The Biogeochemical Cycle of Silicon in the Ocean 9781848218154

385 65 10MB

English Pages 149 Year 2016

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

The Biogeochemical Cycle of Silicon in the Ocean
 9781848218154

Table of contents :
Cover......Page 1
Title Page......Page 5
Copyright......Page 6
Contents......Page 7
Preface......Page 11
1.1. The element “silicon”......Page 15
1.2. Orthosilicic acid......Page 16
1.3. Particulate silicas......Page 17
1.3.1. Lithogenic silica......Page 18
1.3.2. Biogenic silica......Page 19
2.1. Techniques for the chemical analysis of silicon......Page 21
2.1.1. The sequential digestion method......Page 22
2.1.3. The correction by aluminum method
......Page 23
2.2.1. Labeling with radioactive isotopes......Page 25
2.2.2. Labeling with stable isotopes......Page 28
2.3. Silica deposit labeling and cellular imaging......Page 31
2.4. Isotopic fractionation of silicon and utilization of δ30 Si as a tracer in oceanography......Page 32
2.4.1. Demonstration of the isotopic fractionation by the diatoms......Page 36
2.4.2. Utilization of δ30 Si as a tracer in oceanography......Page 37
2.4.3. The interest of analyses of the isotopic ratio of silicon......Page 38
3.1. Radiolarians......Page 41
3.3. Diatoms......Page 43
3.4. Silicification within the scope of nanoplankton and picoplankton......Page 45
3.5. Siliceous sponges......Page 47
3.6. The functions of biogenic silica......Page 48
3.7. The evolution of the siliceous organisms and the oceanic cycle of the silicon......Page 51
3.8. Sedimentary opal deposits......Page 53
4.1.1. General formulations and kinetics information......Page 57
4.2.1. The model of Riedel and Nelson [RIE 85]......Page 60
4.2.2. The model of Del Amo and Brzezinski......Page 62
4.2.3. The membrane transporters......Page 64
4.3. Cellular mechanisms of orthosilicic acid uptake......Page 65
4.4.1. The Hecky et al. conceptual model......Page 67
4.4.2. Frustulins and silaffins......Page 70
4.4.3. Frustule synthesis, a complex physiological process......Page 72
4.5.1. Stoichiometry in diatoms and limitation by iron......Page 73
4.5.2. The influence of trace metals on the uptake of orthosilicic acid......Page 75
5.1. Reactivity of the particulate silica and dissolution constants......Page 77
5.2. Processes of control of the dissolution in aqueous phase......Page 80
5.2.1. Variation of the solubility of opal with depth......Page 81
5.2.3. Role of temperature......Page 82
5.2.4. Relationship with bacterial degradation process......Page 84
5.2.5. Influence of aluminum concentration......Page 85
5.3. The solubility of opal in natural conditions......Page 87
6.1. The preservation of calcite in ocean sediments......Page 91
6.1.1. Control of alkalinity by organic production......Page 92
6.1.2. The CaCO3/Corg ratio (rain ratio)......Page 93
6.1.3. The distribution of orthosilicic acid in the Global Ocean......Page 94
6.2.1. Subantarctic Mode Water (SAMW)......Page 97
6.2.2. Si* tracer......Page 99
6.2.4. The conceptual model of Sarmiento et al.......Page 101
6.3.1. The last glacial–interglacial transition......Page 103
6.3.2. The sedimentary record......Page 108
7.1. Estimates of production and export of biogenic silica......Page 113
7.1.1. Estimation of the upper limit......Page 114
7.1.2. Estimation of the lower limit......Page 116
7.1.3. General overview of production and export......Page 118
7.2. The biogeochemical cycle of silicon in the Global Ocean......Page 120
Bibliography......Page 123
Index......Page 143
Other titles from ISTE in Earth Systems - Environmental Engineering......Page 145
EULA......Page 0

Citation preview

FOCUS EARTH SYSTEM – ENVIRONMENTAL SCIENCES SERIES

The Biogeochemical Cycle of Silicon in the Ocean

Bernard Quéguiner

The Biogeochemical Cycle of Silicon in the Ocean

FOCUS SERIES Series Editor Paul Tréguer

The Biogeochemical Cycle of Silicon in the Ocean

Bernard Quéguiner

First published 2016 in Great Britain and the United States by ISTE Ltd and John Wiley & Sons, Inc.

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms and licenses issued by the CLA. Enquiries concerning reproduction outside these terms should be sent to the publishers at the undermentioned address: ISTE Ltd 27-37 St George’s Road London SW19 4EU UK

John Wiley & Sons, Inc. 111 River Street Hoboken, NJ 07030 USA

www.iste.co.uk

www.wiley.com

© ISTE Ltd 2016 The rights of Bernard Quéguiner to be identified as the author of this work have been asserted by him in accordance with the Copyright, Designs and Patents Act 1988. Library of Congress Control Number: 2016938935 British Library Cataloguing-in-Publication Data A CIP record for this book is available from the British Library ISSN 2051-2481 (Print) ISSN 2051-249X (Online) ISBN 978-1-84821-815-4

Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

Chapter 1. The Chemical Forms of Silicon in the Marine Domain . . . . . . . . . . . . . . . . . . . . . . .

1

1.1. The element “silicon” 1.2. Orthosilicic acid . . . . 1.3. Particulate silicas . . . 1.3.1. Lithogenic silica . 1.3.2. Biogenic silica . .

. . . . .

1 2 3 4 5

Chapter 2. Techniques for Studying Stocks and Fluxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7

. . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

2.1. Techniques for the chemical analysis of silicon . 2.1.1. The sequential digestion method . . . . . . . 2.1.2. The extraction kinetics method . . . . . . . . 2.1.3. The correction by aluminum method . . . . . 2.2. Techniques for the analysis of silicon fluxes . . . 2.2.1. Labeling with radioactive isotopes . . . . . . 2.2.2. Labeling with stable isotopes . . . . . . . . . 2.3. Silica deposit labeling and cellular imaging . . . 2.4. Isotopic fractionation of silicon and utilization of δ 30 Si as a tracer in oceanography . . . . . . . . . . 2.4.1. Demonstration of the isotopic fractionation by the diatoms . . . . . . . . . . . . . .

. . . . .

. . . . . . . .

. . . . .

. . . . . . . .

. . . . .

. . . . . . . .

. . . . .

. . . . . . . .

. . . . .

. . . . . . . .

. . . . .

. . . . . . . .

. . . . .

. . . . . . . .

. . . . .

. . . . . . . .

. . . . . . . .

7 8 9 9 11 11 14 17

. . . . . . . . .

18

. . . . . . . . .

22

vi

The Biogeochemical Cycle of Silicon in the Ocean

2.4.2. Utilization of δ30Si as a tracer in oceanography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.3. The interest of analyses of the isotopic ratio of silicon . . . . . . . . . . . . . . . . . . . . . . . . . Chapter 3. The Marine Producers of Biogenic Silica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Radiolarians. . . . . . . . . . . . . . . . . . 3.2. Silicoflagellates . . . . . . . . . . . . . . . 3.3. Diatoms . . . . . . . . . . . . . . . . . . . . 3.4. Silicification within the scope of nanoplankton and picoplankton . . . . . . . . . 3.5. Siliceous sponges . . . . . . . . . . . . . . 3.6. The functions of biogenic silica . . . . . . 3.7. The evolution of the siliceous organisms and the oceanic cycle of the silicon . . . . . . 3.8. Sedimentary opal deposits . . . . . . . . .

23 24 27

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

27 29 29

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

31 33 34

. . . . . . . . . . . . . . . . . . . . . . . . . .

37 39

Chapter 4. Cellular Mechanisms of Silica Deposition by Diatoms. . . . . . . . . . . . . . . . . . . . . . . . . . .

43

4.1. Influence of orthosilicic acid availability on uptake and diatom growth . . . . . . . . . . . . . . 4.1.1. General formulations and kinetics information . . . . . . . . . . . . . . . . . . 4.2. The chemical form of dissolved Si available for diatoms . . . . . . . . . . . . . . . . . . . 4.2.1. The model of Riedel and Nelson . . . . . . . 4.2.2. The model of Del Amo and Brzezinski . . . 4.2.3. The membrane transporters . . . . . . . . . . 4.3. Cellular mechanisms of orthosilicic acid uptake 4.4. Intervention of specific proteins in the deposition mechanism . . . . . . . . . . . . . . . . . . 4.4.1. The Hecky et al. conceptual model . . . . . 4.4.2. Frustulins and silaffins . . . . . . . . . . . . . 4.4.3. Frustule synthesis, a complex physiological process . . . . . . . . . . . . . . . . . 4.5. The stoichiometric ratios Si/C/N of diatoms . .

. . . . . . . . .

43

. . . . . . . . .

43

. . . . .

. . . . .

46 46 48 50 51

. . . . . . . . . . . . . . . . . . . . . . . . . . .

53 53 56

. . . . . . . . . . . . . . . . . .

58 59

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

Contents

vii

4.5.1. Stoichiometry in diatoms and limitation by iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.2. The influence of trace metals on the uptake of orthosilicic acid . . . . . . . . . . . . . . . . . . . . . . . .

61

Chapter 5. Dissolution of Biogenic Silica and Orthosilicic Acid Regeneration . . . . . . . . . . . . . .

63

5.1. Reactivity of the particulate silica and dissolution constants . . . . . . . . . . . . . . . . . . 5.2. Processes of control of the dissolution in aqueous phase . . . . . . . . . . . . . . . . . . . . . . . 5.2.1. Variation of the solubility of opal with depth . 5.2.2. Influence of pH . . . . . . . . . . . . . . . . . . . 5.2.3. Role of temperature . . . . . . . . . . . . . . . . 5.2.4. Relationship with bacterial degradation process . . . . . . . . . . . . . . . . . . . . 5.2.5. Influence of aluminum concentration . . . . . . 5.3. The solubility of opal in natural conditions. . . . .

59

. . . . . . . .

63

. . . .

. . . .

66 67 68 68

. . . . . . . . . . . . . . . . . . . . . . . .

70 71 73

Chapter 6. The Control of Biogeochemistry by Silicon at Global Scale . . . . . . . . . . . . . . . . . . . . . . . . .

77

6.1. The preservation of calcite in ocean sediments . . 6.1.1. Control of alkalinity by organic production . . 6.1.2. The CaCO3/Corg ratio (rain ratio) . . . . . . . 6.1.3. The distribution of orthosilicic acid in the Global Ocean . . . . . . . . . . . . . . . . . . . . 6.2. The central role of the Southern Ocean . . . . . . . 6.2.1. Subantarctic Mode Water (SAMW) . . . . . 6.2.2. Si* tracer . . . . . . . . . . . . . . . . . . . . . . 6.2.3. The influence of SAMW in the Global Ocean 6.2.4. The conceptual model of Sarmiento et al. . . 6.3. The silicic acid leakage hypothesis (SALH) . . . 6.3.1. The last glacial–interglacial transition . . . . . 6.3.2. The sedimentary record . . . . . . . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . . . . . . . . . . . . . . . . . . . . . .

77 78 79

. . . . . . . . .

. . . . . . . . .

80 83 83 85 87 87 89 89 94

Chapter 7. The Global Budget of Silicon in the Oceans . . . . . . . . . . . . . . . . . . . . . . . . . . . .

99

7.1. Estimates of production and export of biogenic silica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.1. Estimation of the upper limit . . . . . . . . . . . . . . . . . . .

99 100

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

viii

The Biogeochemical Cycle of Silicon in the Ocean

7.1.2. Estimation of the lower limit . . . . . . . . . . . . . . . . . . 7.1.3. General overview of production and export . . . . . . . . . 7.2. The biogeochemical cycle of silicon in the Global Ocean . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

102 104 106

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

109

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

129

Preface

Biology, chemistry, physics, mathematics, geosciences, and social sciences provide us with the tools necessary to understand the past, present, and future of the world. These tools allow us to build concepts that are not necessarily accurate but seem sufficiently developed for us to consider them as paradigms. The ancient cosmology of our Western society has long been satisfied with seeing planet Earth as a flat object, a paradigm sufficient for understanding the environment perceived by humanity at that time. We must always keep in mind the approach of Descartes, and progress in our analysis and knowledge of objects in the environment, whilst maintaining the humility essential for knowing that we can sometimes be wrong and that there are limits to our understanding. Geochemistry is the science or study of the elemental chemical composition of Earth, the chemical speciation of the elements in the dynamic aspect of their transfers (= fluxes) between different compartments (= stocks). Whether or not life has little influence on the deep geochemistry of our planet1, it is a feature of its external envelope, and may be unique on the Universe scale. Biological mechanisms have emerged progressively, acting as drivers of the basic dynamics of matter regarding the chemical composition of fluid envelopes and the upper mantle rocks as well as element fluxes between these compartments under a variety of chemical species. How do we explain the elemental 1 Although the discovery of “active” microorganisms within the deep sedimentary layers is likely to change this concept in the near future.

x

The Biogeochemical Cycle of Silicon in the Ocean

composition of our atmosphere and its evolution on a geological time scale without referring to the emergence and development of life on Earth? This question also applies to the hydrosphere and, particularly, the oceans, whose chemical composition changes on that same scale, which are closely associated with the chemical structure of the primordial Earth and the evolution of living organisms. How do we explain the formation and composition of not only sedimentary rocks, but also of metamorphic rocks and even some crystalline rocks of variable nature over geological eras, evading the role of life on land and in the oceans? This short list of questions is not exhaustive and is provided here as an example; it reflects the direction of thinking from the first naturalists to modern scientists involved in studies of the Earth’s environment that ultimately led to the definition of a new, eminently multidisciplinary science: biogeochemistry. This can therefore be defined as the science whose objects are the elemental chemical composition of the Earth, the chemical speciation of its components, stock dynamics of the main reservoirs and fluxes between them, under the simultaneous control of physical, chemical and biological reactions. Such a wide program underlines the inclusiveness of biogeochemical studies, whether it be through the establishment of working concepts on the global scale or through the definition of experimental sampling schemes and processing of supporting data of such concepts! Therefore, biogeochemistry occupies a special place among other more mono-disciplinary sciences. The global carbon cycle is at the center of current concerns of biogeochemists because biogeochemistry is inseparable from climate science. The concept of the biological pump perfectly illustrates this case. Autotrophic organisms living on the surface of our planet are indeed responsible for the annual fixing of around 120 Gt2 of carbon in almost equal proportions between the continents and oceans. In the oceanic compartment, diatoms, microalgae with siliceous cell walls, are responsible for nearly half of the primary production. This is one of the reasons why, albeit belatedly, marine biogeochemists became interested in the silicon cycle. Furthermore, despite the importance of primary marine production, only a small proportion of organic carbon formed is finally deposited and buried in deep sediments. Indeed, through 2 Gt: unit, a gigatonne (109 t), equal to a petagram (Pg, 1015 g).

Preface

xi

heterotrophic respiration, but also through that of autotrophs, the bulk of primary marine production is returned to the atmosphere as CO2 on short time scales ranging from minutes to years. This is one reason why the biological pump plays a smaller role than the physical pump in the annual cycle of carbon, the latter being responsible for more than 90% of the carbon annually swept along in the oceans by CO2 dissolution in surface water and subduction during the formation of intermediate and deep water. However, the carbon carried away by the physical pump will not be permanently sequestered, as the return of the deep water to the surface by the global thermohaline circulation will release it again to the atmosphere as CO2, on time scales less than 1,200 years. Ultimately, the biological pump will play the key role because, although less than 0.5% of the carbon fixed in surface waters accumulates in deep sediments, it is the only way to isolate carbon over geological time scales. The biological pump is a set of processes responsible not only for the processing and vertical advection of dissolved organic material, but also for the passive sedimentation flux of particulate organic material and associated biominerals as well as the active transport by the nycthemeral migration of zooplankton. This definition taken from Robinson et al. [ROB 10] underlines the unique role played by organisms in the pelagic realm, from bacteria to mesozooplankton. In this book, we will focus on a group of particularly important organisms for the functioning of the biological pump. These organisms represent a functional group of “biomineralizers” using silicon and are therefore dependent on the availability of this element to be able to develop. Several groups of organisms are represented here, but the diatom group is undoubtedly the one that plays the major role. Understanding the mechanisms that will govern the ability of diatoms to use the silicon cycle is therefore essential to understand their role in the biological carbon pump.

Bernard QUÉGUINER April 2016

1 The Chemical Forms of Silicon in the Marine Domain

1.1. The element “silicon” Silicon (symbol Si, atomic number 14) is not found in its native state1, but in the form of silicates it is the most abundant element in the Earth’s surface (about 28%), after oxygen. Its name comes from the Latin word silex, meaning stone. Lavoisier had suspected its existence in 1787, but it was not until 1811 that it was discovered by Gay-Lussac and Thénard. In 1823, Berzelius isolated silicon in a sufficient state of purity to be able to approach its study. Natural silicon is a metalloid that actually corresponds to a mixture of three stable isotopes. Atomic mass

Abundance (% atoms)

28

Isotope

279,769,265,325

92,223

29

28,976,494,700

4,685

30

2,997,377,017

3,092

Si Si Si

Table 1.1. Natural atomic masses and abundances of three stable isotopes of silicon [HAY 16]. The weighted average atomic mass of natural silicon is equal to 28.08549871

1 Several chemical methods, however, are used to prepare elemental silicon, particularly from the reduction of SiO2 at very high temperatures (~2000°C).

The Biogeochemical Cycle of Silicon in the Ocean, First Edition. Bernard Quéguiner. © ISTE Ltd 2016. Published by ISTE Ltd and John Wiley & Sons, Inc.

2

The Biogeochemical Cycle of Silicon in the Ocean

The silicon element also presents 20 radioactive isotopes [AUD 03]; the most frequently reported (and used) in biogeochemistry are 31Si, with a period of 2.62 h (transmitter β−, maximum energy of 1.49 Mev), and 32Si, known as a cosmogenic isotope, with a period of 132 years (transmitter β−, maximum energy of 0.22 Mev). 1.2. Orthosilicic acid Dissolved silicon in seawater is present in various chemical forms, eventually available to living organisms. The most abundant form is orthosilicic acid, which has the molecular formula H 4SiO 4 and is a tetraprotonated acid. The four protons in aqueous solution may be successively released during the formation of a series of monosilicate anions: H 3SiO −4 , H 2SiO42− , HSiO34− and SiO44− . In an aqueous saline solution at room temperature, the major forms of orthosilicic acid are H4SiO4 , H 3SiO 4− and H 2SiO 42− , and other anions can be overlooked [ING 78]. At a pH value close to that of seawater, only the undissociated form and the first deprotonated form may co-exist, the second being largely in the minority.

Figure 1.1. Relative distribution of the majority of the chemically dissolved species of silicon under thermodynamic conditions comparable to seawater (0.6 M NaCl, 25ºC) as a function of the pH of the solution

Thermodynamic constants for acid–base reactions of orthosilicic acid in saline solution (25°C, 0.6 M NaCl salinity, i.e. ≈34.5) were

The Chemical Forms of Silicon in the Marine Domain

3

determined by Sjöberg et al. [SJÖ 81]. They allow us to calculate relative concentrations of the different species of orthosilicic acid with the formulation by Riedel and Nelson [RIE 85]:

[H SiO ]⋅ a 3

− 4

H+

[H 4SiO4 ]

= 10−9,47

⎡ H 2SiO 42− ⎤ ⋅ a + ⎣ ⎦ H = 10−12,6 ⎡ H3SiO 4− ⎤ ⎣ ⎦

[1.1]

[1.2]

where aH+ is the activity of protons. The dissolved silicon is equally likely to react with metal cations such as Ca2+ and Mg2+. The values of the equilibrium constants determined by Santschi and Schindler [SAN 74] show that for the pH values of seawater and the usual concentrations ([H4SiO4]total≤ 10–4M; [Ca2+] ≤10–2M; [Mg2+] ≤5 10–2M), the complexes formed are quite negligible. However, the possible formation of such complexes under acidic pH must be considered when preparing the stock solutions used in the different incubation experiments (and acidification solutions during purification for the removal of metals may result in the formation of such complexes). Another aspect of silicon chemistry, which is often overlooked, is the possible formation of complexes with Fe(III). As noted by Ingri [ING 78], these complexes can be present in significant quantities in natural freshwaters and must be taken into account in limnological biogeochemical cycles. In the marine environment, given the low concentrations of mineral iron, the formation of these complexes can be ignored. 1.3. Particulate silicas

The global silicon cycle is governed by the fluxes between dissolved and particulate stocks. The latter is represented by a geological component (the lithogenic silica) and a biogeochemical

4

The Biogeochemical Cycle of Silicon in the Ocean

component (biogenic silica or opal). As part of this work, we will focus mainly on biogenic silica and biogeochemical exchanges controlled by the biological production processes and the physicochemical process of dissolution from tests of organisms with siliceous walls. 1.3.1. Lithogenic silica

The silica contained in rocks and the different forms of silicates that make them up are collectively referred to as lithogenic silica. Given the existence of different crystalline forms, lithogenic silica thus defined has variable physicochemical properties, particularly with respect to the dissolution by basic solvents (including seawater). The geochemical role of silicates is essential because they represent more than 90% by weight of the Earth’s crust. These silicates are quasiregular tetrahedrons of the general formula (Si, Al)O4, the centers of which are occupied by atoms of Si or Al and the tips by O. There are different types of three-dimensional structures, the details of which can be found in books on crystallography. In summary, in tectosilicates, tetrahedra form a three-dimensional framework, (Si, Al)xO2x, wherein each of the oxygen atoms is shared by two tetrahedrons. When the centers of the tetrahedra are all Si atoms, the network, which is electrically neutral, has the formula SiO2 and corresponds to different forms of lithogenic silica such as quartz, tridymite and cristobalite. If some of the Si atoms are replaced by Al atoms, the structure is a macroanion whose negative charge is compensated by cations, such as in feldspars, feldspathoids or zeolites. Phyllosilicates (micas, chlorites, clay minerals, etc.), which are easily cleavable, are characterized by plane sheets of tetrahedrons associated with three of their tips, wherein the degrees of substitution of silicon by aluminum may vary. Quartz, and its many varieties, is by far the most important species; it is also one of the major minerals of many rocks – especially igneous rocks – which are found in excess (saturation) or deficit in silica. The process of alteration of igneous rocks, however, results in the formation of clay minerals, collectively referred to as aluminosilicates, mainly as phyllosilicates. The

The Che emical Forms off Silicon in the Marine M Domain

5

minerallogical compposition of thhese differen nt forms of liithogenic siliica will coontrol their susceptibility s y with respecct to the dissolution in tthe marine environmennt. The dissoolution of liithogenic sillica thus hass a high vaariability at thhe regional level. l 1.3.2. Biogenic silica s Bioggenic silica is the fracttion of partiiculate silicaa produced by living organisms. o I has an amorphous, thaat is to say non-crystallin It n ne, structurre in which aluminum atoms can sometimes be substituted ffor silicon atoms. In biiogenic silicaa, disordered d tetrahedra are a coordinatted with waater moleculles (Figure 1..2).

Figure 1.2. Structural model of bio ogenic silica. Al A enters the network n structu ure nment resultin ng from sharring while prreserving the three-dimenssional environ tetrahed dra SiO4. The substitution of o Si4+ by Al3+ generates a negative n charg rge. Chemical analysis off the diatom frustule f sugge ests charge compensation c by Ca2+catiions [GEH 02]]

Bioggenic silica is i a hydratedd polymer haaving a degreee of hydratiion that vaaries dependding on the nature of the t material.. Hecky et al. [HEC 73] 7 indicatedd that diatom ms can prod duce a range of amorphoous silica structures s off the generaal formula (SiO2, nH2O). O In oceannic sediments, the averrage formulaa of biogenic silica is (S SiO2, 0.4 H2O) F [M MOR 89]. The T degree of accordiing to Morrtlock and Froelich hydratiion of biogenic silica also seems to play an im mportant role in the dissolution proocesses. Bioggenic silica also containns a number of

6

The Biogeochemical Cycle of Silicon in the Ocean

organic molecules originating from the process of polymerization and mainly composed of proteins and polysaccharides. This organic matter can be preserved in sedimentary opal and represents a material of choice for paleoceanography, especially allowing us to trace the evolution of the isotopic ratios of carbon and nitrogen (e.g. [CRO 02]).

2 Techniques for Studying Stocks and Fluxes

2.1. Techniques for the chemical analysis of silicon The determination of orthosilicic acid is generally carried out by spectrophotometry according to the colorimetric method of Mullin and Riley [MUL 55], adapted by Strickland and Parsons [STR 72]. Leblanc [LEB 02] conducted a review of the main methods for the determination of particulate silica. The dosage of biogenic silica in the marine environment is relatively recent in oceanography, as it was introduced in the late 1950s by Goldberg [GOL 58]. Early techniques that have been implemented (X-ray diffraction, infrared spectroscopy, digestion with hydrofluoric acid (HF) and colorimetric analysis of orthosilicic acid) were not selective and did not allow us to discriminate between biogenic silica and lithogenic silica. Several techniques are available today, according to the type of material being looked at: filtered samples on 0.6 micron porosity polycarbonate membrane (for samples of the water column and some sediment traps, including drifting traps) or concentrated samples after lyophilization (for sediment samples or some benthic sediment traps, including fixed traps deployed for long periods).

The Biogeochemical Cycle of Silicon in the Ocean, First Edition. Bernard Quéguiner. © ISTE Ltd 2016. Published by ISTE Ltd and John Wiley & Sons, Inc.

8

The Biogeochemical Cycle of Silicon in the Ocean

2.1.1. The sequential digestion method The development of an alkaline digestion method by Paasche [PAA 73] has made it possible to achieve the first biogenic silica measurements in the marine environment. Biogenic silica is extracted at a temperature of 100°C in an alkaline medium (NaOH or Na2CO3), resulting in the oxidation of opal into orthosilicic acid. Quantitation is then performed by colorimetric assay of orthosilicic acid. The advantage of this method is to allow the determination of lithogenic silica from the same filter, in a second extraction in a strong acid medium (2.9 N HF) at room temperature for 48 h, following the technique subsequently developed by Nelson et al. [NEL 89]. If the sequential digestion method is most commonly used for routine dosage of opal in the open ocean area, its application remains delicate for coastal environments. According to the sample composition, a variable part of lithogenic silica, more or less labile given its aluminosilicate content, can be attacked at the first basic extraction, leading to an overestimation of the biogenic silica measured. Samples coming from coastal areas may contain high amounts of lithogenic silica, because of river and wind contributions, coastal erosion or sediment resuspension phenomena on continental shelves. Ragueneau and Tréguer [RAG 94] showed a linear relationship between biogenic silica and lithogenic silica in a sample set of the Bay of Brest and the English Channel, and estimated the proportion of lithogenic silica dissolved during the first extraction as 15%, which led them to propose a statistical method of data correction. This correction is determined from the percentage of interference for samples collected in winter, the period in which the maximum interference is assumed, as concentrations of biogenic silica are very low and lithogenic silica concentration is generally high. This correction factor, once determined, is then extrapolated to the data of the rest of the year. Note that the correction is applicable at the very best at regional level and most often at a very local level. Moreover, this correction is usually impossible to achieve during oceanographic cruises, often conducted in a productive period for which winter references are not known, thus preventing the evaluation of the relationship between the two types of silica and, therefore, the generation of such a correction factor.

Techniques fo or Studying Stoccks and Fluxes

9

2.1.2. The extrac ction kinetic cs method d Anoother methood, known as extracction kinetiics, originaally developped for sediment sampples by DeM Master [DEM M 81], is allso widely used and coorrects for mineral m interfference, withhout prejudgiing mposition or the amount of lithogeniic silica present; it is bassed the com on diffferences in thhe dissolutioon rates of bo oth siliceouss fractions. T The sample is hot extraacted in a baasic medium m (NaOH or Na2CO3) forr a ved silica iss measured at period of several hours, and the dissolv differennt times duuring the exxtraction (Fiigure 2.1). Although thhis methodd is one of thhe most reliaable, it is ted dious and diffficult to appply for routtine assays of o a large num mber of samp ples.

Figure 2.1. 2 Typical extraction kinetics of particulate silica. The e first part of the curve re epresents the dissolution of o the BSi (hig gh-dissolution n rate), while the linear ph hase reached d at the end of o extraction corresponds c to o the dissoluttion phase of o LSi (at a theoretically t m much lower dissolution d ratte in an alkalline medium m). The correcction is to dra aw the tangen nt to the linea ar portion of the curve. The T extrapola ated interceptt at time t0 equals e the corrrected bioge enic silica co ontent of the sa ample

2.1.3. The correc ction by alu uminum me ethod Anoother methood, still seeldom used d, uses aluuminum as a discrim minating facttor between lithogenic silica and biogenic b silicca. Aluminnum is in facct an element which is in ncorporated into i the diatoom frustulee but in veryy small propoortions. The Al/Si ratio of o diatoms iss in the ordder of 10−2 while w it is usuually close to o 2 for the liithogenic siliica

10

The Biogeochemical Cycle of Silicon in the Ocean

[LEW 61, KAM 74]. Some authors have, therefore, suggested the use of aluminum to correct the biogenic silica measurements [EGG 80, SCH 98]. One method performs a triple extraction on the same sample and uses the dissolution of aluminum during three successive digestions as an indicator of the dissolution of labile lithogenic silica [RAG 05]. Initially, the sample is extracted in sodium hydroxide for 45 min using the method described by Nelson et al. [NEL 89]. Following this extraction, the respective concentrations of silicon and aluminum in the solution are quantified, according to the conventional colorimetric method for silicon and by inductively coupled plasma– atomic emission spectroscopy (ICP–AES) or the fluorometric lumogallion method of Ren et al. [REN 01] for aluminum (Si1 and Al1). A second extraction is carried out under the same conditions as the first, and the dissolved silicon and aluminum are again measured (Si2 and Al2). The final digestion is carried out with 2.9 N HF, in which only the dissolved silica is measured (Si3) to quantify lithogenic silica. It is considered that the entire biogenic silica is extracted in the first digestion with a fraction of labile lithogenic silica in the alkaline medium while during the second extraction, only the residual labile lithogenic silica is dissolved. In this case, the Si/Al ratio of the second extraction is representative of the composition of the labile lithogenic silica. This ratio is then used to correct for the dissolution of the lithogenic silica during the first extraction. The corrected amount of biogenic silica is then obtained according to the following relationship:

[BSi ]corrected

⎛ Si ⎞ = Si1 − ⎜⎜ 2 × Al1 ⎟⎟ . ⎝ Al2 ⎠

[2.1]

The amount of silica lithogenic is, in turn, determined by the equation:

[ LSi ]corrocted = Si3 + Si2 + Si1 − [ BSi ]corrected

[2.2]

A second method based on the analysis of the Si/Al ratio, the method of Kamatani and Oku [KAM 00], is well suited to samples of

T Techniques for Studying Stockks and Fluxes

11

sediment traps andd sediments. The extraction is carriied out from m a lyophillized amountt of material in a kinetics assay of thee Si/Al ratio ffor 2 h. The T entire BSi B is dissoolved quickly in the firrst 20 min of extractiion while thee silicate minnerals are attacked graduaally.

Figure 2.2. Extraction n kinetics of silicon s a) and aluminum b) from a sediment B under va arious alkaline e conditions (●: 0.2 N NaO OH; sample from Tokyo Bay ○: 0.1 N NaOH; ▲: 0..5 M Na2CO3; △: 0.1 M Na a2CO3). The lin near relationsship between n these two ellements c) ena ables us to calculate the am mount of bioge enic silica off the sample (intercept ( of th he line on the y-axis) as co orrected from the interfere ence of the lith hogenic silica. The slope of the regression n line represents the Si/A Al ratio of the liithogenic mate erial (accordin ng to [KAM 00]])

2.2. Te echniques for the ana alysis of silicon fluxes Orthhosilicic acidd is taken upp by siliceou us organism ms before beiing depositted as biogeenic silica. Taken togetther, these two t processes, of deposittion and upttake, are groouped underr the term “production “ biogeniic silica”. On the death of organism ms, orthosilicic acid will be regenerrated by disssolution of biogenic b silicca. All thesee fluxes can be measurred using isotopic labelling techniqu ues (i.e. usiing radioactiive isotopees or stable isotopes) and, a more recently, r by labeling w with fluoropphores enabling us to identify acctive organiisms within a planktoonic assemblage. 2.2.1. Labeling with w radioac ctive isotop pes Thiss technique allows us too measure only the uptaake rate thatt is consideered equal too the deposition flux and d, therefore, considered c as a measurrement of thhe productioon of biogeenic silica. In practice,, a

12

The Biogeochemical Cycle of Silicon in the Ocean

seawater sample with a known concentration of orthosilicic acid, [H4SiO4]i, is inoculated with a radioactive tracer with a known total activity ASif. After an incubation in conditions as close as possible to the natural environment (in situ mooring or deck incubator), during the time Δt (usually over a period of 24 h), the sample is filtered through a polycarbonate membrane of 0.6–0.8 µm porosity, and the radioactivity of the particulate material retained on the filter (ASif) is then measured. The absolute production rate ρSi is then calculated by the following equation:

ρ Si =

(

ASi f ⋅ [ H 4SiO 4 ]i ASii Δt

)

[2.3]

In the past, very few studies have been conducted using 31Si because the short period of this element (2.62 h) requires us to complete the work in the immediate vicinity of a nuclear research reactor to produce neutrons for irradiating natural silicon (see [AZA 74a]). These difficulties have led us to favor the use of a radioactive isotope of germanium: 68Ge [AZA 74b, AZA 74c]. This is a β– emitter considered as a structural analogue of silicon due to its position in the periodic table. However, its steric hindrance is twice as large as that of silicon, and this element is known to induce, at high concentrations, frustule malformations making it hard to justify its use as a silicon tracer. A number of studies, including the study by Sullivan [SUL 76, SUL 77], have been performed but certain evolutions, including chase kinetics, suggest that germanium is partially adsorbed on the frustule and not totally taken up. Since the end of the 20th Century, it has been possible to use a long-lived isotope: 32Si (period: 132 years). This tracer is regularly produced in the accelerator of Los Alamos by the US Department of Energy. However, its high production cost and reduced availability restrict a wider utilization. The utilization of 32Si was introduced into oceanography by Treguer et al. [TRE 91]. This method, described in detail by Brzezinski and Phillips [BRZ 97], is based on the utilization of 32Si, which is called cosmogenic radioactive as it is produced naturally in

Techniques for Studying Stocks and Fluxes

13

the atmosphere under the action of cosmic rays on 40Ar [LAL 60]. Artificially, there are a number of methods for radiosynthesis from sulfur, chlorine and vanadium. It is produced by the proton-induced spallation of a potassium chloride target by Oak Ridge National Laboratory. 32Si is a β− emitter, with a maximum energy of 221 keV. It disintegrates into a daughter element, 32P, according to the following reaction: β- (221 keV) 32

Si

→ τ = 132 years

β- (1.71 MeV) 32

P



32

S (stable)

τ = 14.2 days

Principle of secular equilibrium (according to [AND 03]): in a closed system inside which there is no physical separation between the parent element and the daughter element, a series of radioactive decays achieves a state of equilibrium known as secular equilibrium, in which the rate of radioactive decay of every daughter element is equal to its rate of production. At a steady state, in the closed system, the rate of production of a daughter element is equal to the rate of decay of its parent element; in other words, the parent element’s activity is equal to the daughter element’s activity. If the system is open and since the physical, chemical or biological processes separate the parent element from its daughter element, for example during the differential uptake of silicon and phosphorus, a transitory1 radioactive disequilibrium emerges, in which the respective activities of parent and daughter are different. 32

P may be measured by the Cerenkov effect. This is a luminous phenomenon that appears when a charged particle crosses a medium with a speed that is higher that the speed of electromagnetic waves in this medium. The minimum energy of particles β- for the production of the Cerenkov effect in water is 0.27 MeV. In practice, the secular equilibrium (after 2 or 3 months) can be achieved when the quantity of the absorbed 32Si is identical to the equilibrium activity of 32P.

1 After several disintegrations, the system will return to secular equilibrium.

14

The Biogeochemical Cycle of Silicon in the Ocean

cytoplasm frustule

Figure 2.3. Principle of the determination of the production of 32 32 silica and phosphorus. Evolution of the activity of P and of Si in the diatoms after the incubation (from [LEY 93])

The Cerenkov emission by 32P allows us to measure the activity without the addition of a scintillation cocktail after returning to the secular equilibrium in the sample filter. In practice, experimenters usually prefer to use the classical liquid scintillation, which allows a faster measurement after the incubation. Theoretically, the 32Si technique may also allow us to evaluate the uptake of phosphorus and silicon simultaneously as orthosilicic acid transforms after the disintegration of silicon into orthophosphoric acid that dissociates into orthophosphate ions, the bioavailable form of phosphorus. In practice, the differences in turnover of the two elements do not allow us to study them simultaneously. Incubations for measuring biogenic silica production usually last 24 h and the resulting incorporation of phosphorus at the end of this period most probably correspond to its storage by microalgae (phosphate is likely to be recycled several times over a 24 h cycle). 2.2.2. Labeling with stable isotopes This technique is the only way to directly evaluate the fluxes of dissolution in water samples. It was introduced by Nelson and Goering in [NEL 77a] for measurements of production and in [NEL 77b] for measurements of dissolution.

Techniques for Studying Stocks and Fluxes

15

The stable isotope generally used is 30Si, in other words, the rarest in nature (3.092%). As for labeling with the radioactive isotope, a sample of seawater, in which the concentration of orthosilicic acid is known [H4SiO4]i, is inoculated by a stable tracer whose isotopic composition is known (contrary to radioactive isotopes, stable tracers are not isotopically pure but enriched in one isotope as compared to the natural element). After the incubation carried out under the identical conditions to those described above, the sample is filtered through a polycarbonate membrane of 0.6–0.8 µm porosity, and the isotopic composition of the material retained on the filter is then measured by using a mass spectrometer. The formula describing the isotopic equilibrium at the end of incubation is as follows: 30

A f ⋅ B f = 30 An ⋅ Bi + 30 Ai ⋅ ΔB

[2.4]

where 30Af = concentration of the tracer 30Si (atom %) in the particulate silica at the end of incubation; 30An = natural concentration of the tracer 30Si (atom %); 30Ai = initial concentration of the tracer 30 Si (atom %) in the dissolved phase; Bf = final concentration of the particulate silica; Bi = initial concentration of the particulate silica; ΔB = augmentation of the particulate silica. Equation [1.5] may be rearranged as follows:

( ΔB =

30

Af ⋅ B f − 30 An ⋅ Bi ) 30

Ai

, [2.5]

With Bf = Bi +ΔB, the relative increase of the element in the particulate phase, may be calculated as follows:

( ΔB =

30

A f ⋅ Bi

+

30

Af ⋅ ΔB − 30 An ⋅ Bi ) 30

Ai

⇔ ΔB Bi = ( 30 Af − 30 An )

(

30

Ai − 30 Af ) .

[2.6] [2.7]

16

The Biogeochemical Cycle of Silicon in the Ocean

In other words, the specific uptake rate, the relative increase of the element in the particulate phase during the incubation Δt, is given by the following expression: VSi =

( 30 Af − 30 An )

[2.8]

( 30 Ai − 30 Af ) ⋅ Δt

The absolute production rate ρSi is finally calculated by the following equation:

ρ Si = B f ⋅

( 30 Af − 30 An )

[2.9]

( 30 Ai − 30 Af ) ⋅ Δt

During the incubations, a part of biogenic silica may also be dissolved, which will increase the isotopic ratio 28Si/30Si. The measurement of this ratio in the dissolved phase allows us to evaluate the rate of dissolution of biogenic silica δSi, according to the calculation presented above. To evaluate the isotopic ratio in the dissolved phase, orthosilicic acid should be extracted as described previously, for example, by co-precipitation using the method introduced by Reynolds et al. [REY 06] and adapted by Rimmelin-Maury et al. [RIM 07]. The absolute dissolution rate ρdis is obtained from the final expression [NEL 77b]:

ρ dis Si = [ H 4SiO 4 ]t ⋅

28

A f − 28 Ai

Δt ⋅ ( 28 An − 28 Ai )

,

[2.10]

where 28Af = concentration of 28Si (atom %) in the particulate silica at the end of incubation; 28An = natural concentration of 28Si (atom %); 30 Ai = initial concentration of 28Si (atom %) in the dissolved phase after adding the tracer 30Si (or 29Si). Under these conditions, Nelson and Goering [NEL 77b] defined the specific dissolution rate, Vdis Si, as the ratio of the absolute rate to the initial concentration of particulate silica:

Vdis Si =

ρ dis Si Bi

[2.11]

Techniques for Studying Stocks and Fluxes

17

Specific dissolution rates are usually one order of magnitude lower than the specific uptake rates. However, some results obtained in the upwelling areas suggest that the orders of magnitude may sometimes be similar. In such a case, from a rigorous point of view, the estimation of production rates should be corrected from the isotopic dilution during the incubation. This correction can be carried out using the nonlinear two-compartment model [DEB 05, ELS 07] that allows us to estimate the specific uptake and dissolution rates simultaneously. The fluxes can be calculated by resolving a system of four equations [CLO 14], in which æ H4SiO4 and æ BSi present the excess of tracer (in atom %) in relation to natural ratio in two phases at the initial time (index i) and the final time (index f):

[H 4 SiO 4 ] f = [H 4 SiO 4 ]i + (δSi − ρSi ) ⋅ Δt

[2.12]

[ BSi ] f = [ BSi ]i + ( ρ Si − δ Si ) ⋅ Δt

[2.13] δ Si

æ H 4 SiO4 f

⎛ ⎞ ρ Si −δ Si δ Si − ρ Si = æ H 4 SiO4i ⋅ ⎜ 1 + ⋅ Δt ⎟ ⎜ [ H 4 SiO4 ] ⎟ i ⎝ ⎠

ρ Si ⎛ ⎞ ⎞ ρ Si −δ Si ⎟ æ H 4 SiO4i ⋅ [ H 4 SiO4 ]i ⎜ ⎛ δ Si − ρ Si ⋅ ⎜1 − ⎜1 + ⋅ Δt ⎟ æ BSi f = ⎟ [ BSi ]i + ( ρ Si − δ Si) ⋅ Δt ⎜ ⎜⎝ [ H 4 SiO4 ]i ⎟⎠ ⎟ ⎝ ⎠

[2.14]

[2.15]

2.3. Silica deposit labeling and cellular imaging

Isotopic labeling techniques allow us to estimate the global production of a community of siliceous organisms. For studying the dynamic of such a community, it is necessary to have access to the information at the organism level. In the past, 32Si was used by Shipe and Brzezinski [SHI 99] to perform microautoradiography2 that helps to determine the levels of production of different species. Such a technique, however, is difficult to apply, as it requires a very long duration of development that can last up to a year. 2 This technique allows us to visualize the areas of the deposit of a radioactive tracer by microscopical examinations.

18

The Biogeochemical Cycle of Silicon in the Ocean

a)

b)

Figure 2.4. a) Visualization with microautoradiography of the synthesis of intercalary bands in Rhizosolenia debyana (scale: 20 μm) [SHI 99]. b) Labeling of a chain of Rhizosolenia styliformis by PDMPO (scale: 100 μm, photograph from M. Lasbleiz)

Therefore, the utilization of fluorescent markers must be favored. The first study was conducted by Brzezinski and Conley [BRZ 94] who used a protocol of labeling with rhodamine 123 to study the silicification of Thalassiosira weissflogii in culture. Today, many markers are available, but we use PDMPO3 [QUE 11]. PDMPO is a pH probe that fluoresces in an acid environment, which is typically the case with diatom silica deposition vesicle (see section 4.2). PDMPO was used for the first time by Shimizu et al. [SHI 01], and a precise protocol was then introduced by Leblanc and Hutchins [LEB 05] who were the first to use this marker to study the dynamics of a natural community of diatoms in the Delaware estuary. The PDMPO method is easy to use, as it demands only an incubation of the natural community in the presence of the marker that can be realized in parallel with the quantitative measurements by isotopic techniques. 2.4. Isotopic fractionation of silicon and utilization of δ30 Si as a tracer in oceanography

All the isotopes of an element have the same electronic structure which gives them the same chemical properties. Conversely, the differences in mass lead to the differences in the vibration energy so that the different isotopes have different vibration frequencies. These differences are responsible for the fractionation between the two 3 [2-(4-Pyridyl)-5{[4-(2-dimethylaminoethylamino-carbamoyl)-methoxy]phenyl}oxazole].

Techniques for Studying Stocks and Fluxes

19

phases at equilibrium. Hence, during a chemical reaction, molecules formed from the light isotope will react slightly more rapidly than the molecules containing the heavy isotope. A kinetic fractionation is added to this fractionation at equilibrium (for example, in a gas, the molecules containing the light isotope move more quickly than the molecules containing the heavy isotope). The concentrations of stable isotopes are generally expressed as the relative difference of the isotopic ratio, δ, which compares isotopic ratio of the sample to the isotopic ratio of a reference material. This notation leads to the higher precision, being independent of the type of measuring equipment. If one examines the number of atoms of an element E, having two isotopes of different mass iE and jE in the same phase, the relative difference of the isotopic ratio can be calculated by the following equation (according to [COP 11]):

δ iEstd =

(i E/ j E)sample (i E/ j E)std

−1

[2.16]

where (iE/jE)sample is the ratio between the atoms’ number of the isotope iE and the atoms’ number of the isotope jE in the sample and (iE/jE)std is the equivalent isotopic ratio of the reference material. By convention, the exponents i and j represent the respective atomic masses of the heaviest and lightest isotopes, and διΕστδ is expressed as “per mil” (‰). Consequently, when a sample is known to be “heavier” than the reference material, the parameter δ is more positive than that of the reference material; in other words, the average atomic mass of the element in the sample is higher than the average mass of the reference material [COP 11]. As for silicon, we generally compare the least abundant heavy isotope in relation to the most abundant one, or 30 Si to 28Si. The standard international reference material is the sand of pure silica (SiO2) with a granulometry between 100 and 177 µm, called “NBS 28 (Silicon and Oxygen Isotopes in Silica Sand)”, prepared by the National Institute of Standards and Technology (NIST) under the reference no. 8546. The relative difference of the isotopic ratio of 30Si in relation to 28Si should be denoted as δ30SiNBS 284.

4 In practice, this ratio is frequently denoted as δ30Si.

20

The Biogeochemical Cycle of Silicon in the Ocean

The evolution of the isotopic ratio during the transformation of a reactant into a product (e.g. passage of orthosilicic acid to biogenic silica) is of the general form of the Rayleigh distillation (as shown in Figure 2.5) if the following three conditions are realized: 1) the product of the reaction is continually extracted from the reaction medium; 2) the fractionation during the reaction is characterized by the fractionation factor α; 3) the factor α is constant during the whole reaction. The Rayleigh equation states that the isotopic ratio (R) of a reactive reservoir during its transformation is a function of the initial isotopic ratio (R0) of the reactive fraction remaining in the reservoir (f), and of the fractionation factor at equilibrium for the reaction (α), as follows:

R = R0 ⋅ f (α −1)

[2.17]

Figure 2.5. Simulation of silicon isotopic fractionation related to the precipitation of the biogenic silica (bSiO2) by diatoms in a closed system (Rayleigh distillation) and in an open system at equilibrium. For a color version of the figure, see www.iste.co.uk/queguiner/silicon.zip

Techniques for Studying Stocks and Fluxes

21

The fractionation factor α was estimated by De La Rocha et al. [DEL 97] by the following equation: ⎛ ⎛ (1 + δ 30 SibSiO acc / 1000) ⋅ (1 − f ) ⎞ ⎞ 2 ln ⎜1 − ⎜ ⎟⎟ ⎟ 30 ⎜ ⎜ ⎟ (1 δ Si0 / 1000) + ⎝ ⎠⎠ ⎝ , α= ln f

[2.18]

where δ30SibSiO2 acc refers to biogenic silica accumulated during the reaction and δ30Si0 refers to orthosilicic acid at f = 1. Sutton et al. [SUT 13] pointed that the fractionation factor at equilibrium α is related to another parameter frequently used in biogeochemistry, the isotopic fractionation ε that represents the difference in δ30Si between orthosilicic acid and biogenic silica, according to the relationship:

ε ≅1000 ⋅(α −1)

[2.19]

The following equations, according to Varela et al. [VAR 04] and Sutton et al. [SUT 13], allow us to represent the isotopic fractionation of silicon in the different phases in an oceanic system:

δ 30 Si H SiO = δ 30 Si 0 + 1000 ⋅ (α − 1) ⋅ ln f

[2.20]

δ 30 Si bSiO

4

4

2

inst

= δ 30 Si H 4SiO4 +1000 ⋅ (α −1)

[2.21]

δ 30 Si bSiO

acc

= δ 30 Si 0 −1000 ⋅ (α −1)⋅ [ f ⋅ ln f / (1− f )]

[2.22]

2

The Rayleigh model ignores the whole processes of exchange of orthosilicic acid at the boundary of the system as well as the dissolution of biogenic silica inside the system. Some authors have considered it to be not especially representative of the natural conditions, and Varela et al. [VAR 04] proposed to use a model of an open system at equilibrium; in other words, continuous inputs of orthosilicic acid, of known isotopic composition, compensate for the loss related to the extraction in the particulate phase. The equations characterizing such a system then become:

δ 30 Si H SiO = δ 30 Si 0 −1000 ⋅ (α −1)⋅ (1− f ) 4

4

[2.23]

22

The Biogeochemical Cycle of Silicon in the Ocean

δ 30 Si bSiO

2

inst

= δ 30 Si H 4 SiO4 + 1000 ⋅ (α − 1) ⋅ f

[2.24]

The open system model is not without any disadvantage, as it is possible that the system may be resupplied by different processes (horizontal advection, deepening of the mixed layer) whose characteristics may vary during the season. Both models strongly differ in the consumption of orthosilicic acid exceeding 70% of the initial stock and should be confronted with the experimental information obtained during cruises. Some situations are actually closer to one or other models. In their study conducted in the Antarctic Circumpolar Current, Varela et al. [VAR 04] concluded that the surface waters evolve during the season between an open system and a closed system. The utilization of the Rayleigh model imposes that the production of the opal should be operated from an invariant reservoir (= of infinite size) of orthosilicic acid. Such conditions are frequent during the initial stages of diatom blooms within the surface mixed layer. However, such an approach is not valid for strongly stratified systems presenting chronically weak nutrient concentrations. It is not valid for weakly stratified systems either whose mixed layer depths are close to the critical depth for diatom growth. In these two cases, a model based on a continuous flux of nutrients in the surface layer will be more appropriate. In the model with a continuous flux, opal and ambient orthosilicic acid δ30Si signatures will gradually increase, as H4SiO4 is utilized until an equilibrium is established between nutrient input and biological utilization. Strongly stratified oceanic ecosystems present the uniform nutrient concentrations over time and space (endless summer); these systems are close to a state of equilibrium (steady state), and the values of δ30Si of opal should not vary with the utilization of H4SiO4 and should be equal to δ30Si values of the orthosilicic acid entering the surface layer. 2.4.1. Demonstration of the isotopic fractionation by the diatoms

In order to use the natural isotopic ratio, it is necessary to show the existence of the biological fractionation and to know if the variations of the isotopic ratio are big enough to be measured with the precision

Techniques for Studying Stocks and Fluxes

23

allowed by the means of analysis. The first evidences of fractionation were obtained in the 1980s. Douthitt [DOU 82] was the first to show the big difference that characterizes the natural abundance of the silicon isotopes (28Si/29Si/30Si) in biogenic silica as compared to terrigenous materials of magmatic or plutonic rocks, as well as clay minerals. Then, Spadaro [SPA 83] was the first to show the isotopic fractionation by Phaeodactylum tricornutum. Isotopic fractionations are particularly known to be sensitive not only to variables of the environment (temperature, salinity, nutrient concentration, etc.) but also to biological conditions (taxonomy, growth rate, etc.) [SHA 73, THO 95, LAW 95]); on the other hand, as Phaeodactylum tricornutum can hardly be considered as indicative of all diatoms, De La Rocha et al. [DEL 97] studied the validity of the results obtained by Spadaro [SPA 83] for other species in culture under different growth conditions. De La Rocha et al. [DEL 97] showed that the fractionation of silicon isotopes takes place for three species of diatoms in culture (Skeletonema costatum, Thalassiosira weissflogii and Thalassiosira sp.). The values of the parameter α are almost identical for the three species with an average value of 0.9989 ± 0.0004, corresponding to an isotopic fractionation ε = −1.1‰. Furthermore, for these three species, the fractionation factor does not seem to vary significantly with temperature in their tested range (between 12 and 22°C). However, the study by Sutton et al. [SUT 13] reported the first denial of the lack of interspecific variations among nine species of diatoms, with the values of parameter ε varying between −0.54‰ for Fragilariopsis kerguelensis and −2.09 ‰ for Chaetoceros brevis, two diatoms from the Southern Ocean. 2.4.2. Utilization of δ 30Si as a tracer in oceanography

The isotopic variation of silicon is a practical means to estimate the production of biogenic silica present in water at a time scale ranging from a couple of weeks to several years (e.g. [CLO 14]). The production of opal may be estimated by determining, using the open or closed system models, the residual fraction of orthosilicic acid f from δ30Si values of orthosilicic acid brought into the surface layer (in other words, the δ30Si of underlying water and from the isotopic composition of diatoms sampled by sediment traps or phytoplankton

24

The Biogeochemical Cycle of Silicon in the Ocean

nets). On a smaller time scale, from hours to a day, the utilization of δ30Si does not present any advantage in relation to the methods of incubation in the presence of an added tracer (30Si or 32Si). On the time scale of the last glaciations, the utilization of δ30Si of diatoms from sediments is a really useful proxy5 to evaluate the relative utilization of orthosilicic acid. This utilization may provide two hypotheses, which deserve to be strictly verified in future studies: 1) the parameter α remains constant at the geological time scale; 2) the δ30Si does not undergo any modification within the sediments at the same time scale. As mentioned by Sutton et al. [SUT 13], the existence of an interspecific variability in the fractionation process makes it necessary to undertake a careful examination of the composition of thanatocœnoses6 to interpret correctly the observed variations of δ30Si. 2.4.3. The interest of analyses of the isotopic ratio of silicon

Since the pioneer works of Urey [URE 47] related to the dependence of water and carbonates δ18O on temperature, the analyses of the isotopic composition of major elements (C, H, N and O) in the biogenic material of the oceans have provided a lot of information about the oceanic processes, from the initial study field of temperature changes [EMI 55] to the estimation of nitrate fluxes brought into the euphotic zone [ALT 85]. Silicon potential as an isotopic tracer was not explored for a long time. The first works concerning this element date back to 1997 with the first methodological development by De La Rocha et al. [DEL 97]. Biogenic silica, particularly produced by diatoms, is an abundant material well-suited to silicon isotopic analyses. Diatoms are the main 5 The term proxy, commonly used in paleoceanography, refers to an indicative parameter used to quantify the intensity of a given process. 6 Thanatocœnoses: fossilized remains of organisms from the same biota deposited in sediments.

Techniques for Studying Stocks and Fluxes

25

organisms that control the oceanic cycle of silicon, whereas the contribution of other groups such as radiolarians and silicoflagellates probably remain minor, at least at the global scale [NEL 95]. Therefore, on average, an atom of silicon that enters the ocean will pass through the diatoms 25 times before being buried in deep sediments [TRE 13]. The latter may contain up to 80% of opal in dry weight in regions where diatoms are abundant [BRO 82]. There are not many possibilities to reconstruct the past productivity of diatoms and evaluate the contribution of these organisms to the ancient biochemical cycles and to the evolution of the climate. The indices, which are based on the accumulation rates of the sedimentary opal, are usually biased by the processes affecting the sediments during their sinking and their accumulation (redistribution by focusing– winnowing, dissolution and diagenetic processes). The normalization of opal records with the help of natural radioelements reacting with the sinking particles may allow us to correct for the effects of horizontal redistribution (230Th) and the dissolution in the bottom sediments (231Pa/230Th) but not from the dissolution during sinking before deposition in the oceanic bottom. Meanwhile, this dissolution during transfer through the water column is high and variable according to the biological, physical and chemical environmental conditions. The isotopic analysis of silicon found a privileged field of application in the Southern Ocean, which is the place where most of the world’s deposits of siliceous sediments take place [DEL 97]. Until the development of this proxy, the isotopic studies, necessary for the paleoceanography of Antarctic, were simply impossible, considering the absence of carbonaceous sediments in this oceanic region [HOW 94]. The isotopic analysis of silicon in the Southern Ocean emerged as a promising tool due to the role played by this ocean in the regulation of the atmospheric concentrations of CO2, as well as considering the fact that modifications in the nutrient utilization in this largest HNLC7 system of the global ocean exert a much bigger impact on the flux of CO2 than that in any other oceanic region [SAR 91].

7 HNLC: High Nutrient Low Chlorophyll [MIN 92].

3 The Marine Producers of Biogenic Silica

In the marine environment, silicification is a process that occurs in many unicellular and pluricellular organisms. Silicification, long considered as the only privilege of diatoms, radiolarians and sponges, seems today much more widespread within the planktonic organisms, independent of their size class1. 3.1. Radiolarians The class of radiolarians belongs to the kingdom of Protoctists and to the superclass of the Actinopods (or Rhizopods), which they share with the Acantharians (that have a skeleton made of celestite, strontium sulfate, SrSO4 but not opal) and Heliozoans (that possess an external siliceous or chitinous skeleton and live principally in freshwater; only several rare marine species belonging to the genus Stilochonte have been reported). The radiolarians appeared in the beginning of the Paleozoic some 600 million years ago. Their name is related to their radial symmetry, usually marked by the presence of radial spines in many species. The skeleton elements do not appear in the center of the organism, 1 The planktonic organisms are categorized according to their size in pico- (0.7 and 1.33 in particular matter) but also the presence of strongly silicified diatoms (production of Si/C between 0.25 and 0.46), as previously mentioned by Nelson and Smith [NEL 86] in Ross Sea during a bloom of Fragilariopsis curta. The work of Lasbleiz [LAS 14] brought a new perspective to the issue by highlighting the essential role of natural community structures. Iron availability appears as a controlling factor of dominant populations, which at least in the Southern Ocean are characterized by strongly silicified diatoms during bloom periods. Therefore, this indirect role superimposes on the direct effect of the decrease in the Si/C ratio when a given species is transferred from limiting to non-limiting iron conditions. The observations of Hutchins and Bruland [HUT 98] and

Cellular Mechanisms of Silica Deposition by Diatoms

61

Takeda [TAK 98] must then be regarded as experimental artifacts, hardly applicable to natural conditions. 4.5.2. The influence of trace metals on the uptake of orthosilicic acid Rueter and Morel [RUE 81] were the first to mention the interaction between trace metals and silicic acid in the diatom Thalassiosira pseudonana, and they thus hypothesized a zincdependent site whose activity is involved in the uptake process. This involvement of zinc had the advantage of providing an explanation for the similarity of profiles of zinc concentrations and orthosilicic acid concentrations in ocean waters. However, the work of Thamatrakoln and Hildebrand [THA 08] has questioned the intervention of zinc that does not seem to play a direct role, as its sequestration by chelators does not imply a modification of the uptake kinetics of orthosilicic acid. Diatoms of the Southern Ocean have particularly high KSi values compared to those in other regions of the Global Ocean [JAC 83, NEL 92]. This feature was quickly related to the availability of iron in these systems. In natural populations of the Polar Front Zone of the Australian sector, Quéguiner [QUÉ 01] showed that the uptake kinetics vary as a function of iron availability, and KSi values decrease when the iron concentration increases (Figure 4.9). Leynaert et al. [LEY 04] investigated in detail the effect of the availability of iron on the uptake of orthosilicic acid by the diatom Cylindrotheca fusiformis. They concluded to the existence of a colimitation process, which modulates the specific uptake rate VSi depending on the concentration of orthosilicic acid and dissolved iron. Surprisingly, in this species, VSi max and KSi evolve in parallel (VSi max and KSi increase when the iron concentration increases) and affinity remains unchanged. This adaptability is connected by Leynaert et al. [LEY 04] to changes in cell size, which would lead to an optimization of the surface/volume ratio in conditions of limitation by iron.

62

The Biogeochemical Cycle of Silicon in the Ocean

Figure 4.9. Orthosilicic acid uptake kinetics by natural populations of diatoms under different concentrations of dissolved iron in the Polar Front area of the Southern Ocean, Australian sector (53° 48’S, 141° 53’E); a) in the surface layer (20 m, [Fe]