.243
-0.229
-0.099
-0.249
-0.153
-0.258
-0.283
-0.271
-0.240
-.680**
-.519**
-.532**
-.528**
log(Mn/Al)
1
.357*
.419*
.403*
.392*
.380*
0.000
0.033
0.560
0.101
0.008
0.000
0.067
0.000
0.000
0.001
0.011
0.000
0.015
0.071
0.093
0.307
0.093
0.026
0.006
0.095
0.306
0.002
0.001
0.020
0.022
0.033
0.007
0.997
0.001
0.414
0.141
0.669
0.075
-.399*
-.411*
.631**
.849**
.721**
.539**
.672**
.480**
.546**
-0.309
-0.304
-0.307
-.706**
-.433**
-.505**
-.440**
log(Mg/Al)
1
.419*
0.000
0.000
0.000
0.000
0.000
0.000
0.961
0.008
0.566
0.000
0.011
0.000
0.001
0.000
0.360
0.360
0.170
0.258
0.999
0.000
0.361
0.088
0.293
0.194
0.015
0.218
0.210
0.000
0.000
0.003
0.000
0.041
-.407*
-.346*
.817**
.636**
.555**
.712**
.602**
-0.099
-0.170
-0.210
-0.170
-0.222
-.787**
-.582**
-.858**
-.894**
-.520**
-.642**
-.870**
-.476**
-.599**
tailed tailed test (p) of significance
log(K/Al)
-
1
.333*
0.000
0.281
0.064
0.312
0.000
0.000
0.000
0.047
0.305
0.176
0.001
0.000
0.000
0.649
0.206
0.216
0.950
0.950
0.012
0.005
0.235
0.220
0.944
0.995
0.913
0.019
0.640
0.194
0.221
0.015
0.000
0.310
0.174
0.058
0.348
-.465*
.800**
.537**
.672**
.712**
-0.184
-0.079
-0.012
-0.013
-0.001
-0.082
-0.319
-0.164
two
-.724**
-.636**
-.628**
-.492**
-.674**
log(Fe/Al)
1
.403*
.416*
.459*
.412*
0.416
0.140
0.000
0.000
0.001
0.305
0.000
0.000
0.000
0.179
0.015
0.001
0.649
0.000
0.006
0.006
0.151
0.020
0.006
0.000
0.000
0.000
0.825
0.038
0.016
0.000
0.002
0.000
0.014
.978**
.515**
.659**
.700**
.649**
.481**
.482**
.800**
.915**
.556**
.501**
.761**
-0.176
-0.229
-0.079
-0.264
-.737**
-.520**
-.950**
-.481**
-.856**
log(Ca/Al)
1
0.098
0.000
0.000
0.000
0.130
0.257
0.000
0.000
0.000
0.210
0.214
0.071
0.000
0.206
0.216
0.000
0.010
0.010
0.556
0.110
0.025
0.009
0.000
0.000
0.000
0.834
0.036
0.011
0.000
0.005
0.000
0.020
-.403*
-.480*
-.390*
.760**
.602**
.458**
.804**
-0.280
-0.304
-.957**
-.625**
-.757**
-.664**
-.706**
-.950**
-.458**
-.459**
-.762**
-.878**
-.637**
-.457**
-.774**
Al (wt%) Al
1
Fm
.429*
.408*
(%)
0.354
0.003
0.074
0.883
0.924
0.427
0.148
0.016
0.010
0.596
0.099
0.093
0.307
0.360
0.950
0.006
0.010
0.000
0.023
0.000
0.000
0.000
0.000
0.000
0.001
0.988
0.840
0.038
0.114
0.289
0.000
0.000
.508**
.456**
.481**
.872**
.668**
.646**
.560**
.850**
.816**
-0.172
-0.326
-0.028
-0.018
-0.170
-0.012
-0.003
-.458**
-.678**
1.000**
-1.000**
1
0.354
0.172
0.003
0.074
0.326
0.883
0.028
0.924
0.018
0.427
0.016
0.010
0.596
0.093
0.360
0.170
0.950
0.012
0.006
0.010
0.000
0.023
0.000
0.000
0.000
0.000
0.000
0.001
0.988
0.003
0.840
0.114
0.000
0.000
-.429*
-.408*
.458**
.678**
-0.148
-0.099
-0.307
-0.038
-0.289
-.508**
-.456**
-.481**
-.872**
-.668**
-.646**
-.560**
-.850**
-.816**
FT (%) FT
-1.000**
-1.000**
1
.408*
.377*
.407*
0.145
0.268
0.378
0.984
0.664
0.081
0.009
0.481
0.131
0.158
0.136
0.177
0.026
0.258
0.005
0.151
0.556
0.110
0.023
0.023
0.037
0.023
0.307
0.239
0.218
0.507
0.126
0.243
0.216
0.508
0.146
0.810
0.045
0.896
0.252
0.212
0.002
-.399*
-.408*
.536**
-0.164
-0.004
-0.260
-0.274
-0.249
-0.210
-0.264
-0.193
-0.025
-.459**
-.492**
Atomic N/C Atomic
1
.442*
.416*
.377*
(‰)
0.033
0.013
0.313
0.247
0.879
0.029
0.718
0.001
0.001
0.055
0.348
0.006
0.999
0.000
0.235
0.220
0.020
0.025
0.000
0.000
0.037
0.000
0.000
0.001
0.008
0.001
0.956
0.543
0.517
0.121
0.000
0.000
-.384*
-.403*
.548**
.580**
.480**
.872**
.870**
.557**
.475**
.557**
.724**
.660**
-0.187
-0.214
-0.067
-0.012
-0.113
δ15N
-.872**
-.604**
1
.428*
.455*
.407*
(‰)
0.356
0.003
0.073
0.888
0.925
0.422
0.150
0.016
0.010
0.611
0.095
0.095
0.306
0.361
0.944
0.006
0.009
0.000
0.000
0.023
0.000
0.000
0.000
0.000
0.001
0.995
0.840
0.038
0.115
0.289
0.000
0.000
.509**
.482**
.870**
.668**
.647**
.560**
.850**
.818**
-0.171
-0.327
-0.026
-0.018
-0.170
-0.013
-0.002
δ13C
-.459**
-.679**
1.000**
-1.000**
Pearson correlation coefficient (r), coefficient correlation Pearson
1
0.960
0.009
0.000
0.008
0.387
0.340
0.166
0.037
0.000
0.000
0.307
0.178
0.002
0.088
0.293
0.995
0.000
0.000
0.000
0.000
0.307
0.000
0.000
0.000
0.000
0.020
0.532
0.196
0.261
0.000
0.008
-.354*
-.393*
.439**
.804**
.678**
-0.151
-0.001
-0.193
-0.126
-0.224
-0.195
-.887**
-.688**
-.590**
-.505**
-.856**
-.678**
-.604**
-.679**
-.969**
-.951**
-.580**
-.444**
(wt%)
Lithogeni
c Fractionc
1
.410*
0.888
0.000
0.044
0.581
0.095
0.703
0.076
0.299
0.000
0.000
0.371
0.001
0.194
0.913
0.019
0.000
0.000
0.000
0.000
0.239
0.218
0.001
0.000
0.000
0.000
0.009
0.936
0.424
0.137
0.392
0.147
0.000
0.014
-.338*
.824**
.722**
.612**
.546**
.800**
.668**
.557**
.668**
.886**
.430**
.575**
-0.024
-0.066
-0.153
-0.222
-0.016
-.762**
-.668**
-.969**
(wt%)
Matter Matter
Organic
1
.354*
.392*
.402*
.362*
0.425
0.139
0.000
0.000
0.037
0.185
0.002
0.000
0.001
0.134
0.020
0.015
0.640
0.000
0.000
0.000
0.000
0.507
0.126
0.008
0.000
0.000
0.000
0.554
0.103
0.119
0.307
0.017
0.081
0.299
0.000
0.035
-.407*
.944**
.515**
.642**
.539**
.915**
.646**
.475**
.647**
.886**
.610**
-0.230
-0.258
-0.082
-.581**
-.878**
-.646**
-.951**
(wt%)
CaCO3
1
.380*
.427*
0.010
0.705
0.065
0.126
0.260
0.000
0.884
0.025
0.023
0.165
0.237
0.201
0.218
0.301
0.177
0.022
0.218
0.210
0.194
0.221
0.825
0.038
0.834
0.036
0.001
0.001
0.243
0.216
0.001
0.001
0.020
0.009
0.554
0.103
0.036
0.010
0.177
0.063
0.265
0.010
-.422*
-.379*
-.393*
-.405*
-.423*
.560**
.557**
.560**
.430**
-0.230
-.557**
-.560**
(wt%)
Bio-Opal
1
.482*
.438*
.459*
.466*
0.011
0.022
0.000
0.000
0.000
0.000
0.458
0.805
0.050
0.012
0.033
0.000
0.015
0.016
0.011
0.988
0.988
0.003
0.508
0.146
0.956
0.995
0.532
0.936
0.119
0.307
0.036
0.000
0.005
0.014
0.203
0.253
-.475*
-.411*
-.465*
-.480*
-.405*
.723**
.717**
.838**
.521**
-0.149
-0.003
-0.012
-0.002
-0.126
-0.016
-.770**
-.697**
-.642**
(wt%)
IRD150
1
.402*
0.000
0.000
0.000
0.000
0.000
0.000
0.895
0.455
0.128
0.001
0.007
0.000
0.000
0.000
0.000
0.840
0.038
0.840
0.810
0.045
0.543
0.840
0.038
0.196
0.424
0.137
0.017
0.010
0.000
0.002
0.000
0.183
0.231
-.423*
.693**
.565**
.932**
.932**
.556**
.838**
.496**
.653**
-0.023
-0.038
-0.113
-0.224
-.835**
-.691**
-.528**
-.440**
-.870**
-.674**
-.637**
Size (µm) Size
free Grainfree
Mean IRD Mean
1
.421*
0.050
0.329
0.011
0.000
0.001
0.979
0.005
0.001
0.348
0.161
0.459
0.127
0.094
0.997
0.001
0.003
0.310
0.174
0.002
0.005
0.114
0.289
0.114
0.896
0.517
0.121
0.115
0.289
0.261
0.392
0.147
0.081
0.299
0.177
0.005
0.002
0.006
0.143
0.253
.528**
.532**
.501**
.521**
.496**
.452**
-0.283
-0.289
-0.025
-0.195
-0.230
-.776**
-.476**
-.457**
(10-6SI)
Magnetic
Susceptibility
Mass-specific
1
.466*
0.157
0.241
0.000
0.000
0.000
0.039
0.000
0.002
0.000
0.109
0.414
0.141
0.000
0.058
0.000
0.000
0.000
0.000
0.252
0.212
0.000
0.000
0.000
0.000
0.000
0.063
0.265
0.014
0.000
0.006
0.000
-.345*
.729**
.588**
.692**
.496**
.559**
.761**
.850**
.724**
.850**
.575**
.610**
.653**
.452**
.636**
-0.271
-0.319
-.702**
-.599**
-.774**
-.850**
-.580**
e (km)e
Distanc
1
.420*
.412*
.410*
.362*
.427*
0.779
0.049
0.012
0.020
0.273
0.190
0.126
0.064
0.317
0.143
0.253
0.144
0.252
0.165
0.669
0.075
0.041
0.348
0.014
0.020
0.000
0.000
0.002
0.000
0.000
0.008
0.014
0.035
0.010
0.203
0.253
0.183
0.231
0.143
0.253
0.000
-.392*
-.346*
-.390*
.816**
.536**
.660**
.818**
.636**
-0.264
-0.240
-0.164
(psu)
-.816**
-.444**
Salinity
Surface
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
p
r
0.05 level (*), correlation is significant at the 0.01 level (**). level 0.01 the at significant is correlation level (*), 0.05
SI)
-6
N (‰) N
C (‰) C
(%)
(%)
m
T
15
13
log(Zr/Al)
log(Sr/Al)
log(Ba/Al)
log(litho-Si/Al)
log(Ti/Al)
log(Si/Al)
log(P/Al)
log(Na/Al)
log(Mn/Al)
log(Mg/Al)
log(K/Al)
log(Fe/Al)
log(Ca/Al)
Al (wt%) Al
F
F
Atomic N/C Atomic
δ
δ
(wt%)
Fraction
Lithogenic
Matter (wt%) Matter
Organic
CaCO3(wt%)
(wt%)
Bio-Opal
IRD150(wt%)
Size (µm) Size
free Grainfree
Mean IRD Mean
(10
Susceptibility Susceptibility
Magnetic
Mass-specific
Distance(km)
Salinity (psu)Salinity
Surface
Table 8.4.1. Correlation matrix of the Table entire Correlation area. 8.4.1. study the
81
Table 8.4.2. Correlation matrix of the Baker Channel. Pearson correlation coefficient (r), two-tailed test of significance (p), correlation is significant at the 0.05 level (*), correlation is significant at the 0.01 level (**).
Mass-specific Mean Surface Magnetic IRD free Organic Lithogeni Salinity Distanc Susceptibility Grain IRD150 Bio-Opal CaCO3 Matter c Fraction δ13C δ15N Atomic Fm log(litho- (psu) e (km) (10-6 SI) Size (µm) (wt%) (wt%) (wt%) (wt%) (wt%) (‰) (‰) N/C FT (%) (%) Al (wt%) log(Ca/Al) log(Fe/Al) log(K/Al) log(Mg/Al) log(Mn/Al) log(Na/Al) log(P/Al) log(Si/Al) log(Ti/Al) Si/Al) log(Ba/Al) log(Sr/Al) log(Zr/Al) Surface r 1 .798** 0.571 -0.400 0.028 .621* 0.320 0.444 -0.469 .843** .838** 0.021 -.841** .841** -0.357 .617* .738** .837** .766** 0.289 0.444 .668* -0.074 .809** -0.328 -0.302 0.543 -.884** Salinity (psu) p 0.002 0.053 0.198 0.947 0.031 0.310 0.148 0.124 0.001 0.001 0.950 0.001 0.001 0.254 0.032 0.006 0.001 0.004 0.363 0.148 0.018 0.819 0.001 0.298 0.340 0.068 0.000 Distance r .798** 1 .727** -0.260 0.209 .775** 0.437 .783** -.689* .978** .982** 0.021 -.978** .978** -.615* .805** .813** .685* .947** 0.365 .702* .889** 0.240 .944** -0.078 -0.516 .746** -.869** (km) p 0.002 0.007 0.415 0.620 0.000 0.156 0.003 0.013 0.000 0.000 0.951 0.000 0.000 0.033 0.002 0.001 0.014 0.000 0.243 0.011 0.000 0.453 0.000 0.809 0.086 0.005 0.000 Mass-specific r 0.571 .727** 1 0.295 0.547 0.452 .919** .874** -.978** .921** .881** -0.022 -.921** .921** -.850** .980** .886** 0.292 .610* 0.040 0.504 .695* 0.564 .701* 0.392 -.901** .960** -0.467 Magnetic p 0.053 0.007 0.352 0.160 0.140 0.000 0.000 0.000 0.000 0.000 0.948 0.000 0.000 0.000 0.000 0.000 0.357 0.035 0.903 0.095 0.012 0.056 0.011 0.208 0.000 0.000 0.126 Susceptibility (10-6 SI) Mean IRD r -0.400 -0.260 0.295 1 .758* -0.162 0.494 0.327 -0.409 -0.593 -0.581 -0.533 0.595 -0.595 -0.555 0.251 -0.058 -.685* -0.304 -0.232 0.099 0.022 .832** -0.232 .933** -.626* 0.368 .651* free Grain p 0.198 0.415 0.352 0.029 0.616 0.103 0.300 0.187 0.054 0.061 0.091 0.053 0.053 0.061 0.432 0.857 0.014 0.337 0.469 0.759 0.945 0.001 0.469 0.000 0.029 0.240 0.022 Size (µm) IRD150 (wt%) r 0.028 0.209 0.547 .758* 1 0.425 .740* 0.590 -0.574 0.265 0.148 -0.313 -0.262 0.262 -0.617 0.428 -0.097 -0.243 0.152 -0.053 0.274 0.404 .872** 0.280 .915** -0.666 0.521 0.050 p 0.947 0.620 0.160 0.029 0.294 0.036 0.123 0.136 0.527 0.726 0.451 0.531 0.531 0.103 0.290 0.818 0.562 0.719 0.901 0.511 0.321 0.005 0.502 0.001 0.071 0.185 0.906 Bio-Opal r .621* .775** 0.452 -0.162 0.425 1 0.135 .658* -0.467 .856** .877** -0.251 -.854** .854** -0.564 .583* 0.520 0.494 .872** 0.492 .786** .914** 0.335 .825** -0.002 -0.367 .578* -.679* (wt%) p 0.031 0.000 0.140 0.616 0.294 0.675 0.020 0.126 0.001 0.000 0.456 0.001 0.001 0.056 0.047 0.083 0.102 0.000 0.105 0.002 0.000 0.287 0.001 0.996 0.241 0.049 0.015 CaCO3 (wt%) r 0.320 0.437 .919** 0.494 .740* 0.135 1 .718** -.933** .748** .685* -0.074 -.747** .747** -.778** .866** .775** 0.027 0.307 -0.135 0.297 0.437 0.556 0.394 0.488 -.918** .866** -0.164 p 0.310 0.156 0.000 0.103 0.036 0.675 0.009 0.000 0.008 0.020 0.828 0.008 0.008 0.003 0.000 0.003 0.935 0.331 0.675 0.349 0.156 0.061 0.206 0.107 0.000 0.000 0.611 Organic r 0.444 .783** .874** 0.327 0.590 .658* .718** 1 -.908** .725* .695* -0.091 -.725* .725* -.884** .896** .680* 0.169 .677* 0.081 .662* .769** .726** .702* 0.514 -.859** .898** -0.434 Matter (wt%) p 0.148 0.003 0.000 0.300 0.123 0.020 0.009 0.000 0.012 0.018 0.790 0.012 0.012 0.000 0.000 0.015 0.600 0.016 0.802 0.019 0.003 0.008 0.011 0.087 0.000 0.000 0.159 Lithogenic r -0.469 -.689* -.978** -0.409 -0.574 -0.467 -.933** -.908** 1 -.819** -.801** 0.157 .818** -.818** .902** -.971** -.837** -0.161 -0.570 -0.021 -0.541 -.695* -.656* -.629* -0.485 .957** -.971** 0.365 Fraction p 0.124 0.013 0.000 0.187 0.136 0.126 0.000 0.000 0.002 0.003 0.645 0.002 0.002 0.000 0.000 0.001 0.618 0.053 0.949 0.069 0.012 0.021 0.028 0.110 0.000 0.000 0.244 (wt%) δ13C (‰) r .843** .978** .921** -0.593 0.265 .856** .748** .725* -.819** 1 .986** 0.141 -1.000**1.000** -0.501 .923** .889** .764** .924** 0.424 .624* .823** 0.012 .922** -0.332 -0.405 .805** -.930** p 0.001 0.000 0.000 0.054 0.527 0.001 0.008 0.012 0.002 0.000 0.679 0.000 0.000 0.116 0.000 0.000 0.006 0.000 0.194 0.040 0.002 0.973 0.000 0.318 0.216 0.003 0.000 δ15N (‰) r .838** .982** .881** -0.581 0.148 .877** .685* .695* -.801** .986** 1 0.069 -.986** .986** -0.523 .923** .915** .774** .945** 0.513 .679* .854** 0.013 .928** -0.346 -0.414 .817** -.912** p 0.001 0.000 0.000 0.061 0.726 0.000 0.020 0.018 0.003 0.000 0.840 0.000 0.000 0.098 0.000 0.000 0.005 0.000 0.107 0.021 0.001 0.970 0.000 0.297 0.206 0.002 0.000 Atomic N/C r 0.021 0.021 -0.022 -0.533 -0.313 -0.251 -0.074 -0.091 0.157 0.141 0.069 1 -0.142 0.142 0.492 -0.166 0.007 0.158 -0.131 -0.325 -0.397 -0.359 -0.508 -0.155 -0.386 0.517 -0.324 -0.208 p 0.950 0.951 0.948 0.091 0.451 0.456 0.828 0.790 0.645 0.679 0.840 0.677 0.677 0.124 0.625 0.985 0.643 0.701 0.330 0.227 0.278 0.110 0.649 0.240 0.104 0.332 0.540
FT (%) r -.841** -.978** -.921** 0.595 -0.262 -.854** -.747** -.725* .818** -1.000** -.986** -0.142 1-1.000** 0.501 -.923** -.889** -.764** -.925** -0.425 -.625* -.822** -0.011 -.923** 0.333 0.405 -.805** .931** p 0.001 0.000 0.000 0.053 0.531 0.001 0.008 0.012 0.002 0.000 0.000 0.677 0.000 0.117 0.000 0.000 0.006 0.000 0.192 0.040 0.002 0.975 0.000 0.317 0.217 0.003 0.000
Fm (%) r .841** .978** .921** -0.595 0.262 .854** .747** .725* -.818** 1.000** .986** 0.142 -1.000** 1 -0.501 .923** .889** .764** .925** 0.425 .625* .822** 0.011 .923** -0.333 -0.405 .805** -.931** p 0.001 0.000 0.000 0.053 0.531 0.001 0.008 0.012 0.002 0.000 0.000 0.677 0.000 0.117 0.000 0.000 0.006 0.000 0.192 0.040 0.002 0.975 0.000 0.317 0.217 0.003 0.000 Al (wt%) r -0.357 -.615* -.850** -0.555 -0.617 -0.564 -.778** -.884** .902** -0.501 -0.523 0.492 0.501 -0.501 1 -.897** -.680* 0.031 -.601* -0.292 -.771** -.791** -.841** -.625* -.653* .941** -.951** 0.229 p 0.254 0.033 0.000 0.061 0.103 0.056 0.003 0.000 0.000 0.116 0.098 0.124 0.117 0.117 0.000 0.015 0.924 0.039 0.356 0.003 0.002 0.001 0.030 0.021 0.000 0.000 0.474 log(Ca/Al) r .617* .805** .980** 0.251 0.428 .583* .866** .896** -.971** .923** .923** -0.166 -.923** .923** -.897** 1 .914** 0.325 .728** 0.201 .654* .811** 0.575 .783** 0.354 -.898** .988** -0.533 p 0.032 0.002 0.000 0.432 0.290 0.047 0.000 0.000 0.000 0.000 0.000 0.625 0.000 0.000 0.000 0.000 0.302 0.007 0.532 0.021 0.001 0.050 0.003 0.258 0.000 0.000 0.074 log(Fe/Al) r .738** .813** .886** -0.058 -0.097 0.520 .775** .680* -.837** .889** .915** 0.007 -.889** .889** -.680* .914** 1 0.562 .756** 0.280 0.541 .746** 0.218 .800** -0.015 -.695* .862** -.692* p 0.006 0.001 0.000 0.857 0.818 0.083 0.003 0.015 0.001 0.000 0.000 0.985 0.000 0.000 0.015 0.000 0.057 0.004 0.379 0.069 0.005 0.496 0.002 0.963 0.012 0.000 0.013 log(K/Al) r .837** .685* 0.292 -.685* -0.243 0.494 0.027 0.169 -0.161 .764** .774** 0.158 -.764** .764** 0.031 0.325 0.562 1 .627* 0.211 0.159 0.441 -0.397 .687* -.617* 0.073 0.207 -.863** p 0.001 0.014 0.357 0.014 0.562 0.102 0.935 0.600 0.618 0.006 0.005 0.643 0.006 0.006 0.924 0.302 0.057 0.029 0.510 0.623 0.152 0.201 0.014 0.033 0.822 0.519 0.000 log(Mg/Al) r .766** .947** .610* -0.304 0.152 .872** 0.307 .677* -0.570 .924** .945** -0.131 -.925** .925** -.601* .728** .756** .627* 1 .614* .837** .925** 0.207 .961** -0.124 -0.429 .691* -.859** p 0.004 0.000 0.035 0.337 0.719 0.000 0.331 0.016 0.053 0.000 0.000 0.701 0.000 0.000 0.039 0.007 0.004 0.029 0.034 0.001 0.000 0.519 0.000 0.700 0.164 0.013 0.000 log(Mn/Al) r 0.289 0.365 0.040 -0.232 -0.053 0.492 -0.135 0.081 -0.021 0.424 0.513 -0.325 -0.425 0.425 -0.292 0.201 0.280 0.211 .614* 1 .758** 0.573 0.046 0.505 -0.154 -0.010 0.237 -0.355 p 0.363 0.243 0.903 0.469 0.901 0.105 0.675 0.802 0.949 0.194 0.107 0.330 0.192 0.192 0.356 0.532 0.379 0.510 0.034 0.004 0.052 0.887 0.094 0.632 0.975 0.458 0.258 log(Na/Al) r 0.444 .702* 0.504 0.099 0.274 .786** 0.297 .662* -0.541 .624* .679* -0.397 -.625* .625* -.771** .654* 0.541 0.159 .837** .758** 1 .890** 0.531 .749** 0.254 -0.547 .700* -0.466 p 0.148 0.011 0.095 0.759 0.511 0.002 0.349 0.019 0.069 0.040 0.021 0.227 0.040 0.040 0.003 0.021 0.069 0.623 0.001 0.004 0.000 0.076 0.005 0.425 0.065 0.011 0.127 log(P/Al) r .668* .889** .695* 0.022 0.404 .914** 0.437 .769** -.695* .823** .854** -0.359 -.822** .822** -.791** .811** .746** 0.441 .925** 0.573 .890** 1 0.487 .917** 0.162 -.620* .818** -.658* p 0.018 0.000 0.012 0.945 0.321 0.000 0.156 0.003 0.012 0.002 0.001 0.278 0.002 0.002 0.002 0.001 0.005 0.152 0.000 0.052 0.000 0.109 0.000 0.616 0.032 0.001 0.020 log(Si/Al) r -0.074 0.240 0.564 .832** .872** 0.335 0.556 .726** -.656* 0.012 0.013 -0.508 -0.011 0.011 -.841** 0.575 0.218 -0.397 0.207 0.046 0.531 0.487 1 0.263 .937** -.794** .672* 0.220 p 0.819 0.453 0.056 0.001 0.005 0.287 0.061 0.008 0.021 0.973 0.970 0.110 0.975 0.975 0.001 0.050 0.496 0.201 0.519 0.887 0.076 0.109 0.409 0.000 0.002 0.017 0.492 log(Ti/Al) r .809** .944** .701* -0.232 0.280 .825** 0.394 .702* -.629* .922** .928** -0.155 -.923** .923** -.625* .783** .800** .687* .961** 0.505 .749** .917** 0.263 1 -0.047 -0.480 .739** -.848** p 0.001 0.000 0.011 0.469 0.502 0.001 0.206 0.011 0.028 0.000 0.000 0.649 0.000 0.000 0.030 0.003 0.002 0.014 0.000 0.094 0.005 0.000 0.409 0.884 0.115 0.006 0.000 log(litho-Si/Al) r -0.328 -0.078 0.392 .933** .915** -0.002 0.488 0.514 -0.485 -0.332 -0.346 -0.386 0.333 -0.333 -.653* 0.354 -0.015 -.617* -0.124 -0.154 0.254 0.162 .937** -0.047 1 -.667* 0.461 0.497 p 0.298 0.809 0.208 0.000 0.001 0.996 0.107 0.087 0.110 0.318 0.297 0.240 0.317 0.317 0.021 0.258 0.963 0.033 0.700 0.632 0.425 0.616 0.000 0.884 0.018 0.132 0.100 log(Ba/Al) r -0.302 -0.516 -.901** -.626* -0.666 -0.367 -.918** -.859** .957** -0.405 -0.414 0.517 0.405 -0.405 .941** -.898** -.695* 0.073 -0.429 -0.010 -0.547 -.620* -.794** -0.480 -.667* 1 -.936** 0.136 p 0.340 0.086 0.000 0.029 0.071 0.241 0.000 0.000 0.000 0.216 0.206 0.104 0.217 0.217 0.000 0.000 0.012 0.822 0.164 0.975 0.065 0.032 0.002 0.115 0.018 0.000 0.673 log(Sr/Al) r 0.543 .746** .960** 0.368 0.521 .578* .866** .898** -.971** .805** .817** -0.324 -.805** .805** -.951** .988** .862** 0.207 .691* 0.237 .700* .818** .672* .739** 0.461 -.936** 1 -0.431 p 0.068 0.005 0.000 0.240 0.185 0.049 0.000 0.000 0.000 0.003 0.002 0.332 0.003 0.003 0.000 0.000 0.000 0.519 0.013 0.458 0.011 0.001 0.017 0.006 0.132 0.000 0.161 log(Zr/Al) r -.884** -.869** -0.467 .651* 0.050 -.679* -0.164 -0.434 0.365 -.930** -.912** -0.208 .931** -.931** 0.229 -0.533 -.692* -.863** -.859** -0.355 -0.466 -.658* 0.220 -.848** 0.497 0.136 -0.431 1 p 0.000 0.000 0.126 0.022 0.906 0.015 0.611 0.159 0.244 0.000 0.000 0.540 0.000 0.000 0.474 0.074 0.013 0.000 0.000 0.258 0.127 0.020 0.492 0.000 0.100 0.673 0.161
82
Table 8.4.3. Correlation matrix of the Martinez Channel. Pearson correlation coefficient (r), two-tailed test of significance (p), correlation is significant at the 0.05 level (*), correlation is significant at the 0.01 level (**). Mean Mass-specific IRD free Surface Magnetic Grain CaCO Organic Lithogenic Salinity Distance Susceptibility Size IRD150 Bio-Opal 3 Matter Fraction δ13C δ15N Atomic Fm log(litho- (psu) (km) (10-6 SI) (µm) (wt%) (wt%) (wt%) (wt%) (wt%) (‰) (‰) N/C FT (%) (%) Al (wt%) log(Ca/Al) log(Fe/Al) log(K/Al) log(Mg/Al) log(Mn/Al) log(Na/Al) log(P/Al) log(Si/Al) log(Ti/Al) Si/Al) log(Ba/Al) log(Sr/Al) log(Zr/Al) Surface r 1 .764* -0.660 -.761* 0.853 .889** .798* 0.408 -.920** .786* .672* 0.380 -.787* .787* 0.164 -0.176 .738* .770* 0.594 0.540 0.327 0.577 -.692* -0.467 -.735* -0.260 0.109 -.746* Salinity (psu) p 0.017 0.053 0.017 0.066 0.001 0.018 0.275 0.001 0.012 0.047 0.314 0.012 0.012 0.673 0.650 0.023 0.015 0.092 0.134 0.390 0.104 0.039 0.205 0.024 0.500 0.779 0.021
Distance (km) r .764* 1 -.781** -.783** .961** .918** .778* 0.363 -.924** .966** .873** 0.223 -.965** .965** -0.376 -0.099 .901** .824** .833** .796** .736* .819** -.796** -0.590 -.838** -0.129 0.396 -.861** p 0.017 0.008 0.007 0.009 0.000 0.014 0.302 0.000 0.000 0.001 0.535 0.000 0.000 0.284 0.786 0.000 0.003 0.003 0.006 0.015 0.004 0.006 0.072 0.002 0.723 0.257 0.001 Mass-specific r -0.660 -.781** 1 .959** -0.712 -.852** -0.448 0.207 .750* -.741* -.656* 0.069 .737* -.737* 0.217 0.412 -.955** -.853** -.862** -.710* -.697* -0.456 .959** 0.603 .948** 0.400 -0.082 .970** Magnetic p 0.053 0.008 0.000 0.178 0.002 0.227 0.567 0.020 0.014 0.040 0.849 0.015 0.015 0.547 0.237 0.000 0.002 0.001 0.021 0.025 0.186 0.000 0.065 0.000 0.252 0.822 0.000 Susceptibility (10-6 SI) Mean IRD free r -.761* -.783** .959** 1 -0.620 -.873** -0.441 0.119 .752* -.745* -.687* -0.119 .740* -.740* 0.095 0.316 -.964** -.912** -.847** -.641* -0.606 -0.495 .933** 0.502 .928** 0.488 -0.102 .983** Grain Size p 0.017 0.007 0.000 0.264 0.001 0.234 0.744 0.019 0.013 0.028 0.743 0.014 0.014 0.793 0.373 0.000 0.000 0.002 0.046 0.063 0.146 0.000 0.139 0.000 0.152 0.778 0.000 (µm) IRD150 (wt%) r 0.853 .961** -0.712 -0.620 1 .922* .926* .891* -.970** .972** .937* 0.836 -.976** .976** 0.522 -0.178 0.756 0.790 0.569 .933* 0.620 .914* -0.742 -0.729 -0.798 0.169 0.172 -0.714 p 0.066 0.009 0.178 0.264 0.026 0.024 0.042 0.006 0.006 0.019 0.078 0.005 0.005 0.367 0.775 0.140 0.111 0.317 0.020 0.264 0.030 0.151 0.162 0.106 0.786 0.782 0.176 Bio-Opal (wt%) r .889** .918** -.852** -.873** .922* 1 .785* 0.208 -.969** .881** .747* 0.129 -.879** .879** -0.159 -0.207 .908** .846** .818** .684* 0.631 .695* -.836** -0.541 -.894** -0.365 0.266 -.906** p 0.001 0.000 0.002 0.001 0.026 0.012 0.564 0.000 0.001 0.013 0.723 0.001 0.001 0.661 0.566 0.000 0.002 0.004 0.029 0.051 0.026 0.003 0.107 0.000 0.300 0.457 0.000 CaCO3 (wt%) r .798* .778* -0.448 -0.441 .926* .785* 1 .800** -.904** .814** 0.614 0.663 -.816** .816** -0.218 0.086 0.539 0.526 0.467 0.572 0.434 .729* -0.454 -0.560 -0.548 0.041 0.415 -0.525 p 0.018 0.014 0.227 0.234 0.024 0.012 0.010 0.001 0.008 0.079 0.051 0.007 0.007 0.572 0.826 0.134 0.145 0.205 0.108 0.243 0.026 0.220 0.117 0.127 0.917 0.267 0.146 Organic Matter r 0.408 0.363 0.207 0.119 .891* 0.208 .800** 1 -0.562 0.409 0.377 .662* -0.418 0.418 -0.070 0.235 0.009 0.116 -0.051 0.234 -0.024 0.507 0.088 -0.169 0.015 0.482 0.298 0.054 (wt%) p 0.275 0.302 0.567 0.744 0.042 0.564 0.010 0.115 0.241 0.283 0.037 0.230 0.230 0.849 0.513 0.981 0.750 0.888 0.515 0.947 0.134 0.810 0.640 0.967 0.159 0.403 0.882 Lithogenic r -.920** -.924** .750* .752* -.970** -.969** -.904** -0.562 1 -.911** -.791* -.702* .911** -.911** 0.169 0.123 -.822** -.794* -.717* -.679* -0.570 -.781* .746* 0.582 .818** 0.191 -0.328 .809** Fraction (wt%) p 0.001 0.000 0.020 0.019 0.006 0.000 0.001 0.115 0.001 0.011 0.035 0.001 0.001 0.665 0.753 0.007 0.011 0.030 0.044 0.109 0.013 0.021 0.100 0.007 0.623 0.388 0.008 δ13C (‰) r .786* .966** -.741* -.745* .972** .881** .814** 0.409 -.911** 1 .916** 0.330-1.000** 1.000** -0.507 0.044 .862** .722* .857** .881** .799** .883** -.725* -0.550 -.776** -0.124 0.553 -.828** p 0.012 0.000 0.014 0.013 0.006 0.001 0.008 0.241 0.001 0.000 0.351 0.000 0.000 0.135 0.904 0.001 0.018 0.002 0.001 0.006 0.001 0.018 0.100 0.008 0.733 0.097 0.003 δ15N (‰) r .672* .873** -.656* -.687* .937* .747* 0.614 0.377 -.791* .916** 1 0.511 -.918** .918** -0.548 0.073 .810** 0.618 .864** .874** .790** .895** -.644* -0.326 -.670* -0.142 0.589 -.751* p 0.047 0.001 0.040 0.028 0.019 0.013 0.079 0.283 0.011 0.000 0.131 0.000 0.000 0.101 0.840 0.004 0.057 0.001 0.001 0.007 0.000 0.045 0.358 0.034 0.695 0.073 0.012 Atomic N/C r 0.380 0.223 0.069 -0.119 0.836 0.129 0.663 .662* -.702* 0.330 0.511 1 -0.337 0.337 -0.100 0.276 0.143 0.164 0.158 0.314 0.066 0.449 -0.014 0.077 -0.036 0.166 0.347 -0.103 p 0.314 0.535 0.849 0.743 0.078 0.723 0.051 0.037 0.035 0.351 0.131 0.340 0.340 0.782 0.439 0.694 0.651 0.664 0.377 0.857 0.193 0.969 0.833 0.921 0.647 0.326 0.777
FT (%) r -.787* -.965** .737* .740* -.976** -.879** -.816** -0.418 .911** -1.000** -.918** -0.337 1-1.000** 0.507 -0.039 -.858** -.719* -.852** -.884** -.796** -.882** .724* 0.553 .774** 0.112 -0.550 .824** p 0.012 0.000 0.015 0.014 0.005 0.001 0.007 0.230 0.001 0.000 0.000 0.340 0.000 0.135 0.916 0.001 0.019 0.002 0.001 0.006 0.001 0.018 0.097 0.009 0.758 0.099 0.003
Fm (%) r .787* .965** -.737* -.740* .976** .879** .816** 0.418 -.911** 1.000** .918** 0.337-1.000** 1 -0.507 0.039 .858** .719* .852** .884** .796** .882** -.724* -0.553 -.774** -0.112 0.550 -.824** p 0.012 0.000 0.015 0.014 0.005 0.001 0.007 0.230 0.001 0.000 0.000 0.340 0.000 0.135 0.916 0.001 0.019 0.002 0.001 0.006 0.001 0.018 0.097 0.009 0.758 0.099 0.003 Al (wt%) r 0.164 -0.376 0.217 0.095 0.522 -0.159 -0.218 -0.070 0.169 -0.507 -0.548 -0.100 0.507 -0.507 1 -0.406 -0.281 0.078 -0.566 -.736* -.830** -0.599 0.116 0.136 0.150 -0.115 -.800** 0.238 p 0.673 0.284 0.547 0.793 0.367 0.661 0.572 0.849 0.665 0.135 0.101 0.782 0.135 0.135 0.245 0.432 0.830 0.088 0.015 0.003 0.067 0.749 0.708 0.679 0.752 0.006 0.508 log(Ca/Al) r -0.176 -0.099 0.412 0.316 -0.178 -0.207 0.086 0.235 0.123 0.044 0.073 0.276 -0.039 0.039 -0.406 1 -0.263 -0.422 -0.032 -0.020 0.079 0.390 0.554 0.566 0.496 -0.216 .776** 0.272 p 0.650 0.786 0.237 0.373 0.775 0.566 0.826 0.513 0.753 0.904 0.840 0.439 0.916 0.916 0.245 0.463 0.224 0.929 0.957 0.827 0.265 0.096 0.088 0.145 0.549 0.008 0.447 log(Fe/Al) r .738* .901** -.955** -.964** 0.756 .908** 0.539 0.009 -.822** .862** .810** 0.143 -.858** .858** -0.281 -0.263 1 .905** .922** .773** .747* .650* -.936** -0.551 -.947** -0.368 0.244 -.992** p 0.023 0.000 0.000 0.000 0.140 0.000 0.134 0.981 0.007 0.001 0.004 0.694 0.001 0.001 0.432 0.463 0.000 0.000 0.009 0.013 0.042 0.000 0.099 0.000 0.296 0.498 0.000 log(K/Al) r .770* .824** -.853** -.912** 0.790 .846** 0.526 0.116 -.794* .722* 0.618 0.164 -.719* .719* 0.078 -0.422 .905** 1 .681* 0.527 0.434 0.446 -.914** -0.628 -.917** -0.247 -0.069 -.906** p 0.015 0.003 0.002 0.000 0.111 0.002 0.145 0.750 0.011 0.018 0.057 0.651 0.019 0.019 0.830 0.224 0.000 0.030 0.118 0.210 0.197 0.000 0.052 0.000 0.492 0.851 0.000 log(Mg/Al) r 0.594 .833** -.862** -.847** 0.569 .818** 0.467 -0.051 -.717* .857** .864** 0.158 -.852** .852** -0.566 -0.032 .922** .681* 1 .860** .916** .750* -.791** -0.384 -.821** -0.400 0.523 -.906** p 0.092 0.003 0.001 0.002 0.317 0.004 0.205 0.888 0.030 0.002 0.001 0.664 0.002 0.002 0.088 0.929 0.000 0.030 0.001 0.000 0.012 0.006 0.274 0.004 0.252 0.121 0.000 log(Mn/Al) r 0.540 .796** -.710* -.641* .933* .684* 0.572 0.234 -.679* .881** .874** 0.314 -.884** .884** -.736* -0.020 .773** 0.527 .860** 1 .911** .772** -.669* -0.527 -.696* 0.000 0.575 -.741* p 0.134 0.006 0.021 0.046 0.020 0.029 0.108 0.515 0.044 0.001 0.001 0.377 0.001 0.001 0.015 0.957 0.009 0.118 0.001 0.000 0.009 0.034 0.118 0.025 1.000 0.082 0.014 log(Na/Al) r 0.327 .736* -.697* -0.606 0.620 0.631 0.434 -0.024 -0.570 .799** .790** 0.066 -.796** .796** -.830** 0.079 .747* 0.434 .916** .911** 1 .728* -0.618 -0.411 -.656* -0.153 .654* -.719* p 0.390 0.015 0.025 0.063 0.264 0.051 0.243 0.947 0.109 0.006 0.007 0.857 0.006 0.006 0.003 0.827 0.013 0.210 0.000 0.000 0.017 0.057 0.238 0.039 0.672 0.040 0.019 log(P/Al) r 0.577 .819** -0.456 -0.495 .914* .695* .729* 0.507 -.781* .883** .895** 0.449 -.882** .882** -0.599 0.390 .650* 0.446 .750* .772** .728* 1 -0.386 -0.131 -0.462 -0.214 .798** -0.593 p 0.104 0.004 0.186 0.146 0.030 0.026 0.026 0.134 0.013 0.001 0.000 0.193 0.001 0.001 0.067 0.265 0.042 0.197 0.012 0.009 0.017 0.271 0.719 0.179 0.553 0.006 0.071 log(Si/Al) r -.692* -.796** .959** .933** -0.742 -.836** -0.454 0.088 .746* -.725* -.644* -0.014 .724* -.724* 0.116 0.554 -.936** -.914** -.791** -.669* -0.618 -0.386 1 .733* .986** 0.199 0.060 .941** p 0.039 0.006 0.000 0.000 0.151 0.003 0.220 0.810 0.021 0.018 0.045 0.969 0.018 0.018 0.749 0.096 0.000 0.000 0.006 0.034 0.057 0.271 0.016 0.000 0.581 0.870 0.000 log(Ti/Al) r -0.467 -0.590 0.603 0.502 -0.729 -0.541 -0.560 -0.169 0.582 -0.550 -0.326 0.077 0.553 -0.553 0.136 0.566 -0.551 -0.628 -0.384 -0.527 -0.411 -0.131 .733* 1 .730* -0.406 0.187 0.558 p 0.205 0.072 0.065 0.139 0.162 0.107 0.117 0.640 0.100 0.100 0.358 0.833 0.097 0.097 0.708 0.088 0.099 0.052 0.274 0.118 0.238 0.719 0.016 0.017 0.244 0.605 0.094 log(litho-Si/Al) r -.735* -.838** .948** .928** -0.798 -.894** -0.548 0.015 .818** -.776** -.670* -0.036 .774** -.774** 0.150 0.496 -.947** -.917** -.821** -.696* -.656* -0.462 .986** .730* 1 0.210 -0.022 .951** p 0.024 0.002 0.000 0.000 0.106 0.000 0.127 0.967 0.007 0.008 0.034 0.921 0.009 0.009 0.679 0.145 0.000 0.000 0.004 0.025 0.039 0.179 0.000 0.017 0.561 0.951 0.000 log(Ba/Al) r -0.260 -0.129 0.400 0.488 0.169 -0.365 0.041 0.482 0.191 -0.124 -0.142 0.166 0.112 -0.112 -0.115 -0.216 -0.368 -0.247 -0.400 0.000 -0.153 -0.214 0.199 -0.406 0.210 1 -0.202 0.416 p 0.500 0.723 0.252 0.152 0.786 0.300 0.917 0.159 0.623 0.733 0.695 0.647 0.758 0.758 0.752 0.549 0.296 0.492 0.252 1.000 0.672 0.553 0.581 0.244 0.561 0.575 0.231 log(Sr/Al) r 0.109 0.396 -0.082 -0.102 0.172 0.266 0.415 0.298 -0.328 0.553 0.589 0.347 -0.550 0.550 -.800** .776** 0.244 -0.069 0.523 0.575 .654* .798** 0.060 0.187 -0.022 -0.202 1 -0.209 p 0.779 0.257 0.822 0.778 0.782 0.457 0.267 0.403 0.388 0.097 0.073 0.326 0.099 0.099 0.006 0.008 0.498 0.851 0.121 0.082 0.040 0.006 0.870 0.605 0.951 0.575 0.561 log(Zr/Al) r -.746* -.861** .970** .983** -0.714 -.906** -0.525 0.054 .809** -.828** -.751* -0.103 .824** -.824** 0.238 0.272 -.992** -.906** -.906** -.741* -.719* -0.593 .941** 0.558 .951** 0.416 -0.209 1 p 0.021 0.001 0.000 0.000 0.176 0.000 0.146 0.882 0.008 0.003 0.012 0.777 0.003 0.003 0.508 0.447 0.000 0.000 0.000 0.014 0.019 0.071 0.000 0.094 0.000 0.231 0.561
83
Table 8.4.4. Correlation matrix of the channels west of the SPI. Pearson correlation coefficient (r), two-tailed test of significance (p), correlation is significant at the 0.05 level (*), correlation is significant at the 0.01 level (**).
Mass-specific Mean Surface Magnetic IRD free Organic Lithogenic Salinity Distance Susceptibility Grain IRD150 Bio-Opal CaCO3 Matter Fraction δ13C δ15N Atomic log(litho- (psu) (km) (10-6 SI) Size (µm) (wt%) (wt%) (wt%) (wt%) (wt%) (‰) (‰) N/C FT (%) Fm (%) Al (wt%) log(Ca/Al) log(Fe/Al) log(K/Al) log(Mg/Al) log(Mn/Al) log(Na/Al) log(P/Al) log(Si/Al) log(Ti/Al) Si/Al) log(Ba/Al) log(Sr/Al) log(Zr/Al) Surface Salinity r 1 .640* 0.323 0.424 0.308 -0.495 .606* 0.485 -0.479 0.491 -0.028 .849** -0.484 0.484 -.747** .623* -0.182 -0.438 0.176 -0.391 .603* 0.475 .636* -0.099 .576* -.585* .610* 0.517 (psu) p 0.019 0.282 0.149 0.305 0.086 0.028 0.093 0.098 0.180 0.943 0.004 0.187 0.187 0.003 0.023 0.551 0.134 0.565 0.187 0.029 0.101 0.019 0.749 0.040 0.036 0.027 0.071 Distance (km) r .640* 1 0.340 .813** .613* -0.458 .744** 0.544 -.603* .936** .774* 0.115 -.937** .937** -.904** .862** -0.485 -.824** 0.117 -0.339 .743** 0.520 .845** -.646* .749** -.794** .840** 0.511 p 0.019 0.255 0.001 0.026 0.115 0.004 0.055 0.029 0.000 0.014 0.767 0.000 0.000 0.000 0.000 0.093 0.001 0.705 0.257 0.004 0.069 0.000 0.017 0.003 0.001 0.000 0.074 Mass-specific r 0.323 0.340 1 0.430 .644* -0.344 0.266 0.052 -0.142 -0.334 -0.304 0.075 0.336 -0.336 -0.389 0.432 0.361 -0.371 0.058 -0.228 0.135 0.085 0.408 -0.103 0.403 -.795** 0.349 0.195 Magnetic p 0.282 0.255 0.142 0.018 0.250 0.380 0.865 0.643 0.380 0.426 0.849 0.376 0.376 0.188 0.141 0.225 0.212 0.851 0.454 0.661 0.783 0.167 0.738 0.172 0.001 0.242 0.522 Susceptibility (10-6 SI) Mean IRD free r 0.424 .813** 0.430 1 .882** -.737** 0.352 0.017 -0.120 -0.042 -0.283 0.453 0.051 -0.051 -.630* .571* -.670* -.966** -0.391 -.591* 0.264 0.023 .943** -.878** .940** -.835** 0.528 .796** Grain Size (µm) p 0.149 0.001 0.142 0.000 0.004 0.238 0.957 0.696 0.915 0.460 0.221 0.897 0.897 0.021 0.042 0.012 0.000 0.186 0.033 0.383 0.942 0.000 0.000 0.000 0.000 0.063 0.001 IRD150 (wt%) r 0.308 .613* .644* .882** 1 -.735** 0.182 -0.157 0.051 -0.142 -0.059 -0.073 0.146 -0.146 -0.435 0.414 -0.393 -.791** -0.436 -.557* 0.054 -0.148 .784** -.729** .817** -.837** 0.346 .690** p 0.305 0.026 0.018 0.000 0.004 0.552 0.609 0.869 0.716 0.881 0.852 0.708 0.708 0.137 0.159 0.184 0.001 0.136 0.048 0.860 0.629 0.002 0.005 0.001 0.000 0.247 0.009 Bio-Opal (wt%) r -0.495 -0.458 -0.344 -.737** -.735** 1 -0.115 0.256 -0.178 0.246 0.605 -.750* -0.253 0.253 0.355 -0.218 .562* .660* .562* .570* 0.093 0.194 -.810** .559* -.886** .614* -0.217 -.952** p 0.086 0.115 0.250 0.004 0.004 0.709 0.398 0.560 0.523 0.084 0.020 0.512 0.512 0.235 0.474 0.046 0.014 0.046 0.042 0.763 0.525 0.001 0.047 0.000 0.026 0.475 0.000 CaCO3 (wt%) r .606* .744** 0.266 0.352 0.182 -0.115 1 .909** -.956** 0.647 0.514 -0.142 -0.645 0.645 -.928** .949** -0.033 -0.464 .670* -0.202 .865** .932** 0.509 -0.086 0.340 -.589* .975** 0.110 p 0.028 0.004 0.380 0.238 0.552 0.709 0.000 0.000 0.060 0.157 0.716 0.061 0.061 0.000 0.000 0.914 0.110 0.012 0.507 0.000 0.000 0.076 0.779 0.256 0.034 0.000 0.722 Organic Matter r 0.485 0.544 0.052 0.017 -0.157 0.256 .909** 1 -.985** .724* .782* -0.411 -.726* .726* -.761** .807** 0.157 -0.146 .855** 0.021 .904** .967** 0.171 0.179 -0.017 -0.277 .832** -0.235 (wt%) p 0.093 0.055 0.865 0.957 0.609 0.398 0.000 0.000 0.027 0.013 0.272 0.027 0.027 0.003 0.001 0.608 0.633 0.000 0.947 0.000 0.000 0.576 0.558 0.955 0.360 0.000 0.440 Lithogenic r -0.479 -.603* -0.142 -0.120 0.051 -0.178 -.956** -.985** 1 -0.518 -.728* 0.619 0.522 -0.522 .816** -.870** -0.127 0.252 -.830** 0.033 -.896** -.981** -0.262 -0.093 -0.073 0.388 -.897** 0.167 Fraction (wt%) p 0.098 0.029 0.643 0.696 0.869 0.560 0.000 0.000 0.153 0.026 0.076 0.149 0.149 0.001 0.000 0.680 0.407 0.000 0.916 0.000 0.000 0.387 0.762 0.812 0.191 0.000 0.586 δ13C (‰) r 0.491 .936** -0.334 -0.042 -0.142 0.246 0.647 .724* -0.518 1 .836** 0.156 -1.000** 1.000** -0.629 0.365 0.125 0.597 0.478 0.332 .880** .715* -0.137 -0.098 -0.186 0.005 0.112 -0.209 p 0.180 0.000 0.380 0.915 0.716 0.523 0.060 0.027 0.153 0.005 0.689 0.000 0.000 0.070 0.334 0.749 0.090 0.193 0.382 0.002 0.031 0.726 0.803 0.632 0.989 0.774 0.590 δ15N (‰) r -0.028 .774* -0.304 -0.283 -0.059 0.605 0.514 .782* -.728* .836** 1 -0.362 -.840** .840** -0.257 0.485 0.630 0.651 .835** 0.409 .887** .804** -0.557 -0.309 -0.594 0.431 0.284 -0.658 p 0.943 0.014 0.426 0.460 0.881 0.084 0.157 0.013 0.026 0.005 0.339 0.005 0.005 0.504 0.186 0.069 0.057 0.005 0.274 0.001 0.009 0.119 0.419 0.091 0.247 0.458 0.054 Atomic N/C r .849** 0.115 0.075 0.453 -0.073 -.750* -0.142 -0.411 0.619 0.156 -0.362 1 -0.150 0.150 -0.444 -0.472 -.848** -0.114 -.699* -0.099 -0.090 -0.234 .845** 0.555 .846** -0.650 -0.467 .835** p 0.004 0.767 0.849 0.221 0.852 0.020 0.716 0.272 0.076 0.689 0.339 0.699 0.699 0.232 0.199 0.004 0.770 0.036 0.801 0.817 0.545 0.004 0.121 0.004 0.058 0.205 0.005
FT (%) r -0.484 -.937** 0.336 0.051 0.146 -0.253 -0.645 -.726* 0.522 -1.000** -.840** -0.150 1 -1.000** 0.623 -0.365 -0.132 -0.600 -0.483 -0.339 -.881** -.720* 0.145 0.105 0.193 -0.014 -0.113 0.217 p 0.187 0.000 0.376 0.897 0.708 0.512 0.061 0.027 0.149 0.000 0.005 0.699 0.000 0.073 0.335 0.735 0.087 0.188 0.373 0.002 0.029 0.711 0.788 0.619 0.971 0.773 0.576
Fm (%) r 0.484 .937** -0.336 -0.051 -0.146 0.253 0.645 .726* -0.522 1.000** .840** 0.150 -1.000** 1 -0.623 0.365 0.132 0.600 0.483 0.339 .881** .720* -0.145 -0.105 -0.193 0.014 0.113 -0.217 p 0.187 0.000 0.376 0.897 0.708 0.512 0.061 0.027 0.149 0.000 0.005 0.699 0.000 0.073 0.335 0.735 0.087 0.188 0.373 0.002 0.029 0.711 0.788 0.619 0.971 0.773 0.576 Al (wt%) r -.747** -.904** -0.389 -.630* -0.435 0.355 -.928** -.761** .816** -0.629 -0.257 -0.444 0.623 -0.623 1 -.976** 0.245 .704** -0.410 0.386 -.842** -.759** -.757** 0.336 -.622* .780** -.971** -0.386 p 0.003 0.000 0.188 0.021 0.137 0.235 0.000 0.003 0.001 0.070 0.504 0.232 0.073 0.073 0.000 0.419 0.007 0.164 0.192 0.000 0.003 0.003 0.261 0.023 0.002 0.000 0.193 log(Ca/Al) r .623* .862** 0.432 .571* 0.414 -0.218 .949** .807** -.870** 0.365 0.485 -0.472 -0.365 0.365 -.976** 1 -0.129 -.663* 0.505 -0.335 .848** .802** .669* -0.289 0.517 -.769** .990** 0.237 p 0.023 0.000 0.141 0.042 0.159 0.474 0.000 0.001 0.000 0.334 0.186 0.199 0.335 0.335 0.000 0.675 0.013 0.079 0.263 0.000 0.001 0.012 0.339 0.071 0.002 0.000 0.436 log(Fe/Al) r -0.182 -0.485 0.361 -.670* -0.393 .562* -0.033 0.157 -0.127 0.125 0.630 -.848** -0.132 0.132 0.245 -0.129 1 .671* .580* 0.441 -0.030 0.174 -.637* .816** -.668* 0.189 -0.151 -.740** p 0.551 0.093 0.225 0.012 0.184 0.046 0.914 0.608 0.680 0.749 0.069 0.004 0.735 0.735 0.419 0.675 0.012 0.038 0.132 0.922 0.570 0.019 0.001 0.013 0.536 0.623 0.004 log(K/Al) r -0.438 -.824** -0.371 -.966** -.791** .660* -0.464 -0.146 0.252 0.597 0.651 -0.114 -0.600 0.600 .704** -.663* .671* 1 0.287 .604* -0.336 -0.150 -.946** .826** -.915** .818** -.635* -.755** p 0.134 0.001 0.212 0.000 0.001 0.014 0.110 0.633 0.407 0.090 0.057 0.770 0.087 0.087 0.007 0.013 0.012 0.343 0.029 0.261 0.625 0.000 0.001 0.000 0.001 0.020 0.003 log(Mg/Al) r 0.176 0.117 0.058 -0.391 -0.436 .562* .670* .855** -.830** 0.478 .835** -.699* -0.483 0.483 -0.410 0.505 .580* 0.287 1 0.293 .686** .866** -0.264 .566* -0.432 0.008 0.535 -.620* p 0.565 0.705 0.851 0.186 0.136 0.046 0.012 0.000 0.000 0.193 0.005 0.036 0.188 0.188 0.164 0.079 0.038 0.343 0.331 0.010 0.000 0.383 0.044 0.141 0.979 0.060 0.024 log(Mn/Al) r -0.391 -0.339 -0.228 -.591* -.557* .570* -0.202 0.021 0.033 0.332 0.409 -0.099 -0.339 0.339 0.386 -0.335 0.441 .604* 0.293 1 -0.046 0.051 -.619* 0.412 -.620* 0.501 -0.300 -.631* p 0.187 0.257 0.454 0.033 0.048 0.042 0.507 0.947 0.916 0.382 0.274 0.801 0.373 0.373 0.192 0.263 0.132 0.029 0.331 0.880 0.870 0.024 0.162 0.024 0.081 0.320 0.021 log(Na/Al) r .603* .743** 0.135 0.264 0.054 0.093 .865** .904** -.896** .880** .887** -0.090 -.881** .881** -.842** .848** -0.030 -0.336 .686** -0.046 1 .863** 0.384 -0.052 0.219 -0.428 .849** -0.055 p 0.029 0.004 0.661 0.383 0.860 0.763 0.000 0.000 0.000 0.002 0.001 0.817 0.002 0.002 0.000 0.000 0.922 0.261 0.010 0.880 0.000 0.195 0.867 0.473 0.144 0.000 0.858 log(P/Al) r 0.475 0.520 0.085 0.023 -0.148 0.194 .932** .967** -.981** .715* .804** -0.234 -.720* .720* -.759** .802** 0.174 -0.150 .866** 0.051 .863** 1 0.190 0.187 0.007 -0.305 .846** -0.200 p 0.101 0.069 0.783 0.942 0.629 0.525 0.000 0.000 0.000 0.031 0.009 0.545 0.029 0.029 0.003 0.001 0.570 0.625 0.000 0.870 0.000 0.533 0.540 0.982 0.310 0.000 0.513 log(Si/Al) r .636* .845** 0.408 .943** .784** -.810** 0.509 0.171 -0.262 -0.137 -0.557 .845** 0.145 -0.145 -.757** .669* -.637* -.946** -0.264 -.619* 0.384 0.190 1 -.722** .980** -.845** .651* .855** p 0.019 0.000 0.167 0.000 0.002 0.001 0.076 0.576 0.387 0.726 0.119 0.004 0.711 0.711 0.003 0.012 0.019 0.000 0.383 0.024 0.195 0.533 0.005 0.000 0.000 0.016 0.000 log(Ti/Al) r -0.099 -.646* -0.103 -.878** -.729** .559* -0.086 0.179 -0.093 -0.098 -0.309 0.555 0.105 -0.105 0.336 -0.289 .816** .826** .566* 0.412 -0.052 0.187 -.722** 1 -.746** 0.533 -0.256 -.675* p 0.749 0.017 0.738 0.000 0.005 0.047 0.779 0.558 0.762 0.803 0.419 0.121 0.788 0.788 0.261 0.339 0.001 0.001 0.044 0.162 0.867 0.540 0.005 0.003 0.060 0.398 0.011 log(litho-Si/Al) r .576* .749** 0.403 .940** .817** -.886** 0.340 -0.017 -0.073 -0.186 -0.594 .846** 0.193 -0.193 -.622* 0.517 -.668* -.915** -0.432 -.620* 0.219 0.007 .980** -.746** 1 -.798** 0.495 .921** p 0.040 0.003 0.172 0.000 0.001 0.000 0.256 0.955 0.812 0.632 0.091 0.004 0.619 0.619 0.023 0.071 0.013 0.000 0.141 0.024 0.473 0.982 0.000 0.003 0.001 0.085 0.000 log(Ba/Al) r -.585* -.794** -.795** -.835** -.837** .614* -.589* -0.277 0.388 0.005 0.431 -0.650 -0.014 0.014 .780** -.769** 0.189 .818** 0.008 0.501 -0.428 -0.305 -.845** 0.533 -.798** 1 -.713** -.574* p 0.036 0.001 0.001 0.000 0.000 0.026 0.034 0.360 0.191 0.989 0.247 0.058 0.971 0.971 0.002 0.002 0.536 0.001 0.979 0.081 0.144 0.310 0.000 0.060 0.001 0.006 0.040 log(Sr/Al) r .610* .840** 0.349 0.528 0.346 -0.217 .975** .832** -.897** 0.112 0.284 -0.467 -0.113 0.113 -.971** .990** -0.151 -.635* 0.535 -0.300 .849** .846** .651* -0.256 0.495 -.713** 1 0.234 p 0.027 0.000 0.242 0.063 0.247 0.475 0.000 0.000 0.000 0.774 0.458 0.205 0.773 0.773 0.000 0.000 0.623 0.020 0.060 0.320 0.000 0.000 0.016 0.398 0.085 0.006 0.442 log(Zr/Al) r 0.517 0.511 0.195 .796** .690** -.952** 0.110 -0.235 0.167 -0.209 -0.658 .835** 0.217 -0.217 -0.386 0.237 -.740** -.755** -.620* -.631* -0.055 -0.200 .855** -.675* .921** -.574* 0.234 1 p 0.071 0.074 0.522 0.001 0.009 0.000 0.722 0.440 0.586 0.590 0.054 0.005 0.576 0.576 0.193 0.436 0.004 0.003 0.024 0.021 0.858 0.513 0.000 0.011 0.000 0.040 0.442
84
8.5 Wavelenghts ICP-AES
Table 8.5. List of elements measured by ICP-AES and their corresponding wavelengths. The selected lines are in bold, these were averaged and processed for the final results (Appendix 8.1, table 8.1.5). rsd (%) = relative standard deviation (%). The measured and reference values of PACS-2 are expressed in wt%, except for Mn, Ba, Sr and Zr which are expressed in ppm. The measured and reference values of STSD3 are expressed in wt%, except for Ba, Sr and Zr which are expressed in ppm. The results from Bertrand et al. (2012b) are expressed in ppm. PACS-2 STSD3 PACS-2 results from Element Line (nm) rsd (%) Measured value Reference Value Measured value Reference Value Bertrand et al. (2012b) Al 236.705 0.51 7.07 6.62 +- 0.32 11.48 10.9 --- Al 257.509 0.39 7.20 6.62 +- 0.32 11.57 10.9 --- Al 308.215 0.49 7.06 6.62 +- 0.32 11.34 10.9 --- Al 309.271 0.50 7.17 6.62 +- 0.32 11.55 10.9 --- Al 394.401 0.44 7.03 6.62 +- 0.32 11.35 10.9 --- Ca 183.738 0.42 2.18 1.96 +- 0.18 3.56 3.3 --- Ca 183.944 0.39 2.21 1.96 +- 0.18 3.62 3.3 --- Ca 219.779 0.41 2.20 1.96 +- 0.18 3.65 3.3 --- Ca 315.887 0.61 2.21 1.96 +- 0.18 3.61 3.3 --- Ca 317.933 0.37 2.22 1.96 +- 0.18 3.64 3.3 --- Ca 318.127 0.37 2.22 1.96 +- 0.18 3.63 3.3 --- Fe 234.350 0.50 4.80 4.09 +- 0.06 6.68 6.2 --- Fe 238.204 0.41 4.75 4.09 +- 0.06 6.61 6.2 --- Fe 259.940 0.46 4.86 4.09 +- 0.06 6.73 6.2 --- Fe 261.187 0.46 4.72 4.09 +- 0.06 6.57 6.2 --- Fe 273.358 0.43 4.47 4.09 +- 0.06 6.25 6.2 --- K 766.491 0.53 1.10 1.24 +- 0.05 1.58 1.8 --- K 769.897 1.26 0.24 1.24 +- 0.05 --- 1.8 --- Mg 277.983 0.42 1.54 1.24 +- 0.05 2.29 2.2 --- Mg 279.078 0.46 1.60 1.24 +- 0.05 --- 2.2 --- Mg 279.800 0.46 1.61 1.24 +- 0.05 --- 2.2 --- Mg 285.213 0.54 1.51 1.24 +- 0.05 --- 2.2 --- Mn 257.610 0.33 502 440 +- 19 0.37 0.3 --- Mn 293.305 0.48 505 440 +- 19 0.37 0.3 --- Mn 293.931 0.51 498 440 +- 19 0.37 0.3 --- Na 588.995 0.54 3.53 3.45 +- 0.17 1.68 1.5 --- Na 589.592 0.76 3.57 3.45 +- 0.17 1.51 1.5 --- P 177.434 0.86 0.10 0.096 +- 0.004 0.38 0.4 --- P 178.222 1.05 0.10 0.096 +- 0.004 0.38 0.4 --- P 178.703 3.54 0.10 0.096 +- 0.004 --- 0.4 --- P 185.878 0.88 0.10 0.096 +- 0.004 0.40 0.4 --- Si 185.005 0.51 28.25 28 51.28 48.6 --- Si 212.412 0.51 28.33 28 49.83 48.6 --- Si 250.690 0.56 28.49 28 51.23 48.6 --- Si 251.611 0.51 27.46 28 49.54 48.6 --- Si 252.851 0.52 29.03 28 52.15 48.6 --- Si 288.158 0.45 28.37 28 50.71 48.6 --- Ti 308.804 0.42 0.50 0.443 +- 0.032 0.71 0.7 --- Ti 334.188 0.39 0.52 0.443 +- 0.032 0.74 0.7 --- Ti 334.941 0.45 0.52 0.443 +- 0.032 0.74 0.7 --- Ti 336.122 0.43 0.53 0.443 +- 0.032 0.76 0.7 --- Ti 337.280 0.50 0.51 0.443 +- 0.032 0.73 0.7 --- Ba 233.527 0.52 1019 --- 1503 1490 969 Ba 455.403 0.46 1035 --- 1483 1490 --- Ba 493.408 0.53 1002 --- 1495 1490 --- Ba 585.367 0.51 1029 --- 1532 1490 --- Sr 407.771 0.47 320 276 ± 30 280 230 288 Sr 421.552 0.44 319 276 ± 30 278 230 --- Zr 257.143 0.69 131 --- 254 196 --- Zr 267.865 1.09 146 ------196 --- Zr 327.307 0.64 140 --- 254 196 --- Zr 339.198 0.58 142 --- 257 196 --- Zr 343.823 0.57 138 ------196 141 Zr 349.619 0.75 141 ------196 ---
85