Regionalisierung und Quantifizierung benthischer

Mineralisationsprozesse

Dissertation zur Erlangung

des Doktorgrades in den Naturwissenschaften

im Fachbereich Geowissenschaften

der Universitat Bremen

vorgelegt von

Katherina Seiter

Bremen, Dezember 2003

Tag des Kolloquiums:

27.02.2004

Gutachter:

PD Dr. Matthias Zabel

Prof. Dr. Kai-Uwe Hinrichs

Priifer:

Prof. Dr.Tilo v. Dobeneck

Prof. Dr. Katrin Huhn

Vonvort I

Vorwort

Die vorliegende Arbeit wurde finanziell von der Deutschen Forschungsgemeinschaft im Rahmen des Projektes ZA 19911-1 mit dem Titel "Bilanzierung und Charakterisierung benthischer StoffkreisHiufe anhand regionaler Verteilungsmuster - Bedeutung filr den ozeanischen Stoffhaushalt" gefOrdert. Der Inhalt der vorliegenden Disseliationsschrift beruht auf den Ergebnissen von vier englischsprachigen Manuskripten (Abschnitt 2.1-2.4), von denen drei Manuskripte von mir als Erst-Autorin verfasst wurden. Letztere liegen als Volltexte vor und sind somit als separate Abschnitte, mit unabhangiger Nummerierung der Tabellen, Gleichungen und Abbildungen sowie eigenen Literaturverzeichnissen zu betrachten. Das 4. Manuskript (Abschnitt 2.4) ist durch den entsprechenden Abstrakt in der vorliegenden Dissertationsschrift berucksichtigt. Im Kartenanhang befinden sich jeweils drei TOC- und J pOCu-V elieilungskarten In 0.10 x 0.10-Auflosung (TOC= Konzentration organischen Kohlenstoffs, Jpocu=partikuHirer Fluss organischen Kohlenstoffs) des SW-Afrikanischen Kontinentalhanges, der Arabischen See und des nordlichen Nordatlantiks. Die Kalien stellen erganzende Arbeiten dar, die im Zusammenhang mit den Ergebnissen des 1. und 2. Manuskriptes zu sehen sind. Ebenfalls im Anhang befinden sich drei glob ale Kartendarstellungen (10 x 10), die den Manuskripten entnommen sind und zur besseren Ubersicht vergrossert dargestellt wurden.

II Danksaglll1g

Danksagung

Mein herzlicher und besonderer Dank gilt Henn Priv. Doz. Dr. Matthias Zabel fur die Vergabe und Betreuung der vorliegenden Disseliationsschrift. Die zahlreichen fruchtbaren Diskussionen und sein stetes Interesse an der Entwicklung der vorliegenden Arbeit waren mir eine grof3e Hilfe und UnterstUtzung. Besonders jedoch mochte ich mich fur seine Geduld und die stets freundschaftliche und positive Stimmung bedanken, die wesentlich dazu beigetragen hat den Spaf3 an der Arbeit bis zum Ende nicht zu verlieren.

Ebenso mochte ich mich bei Herrn Professor Dr. Kai Hinrichs herzlich fur die Ubernahme des schnellen Zweitgutachtens bedanken.

Mein Dank gilt ebenfalls Christian Hensen, der als Mitantragsteller des von der deutschen Forschungsgemeinschaft finanzierten Projektes (DFG Projekt ZA 19911-1 ,,Bilanzierung und Charakterisierung benthischer StojJkreislaufe an hand regionaler Verteilungsmuster -Bedeutung fur den ozeanischen Stojjhaushalt") durch viele konstruktive Anregungen und Diskussionen einen grof3en Anteil an der Entstehung der vorliegenden Arbeit tragt.

Dem Europrox-Graduiertenkolleg, insbesondere Herrn Prof. Willems danke ich fur die Aufnahme als assoziertes Mitglied.

Zahlreiche intensive Diskussionen und anregende Gesprache mit Jurgen Schroter haben es mir moglich gemacht dem Wesen der geostatistischen Datenanalyse auf die Spur zu kommen. - vielen herzlichen Dank. In diesem Zusammenhang mochte ich mich ebenfalls bei Heinz Burger bedanken. Angela Schafer danke ich fur ein offenes Ohr bei der Arbeit mit GIS.

Mein herzlicher Dank gilt Martin Kolling der immer mit Rat und Tat zur Stelle war.

Den Mitgliedern der Arbeitsgruppe Geochemie und Hydrogeologie der Universitat Bremen, insbesondere Herrn Professor H.D. Schulz danke ich fur die schnelle Integration, das gute Arbeitsklima und ihre Hilfsbereitschaft. Ich habe mich in der Arbeitsgruppe immer sehr wohl gefuhlt und die vielen konstruktiven Diskussionen und die vielfaltige Unterstutzung ermoglichten mir ein produktives und angenehmes Arbeiten. In diesem Zusammenhang mochte ich mich besonders bei Henrik Hecht, Kay Hamer, Christian Hensen, Volker Km·ius, Sabine Kasten, Martin Kolling, Tanja Lager, Kerstin Pfeifer, Natascha Riedinger und Jurgen Schroter bedanken.

Ein besonderes Dankeschon gilt auch den wissenschaflichen Hilfskraften Gunter Wegener und Bianca Rajes, die mich bei der aufwendigen Literatunecherche und Datenarchivierung unterstUtzt haben.

Fur die freundliche und kollegiale Bereitstellung von Daten und Datenzusammenstellungen mochte ich ganz herzlich Gesine Mollenhauer, Frank WenzhOfer, Dierk Hebbeln, E.A. Romankevich und A.A. Vetrov, Tom Wagner und Tim Jennerjahn danken.

DanksagZll1g III

Der Arbeitsgruppe Pangaea danke ieh fUr die geduldige und stetige Untersmtzung bei der

Datenreeherche. Nal11entlieh l110ehte ieh Miehael Diepenbroeek, Lydia Gerullis, Rainer Sieger

und Verena Meyer-Sturl11borg danken.

Ein ganz besonderes Dankesehon geht an die FNK (Zentrale ForsehungsfOrderung und

Forderung des wissensehaftlichen Nachwuehses), die es l11ir dureh Gewahrung eines

Prol11otionsabsehluss-Stipendiul11s mit ermogliehte, die vorliegende Arbeit zu beenden -

vielen herzliehen Dank.

Meinen Freunden und Kollegen - Fanni, Henner, Henrik, Ina, Kerstin, Mark, Maik, Nataseha,

Tanja, Urte, Uwe und Volker - der "DoID1erstagsrunde" danke ieh fUr die vielen

abwechslungsreiehen, stimmungsvollen und aueh ruhigen Abende, die l11ir viel bedeutet

haben. Kerstin und Nataseha danke ieh besonders fUr die nette zweite Heimat im TAB und

das abweehlungsreiehe Essen, dass immer zur "Not" parat war. Sadat danke ieh fUr die

Kaffeepausen am Wochenende.

Mein ganz personlicher Dank gilt Maja und Bemd, Christiane, Christian, Harald, Henrik,

Henner und Ina, Urte, Andrea, Angela, Johanna, Sonni, Gerd, Marie und Boris, die fUr mieh

da waren, mit mir gelacht und l11ir Mut gemacht haben.

Christan Seiter danke ieh fUr vieles, besonders jedoeh fUr seme Hilfe und Untersmtzung

vieWiltiger Art.

Maja und Bemd und Christiane danke ieh besonders fUr ihre uneingesehrankte Hilfe und ihr

Verstandnis zujeder Zeit.

Henrik moehte ieh dafUr danken, dass er immer fUr l11ieh da war, mir zugehort hat, dafUr

gesorgt hat, dass der Mut mieh aueh in sehwierigen Situationen nieht verlassen hat und es

immer viel zu laehen fUr uns gab.

Meiner Toehter Lea Marie danke ieh von Herzen. Ohne ihre Frohliehkeit, ihr geduldiges

Wesen und so viel Sehones, das ieh dureh sie erfahren durfte und darf, ware vieles nieht so

bunt und interessant.

Ieh danke meinen Grof3eltem, Sophie, Otti und Kurt fUr die schone Zeit mit ihnen, und die

vielfaltige Untersmtzung, die sie mir haben zukommen lassen.

IV KlIrziasslIl1g

Kurzfassung

Ozean und Atmosphare tauschen Kohlenstoff liber verschiedene biologische, physikalische und chemische Prozesse aus, die als marine Kohlenstoffpumpen bezeichnet werden. In das natlirlichen Schwankungen unterliegende Gleichgewicht zwischen CO2-Aufnahme und Abgabe greift der Mensch zunehmend ein. Ein GroBteil des so entstehenden CO2-

Uberschusses wird von den Ozeanen in der euphotischen Zone durch den Einbau von CO2 in die organische Substanz planktonischer Organismen (organischer Substanz) und durch die Bildung von Karbonatschalen absorbiert. Welcher Anteil des so gebundenen Kohlendioxids letztendlich langfristig im Sediment gespeicheli und somit dem globalen Kohlenstoffkreislauf liber geologische Zeitraume entzogen wird, entscheiden die Abbau- und Losungsprozesse wahrend des Transports durch die Wassersaule und im Sediment.

Wahrend der Losungsprozesse in del' Wassersaule, aber auch an der Sediment -Wasser Grenzschicht, werden Nahrstoffe wie Silizium, Phosphor und Stickstoff ruckgelOst und durch die globale Zirkulation der Wassennassen wieder in euphotische Bereiche transpOliiert. Den Sedimenten als Kohlenstoffsenke, aber auch den Prozessen an der Sediment-Wasser Grenzschicht, kommt daher eine besondere Bedeutung zu.

Mit dem Ziel der Identifizierung regional er Verteilungsmuster benthischer Mineralisationsprozesse und del' Quantifizierung des Stofftransportes libel' die Sediment­Wasser Grenzschicht in del' Tiefsee, liegt ein besonderer Schwerpunkt der vorliegenden Arbeit auf der Charakterisierung benthischer Provinzen. Die Grundlage hierflir bildet eine umfassende Datenkompilation aus weltweit verfugbaren Daten zur Konzentration des ol'ganischen Kohlenstoffs (TOC) im Obel'flachensediment. Die Einbeziehung regionaler, wie ozeanographischer, biogeochemischer und sedimentspezifischer Randbedingungen, und eine umfassende geostatistische Analyse, ermoglichte die Einteilung des Weltozeans in 33 benthische TOC-Provinzen und die Erstellung einer globalen TOC-Velieilungskarte (1 ° x 1°).

Del' Parameter TOC ist ein idealer Stellvelireterparameter (Proxi- und Kontrollparameter) fur viele benthische Abbauprozesse, da das geochemische Milieu 111 den oberen Sedimentabschnitten entscheidend durch die Mineralisation organischer Substanz gesteuert wird. Ebenso bietet diesel' Parameter gegenliber der direkten Nutzung biogeochemischer Parameter, wie diffusiven Stofffllissen libel' die Sediment-Wasser Grenzschicht, den V Olieil einer llOhen Datenverfligbarkeit. Um den Anteil des organischen Materials, del' den Meeresboden nach seinem langen TranspOli von del' Wasserobel'flache zum Meeresboden erreicht, zuverlassig abschatzen zu konnen, stellen die Einteilung in benthische Provinzen und die TOC-Verteilungskarte wichtige Grundlagen dar. Somit kann eine Vielzahl regional spezifischer Prozesse berucksichtigt werden, die das organische Material auf seinem Weg aus der euphotischen Zone bis an die Sediment-Wasser Grenzschicht beeinflussen.

Die Entwicklung regional spezifischer Transferfunktionen zur Darstellung des pmiikularen Flusses organischen Kohlenstoffs zum Sediment stiltzt sich auf die Annahme, dass del' Hauptanteil des Abbaus organischen Materials aerob an del' Sedimentoberflache geschieht. Bei hoher Primarproduktion organischen Materials und damit verbundener hoher

KlIrzfasslIl1g v

Sauel'stoffzehrung kann del' Sauel'stoffgehalt im Bodenwasser regional limitierend sem. Ausgehend von den 33 benthischen TOC-Provinzen konnten in 11 charakteristischen Regionen Korrelationen zwischen benthischen diffusiven Sauerstofffliissen, der TOC­Konzentration im Oberflachensediment und dem Sauerstoffgehalt im Bodenwasser festgestellt und empirische Beziehungen hergeleitet werden. Angewandt auf die globale TOC­Velieilungskarte konnte sowohl regional als auch global die Mindestmenge partikularen organischen Materials, die den Meeresboden erreicht haben muss, dal'gestellt und anderen Ansatzen gegeniibel'gestellt werden. Ein besondel'er Schwerpunkt lag hierbei auf vergleichenden Betl'achtungen zwischen bodemlahen PaIiikelfallen und den abgeschatzten Partikelfliissen. Es konnte vor all em gezeigt wel'den, dass regionale laterale Transportprozesse in der Wassersaule oder in Bodennahe entscheidend zu einer Entkopplung zwischen den Verteilungsmustem der Primarproduktion der euphotischen Zone und benthischen Velieilungsmustem des partikularen organischen Materials beitragen.

Die beschriebene Methode wurde flir den Bel'eich des S-Atlantik auf die Abschatzung des Gesamtflusses biogenen Materials zum Sediment ausgeweitet. Einen wesentlichen Bestandteil stellte dabei die Untersuchung der Kopplung zwischen benthischel' Sauerstoffzehrung und Siliziumruckfluss iibel' die Sediment-Wasser Grenzschicht dar. Die in diesem Zusammenhang entstandene Vel'teilungskarte benthischer Siliziumfl'eisetzung ist die Grundlage zur Abschatzung des Flusses biogenen Opals zum Sediment. Durch die zusatzliche Nutzung vol'ab hergeleitetel' Beziehungen zwischen Sauerstoffzehrung und TOC sowie die Einbeziehung bekannter empirischer Beziehungen zwischen KalzitlOsungsl'aten und aerobem Umsatz organischer Substanz, konnte del' Gesamtfluss biogenen Materials zum Sediment abgeschatzt werden. Die Ergebnisse zeigen, dass die ennittelten PaIiikelfliisse herkommliche Akkumulationsl'aten um ein vielfaches iibel'steigen.

VI KlIrz[assung

lnhalt

VOl"Wort

Danksagung

Kurzfassung

1. Einleitung und Fragestellung

VII

Inhalt

1.1 Bedeutung mariner biogeochemischer StoffkreisHiufe fUr den globalen 2

Stoffhaushalt

1.1.1

1.l.2

1.1.3

1.1.4

1.1.5

1.2

1.3

1.3.1

1.3.2

1.4

Der Kohlenstoffkreislauf

Die anorganische und physikalische Kohlenstoffpumpe

Die organische Kohlenstoffpumpe

Benthische Remineralisierungsprozesse

Der Siliziumkreislauf

Erfassung regionaler Verteilungsmuster benthischer

Min eralisa tio nsp rozesse

Nutzung eines Geo-!nformations-§.Ystems

Datenmodelle

Erstellung und Erfassung hochauflosender Rasterdatenmodelle

Kurzfassung der eingereichten Manuskripte (Abschnitte 2.1-2.4)

Literatur

2. Ergebnisse

2.1

2.2

2.3

2.4

Regionalization of the organic carbon content in surface sediments -

defining a new approach ofbenthic TOe-based regional provinces

[K. Seiter, C. Hensen, 1. Schroter, M. Zabel]

The benthic carbon mineralization on a global scale

[K. Seiter, C. Hensen, M. Zabel]

The benthic silica release and its implication for the estimation of the non­

litho genic particle fluxes to the sea floor

[K. Seiter, 1. M. Holstein, C. Hensen, M. Zabel]

Fluxes at the benthic boundary layer - a global view from the S-Atlantic

[C. Hensen, K. Pfeifer, F. WenzhOfer, A. Volbers, S. Schulz, 1. Holstein, O. Romero, K.

Seiter]

2

3

5

9

10

11

13

14

16

18

21

26

26

60

95

116

VIII

3. Zusammenfassung

4. Kartenanhang

lnhalt

118

120

A I+II TOC-Konzentration 1m Oberflachensediment und partikularer Fluss 120

organischen Kohlenstoffs zum Sediment des SW-Afrikanischen

Kontinentalhanges (0.1 ° x 0.1 0)

A IIl+IV TOC-Konzentration im Oberflachensediment und partikularer Fluss 121

organischen Kohlenstoffs zum Sediment im nordlichen N-Atlantik

(0.10 x 0.1°)

A V+VI TOC-Konzentration 1m Oberflachensediment und partikularer Fluss 122

organischen Kohlenstoffs zum Sediment der Arabischen See

B I TOC-Gehalt im Oberflachensediment « 5 cm Sedimenttiefe) 123

B Il Sauerstoffgehalt im Bodenwasser (BOC) 124

B III Minimaler partikularer Fluss organischen Kohlenstoffs zum Sediment 125

(JPOCa)

5. Datenanhang 126

Appendix I 126

Appendix Il 130

Appendix III 132

Appendix IV 134

Einleitung lInd Fragestellzmg 1

1. Einleitung und FragesteIlung

Untersuchungen zum Verstandnis des globalen Kohlenstoffkreislaufs SOWle semer Bilanzierung und zeitlichen Entwicklung gehoren seit lahrzehnten zu den zentralen Themen geowissenschaftlicher F orschung. Von besonderer Bedeutung ist hierbei die Auswirkung des anthropogen bedingten CO2-Gehaltes del' Atmosphare im Hinblick auf eine globale Erwannung (Treibhauseffekt). Wesentliche SteuergroI3en des Kohlenstoffkreislaufes sind verschiedene biologische, physikalische und chemische Austauschprozesse libel' die Grenzflache Atmosphare-Ozean.

Ein GroI3teil des CO2-Uberschusses wird von den Ozeanen durch den Einbau von CO2 in die Biomasse planktonischer Organismen m Fonn von organischer Substanz und Karbonatschalen absorbiert. Die Abbau- und Losungsprozesse wahrend des TranspOlis durch die Wassersaule und im Sediment entscheiden nach dem Absterben del' Organismen, welcher Anteil des karbonatisch odeI' organisch gebundenen Kohlenstoffs in den kurzfristigen ozeanischen Kohlenstoffkreislauf zurlickgefiihrt odeI' langfristig im Sediment gespeichert wird. Den Sedimenten kommt daher als Kohlenstoffsenke eine besondere Bedeutung zu. Eng an den kurzfristigen Kohlenstoffkreislauf gekoppelt sind Nahrstoffe wie Silizium, Stickstoff und Phosphor, die u.a. von Phytoplankton in del' euphotischen Zone zum Aufbau organischer Substanz und mineralischer Skelettstrukturen benotigt werden. Wahrend del' Losungsprozesse in del' Wassersaule aber auch an del' Sediment-Wasser Grenzschicht werden die Nahrstoffe rlickgelOst und durch die glob ale Zirkulation del' Wasselmassen wieder in den euphotischen Bereich transportiert. Langfristige Schwankungen im COrGehalt del' Atmosphare beeinflussen daher nicht nur oberflachennahe Prozesse, Wle Anderungen in del' Organismenvergesellschaftung, sondern auch die Prozesse an der Sediment-Wasser Grenzschicht und das Speichervermogen del' Sedimente.

Um den globalen Kohlenstoffkreislauf und letztendlich die mittel- und langfristige Klimaentwicklung verstehen zu komlen, mlissen die angekoppelten biogeochemischen Nahrstoffkreislaufe untersucht werden. Von besonderem Interesse sind hierbei globale und regionale Untersuchungen zur Kopplung zwischen Primarproduktion, Partikelfllissen organischen Kohlenstoffs zum Sediment und benthischen Stofffllissen sowie der Einbettung organischen Materials.

Das Ziel del' vorliegenden Arbeit (DFG Projekt ZA 19911-1) bestand in der Erfassung und Quantifizierung regionaler Verteilungsmuster benthischer Mineralisationsprozesse, des Stofftransportes libel' die Sediment-Wasser Grenzschicht in del' Tiefsee, sowie ihrer Bilanzierung hinsichtlich globaler Stoffkreislaufe. Ausgehend von Ergebnissen aus Einzelstudien und regional begrenzter Analysen sollten die vorliegenden Daten zusammengefiihli werden, um groI3raumige und globale Velieilungsmuster benthischer Stofffllisse zu erfassen. Ein besonderer Schwerpunkt del' Arbeit lag hierbei - in Analogie zu bestehenden Ansatzen zur Abschatzung del' globalen Primarproduktion - auf der Charakterisierung benthisch-biogeochemischer Provinzen. Da es ein wesentliches Ziel del' Arbeit war, Korrelationen zwischen benthischen Stofffllissen und anderen haufig gemessenen und gut verfiigbaren Kontrollparametem zu nutzen und die Ubertragung del' Ergebnisse auf

2 Einleitlll1g lInd Fragestel/zll1g

weniger gut untersuchte Gebiete zu ermoglichen, stellt die vorliegende Arbeit eine inhaltliche

Fortsetzung der Arbeiten von Zabel (1994) und Hensen (1996) dar. Diese Arbeiten entstanden

im Rahmen des deutschen Beitrages zum lGOFS Programm (.Joint Global Ocean £lux (itudies) und in enger Zusammenarbeit mit den Untersuchungen im Rahmen des

Sonderforschungsbereiches 261 der Universitat Bremen ("Der Siidatlantik im Spatquartar: Rekonstruktiol1 von Stoffhaushalt und Stromsystemen").

Im folgenden solI ein Uberblick der fUr die Arbeit relevanten Prozesse gegeben werden. Aus

der Vielschichtigkeit del' zueinander in Bezug stehenden Teilbereiche ergibt sich, dass eine

so1che Zusammenfassung nur ansatzweise die Komplexitat del' Einzelsysteme wiedergeben

kam1. Dennoch wird versucht den jeweiligen gegenwartigen Kenntnisstand darzulegen.

1.1 Bedeutung mariner biogeochemischer StoffkreisHiufe fUr den globalen Stoffhaushalt

1.1.1 Del' KohlenstojJkreislm(f

Dem marinen Kohlenstoffkreislauf wird in den letzen lahrzehnten eine zunehmend wichtige

Schltisselfunktion im Verstandnis des Klimageschehens auf der Erde zugeschrieben.

Untersuchungen zu langfristigen und kurzfristigen Klimaschwankungen in bezug auf die

Variabilitaten im CO2-Haushalt del' Atmosphare sind hierbei von besonderem Interesse. Die

Atmosphare und die feste Erde stehen in standi gem Gasaustausch, der durch verschiedene

komplexe physikalische, aber ebenso biochemische Prozesse gesteueli wird. Der

Stoffkreislauf, der die COrFreisetzung an die Atmosphare und die kurzfristige, aber auch

langfristige Speicherung von Kohlenstoff in den Sedimenten der Ozeane, del' terrestrischen

Biosphare und der Lithosphare beschreibt, wird als Kohlenstoffkreislauf bezeichnet. Hierbei

stellt die Lithosphare die groBte Kohlenstoffsenke dar. 600 Millionen Gt (Giga Tom1en)

anorganischen Kohlenstoffs sind in den Karbonatgesteinen und 15 Millionen Gt organischen

Kohlenstoffs sind in Fonn von ErdOl, Erdgas und Kohle gespeicheli (Abb. 1). Im Gegensatz

zur Atmosphare, in der derzeit ca. 750 Gt Kohlenstoff enthalten sind, sowie zur terrestrischen

Biomasse mit 2300 Gt Kohlenstoff, ist die Speicherkapazitat der Ozeane sehr groB

(Siegenthaler & Sarmiento, 1993, Sanniento & Grub er, 2002). Sie betragt schatzungsweise

38150 Gt Kohlenstoff. Davon sind ca. 900 Gt Kohlenstoffim Oberflachenwasser und 150 Gt

Kohlenstoff im Oberflachensediment gespeichert (Siegenthaler & Sanniento, 1993;

Sanniento & Gruber, 2002).

In das narurlichen Schwankungen unterliegende Gleichgewicht zwischen COrFreisetzung

und F estlegung greift der Mensch mit zunehmender Industrialisierung seit den letzten 100

lahren durch die Nutzung fossiler Brennstoffe, Holzrodung und Zementproduktion vennehrt

ein. Nach Schatzungen von Sanniento & Gruber (2002) betragt die derzeitige anthropogen

verursachte CO2-Emission ca. 7.1 GtCllahr (5.4 GtCIJahr + 1.7 GtCllahr, Abb. 1), wovon nur

3.3 GtCllahr in der Atmosphare verbleiben. Neueste Untersuchungen zeigen, dass dieser

"missing link" von ca. 3.8 GtCIJahr gleichennaBen auf die Aufnahme durch die Ozeane und das terrestrische Okosystem zUrUckzufUhren ist. Die Auswirkungen der vennehrten CO2-

Aufnahme del' Atmosphare sind schon heute erkennbar. Eine del' nachhaltigsten Folgen ist die

Einleitlll1g lInd Fragestellllng 3

globale Zunahme del' Oberflachentemperatur seit 1861 um ~O.6°C wie das Intergovernmental

fane1 on Climate Change (IPCC, 2001; http://www.ipcc.ch) feststellen konnte.

T 59.6

c .2 2 'r: '/; CJ

Co<: T

60 /.9 1. 590

~ cD

" " EO ;;;l

f1. '" '" .5 en " Sl)

c 2 " i;, § 0 2 '5 " E. ~ -§ ,E .3 " <

Lithosphiire: 15.000000 C,,,,, 600000000 Cwm•

-+- 161 '3.3 20 21.9 SA 70.6 70

Abbildung 1: Schematische Darstellung des globalen Kohlenstofflcreislaufs (modifiziert nach Ittekkot et aL, 2002 und Sarmiento & Gruber, 2002). Die Pfeile markieren die Kohlenstofffllisse in GtCIJahr (kursiv) zwischen der Atmosphare, Land und Ozean. Das Kohlenstoff-Reservoir ist in GtC angegeben. Anthropogen verursachte Anderungen sind rot markiert. Die jahrliche Zuwachsrate ist griin markiert.

1.1.2 Die anorganische und physikalische Kohlenstojjjmmpe

Del' Austausch von CO2 zwischen Atmosphare und Ozean wird durch den Unterschied in del'

Konzentration, bzw. der Partialdrucke des Kohlendioxids gesteuert. Die Bildung, Festlegung

oder Loslichkeit des Kohlendioxids wird durch verschiedene ineinander greifende Prozesse -die als Kohlenstofipumpen bezeichnet werden - reguliert. Physikalische GroJ3en wie

Salzgehalt, Druck und Temperatur regulieren den CO2-Partialdruck (pC02) und damit den

Gasaustausch mit del' Atmosphare. Dies ist vor allem in den Polarregionen von groJ3er Bedeutung, da bei sinkenden Wassertemperaturen zunehmend CO2 gebunden werden kann.

Vorwiegend im N-Atlantik und im Weddelmeer sinkt kaltes salzreiches Wasser in groJ3e

Tiefen ab und wird zu Tiefen- und Bodenwasser, dass reich an gelOsten Gasen wie 02 und

Kohlendioxid ist. Die gelOsten Gase verbleiben so fur langere Zeit in tieferen Wassermassen und werden durch die thennohaline Tiefenzirkulation (engl. conveyer belt) in den Ozean

Becken verteilt. Wann genau das im Wasser gelOste C02 emeut in Kontakt mit del' Atmosphare gerat, hangt von del' Durchmischung del' Wassennassen ab. Die Verweildauer

einzelner Gas-Molekiile im Tiefenwasser betragt im Durchschnitt ca. 1000 Jahre, d.h. erst

nach Ablauf diesel' Zeitspanne gerat das Tiefenwasser wieder in den Kontakt mit del'

4 Ein/eitlll1g 1ll1d FragestellzlI1g

Atmosphare. Ein anthropogen verursachter CO2-Anstieg wird daher, bis auf wemge Ausnahmen im N-Atlantik, derzeit noch nicht in Wasserschichten unterhalb 2000 m

beobachtet (Broecker & Peng, 1983; Sanniento & Gruber, 2002).

Andern sich die physikalischen Randbedingungen wird das Gleichgewicht des CO2-Austausches zwischen Ozean und Atmosphare gestOrt. So weisen Wannzeiten hohe und

Kaltzeiten niedrige C02-Konzentrationen del' Atmosphare auf. Uber den PolalTegionen bewirkt eine einsetzende Erwarmung del' oberflachennahen Luftschichten eine verstarkte CO2-Aufnahme durch die kalten Wassermassen. Eine zunehmende Erwarmung hatte jedoch

die Reduzienmg del' Tiefenwasserbildung zufolge und wurde auch nachhaltig die globale

Zirkulation del' Wassennassen und das Klimageschehen beeinflussen (z.B. lttekkot et aI., 2002). Die Ursachen fUr die Kopplung sind bislang nicht vollstandig geklali. Sicher ist, dass

eine enge Rilckkopplung zwischen Ozean und Atmosphare besteht und del' Ozean nicht nur

Wannespeicher, sondern auch Kohlenstoffsenke und Quelle zugleich ist.

Eng an die physikalische Pumpe ist die anorganische Kohlenstoffpumpe gekoppelt, da del'

C02-Austausch mit del' Atmosphare auch durch die Bildung kalkschaliger Organismen (z.B.

Coccolithophoriden) in del' euphotischen Zone des Ozeans beeinflusst wird. Die Speziesverteilung del' Karbonat-Ionen in del' Wassersaule ist von del' Lage des

Gleichgewichts nn Kalk-Kohlensaure-System abhangig. Eine Verschiebung des

Gleichgewichts kann zu Losungs- oder Fallungsprozessen fUhren (Gl. 1).

(1)

Sinken abgestorbene kalkschalige Organismen durch die Wassersaule zum Meeresboden, 16sen sie sich in Abhangigkeit des Sattigungszustandes des Wassers in bezug auf die

Konzentration del' Karbonationen auf. Del' Sattigungszustand wird maf3geblich durch den Druck, die Temperatur und die Vennischung von Wassermassen unterschiedlicher Herkunft

und CO2-Gehalte gesteuert. Unterhalb del' hydrographischen Lysokline wird das Loslichkeitsprodukt von Kalzit unterschritten und es kommt zur Karbonat16sung. In den

oberen lichtdurchfluteten Wasserschichten wirken anorganische und organische

Kohlenstoffpumpe entgegengesetzt (Karbonat-"Gegenpumpe"), da die Bildung kalkiger

Skelettstrukturen durch das Phytoplankton die Freisetzung von CO2 in die Atmosphare

begilnstigt und del' notwendige Aufbau organischer Substanz del' Organismen C02 bindet. In tieferen Wasserschichten wendet sich das Bild und die CO2-Freisetzung wahrend des

mikrobiellen Abbaus organischer Substanz fUhrt zur Erhohung del' Alkalitat und begilnstigt

somit die Kalk16sung. Die Mineralisation organischer Substanz und die Kalzit16sung am Meeresboden sind somit eng aneinander gekoppelt.

Entscheidend fUr die Effizienz del' Kohlenstoffaufnahme durch den Ozean (biologische

Pumpe) ist das Verhaltnis zwischen del' Produktion organischen Kohlenstoffs und biogenem Karbonat. Sinkt die Karbonatproduktion relativ zur Gesamtproduktion, etwa durch sinkende

Wasseliemperaturen oder durch die Zunahme silikatschaliger Organismen, wird die Effizienz

del' organischen Kohlenstoffpumpe verstarkt, die Karbonat-"Gegenpumpe" jedoch geschwacht. Del' Ozean kann mehr CO2 aus del' Atmosphare aufnehmen und langfristig

Einleitung und Fragestellung 5

speichern (z.B. Ittekkot et aI., 2002; Sigman & Boyle, 2000). Wie sich die anthropogen

verursachten derzeit steigenden Temperaturen der Atmosphare auf dieses komplexe

Zusammenspiel zwischen biologischer Kohlenstoffpumpe und Atmosphare langfristig

auswirken werden ist bislang ungeklart (z.B. Ittekkot et aI., 2002).

1.1. 3 Die organische KohlenstoiJpumpe

In der euphotischen Zone produzieren Algen mit Hilfe des Sonnenlichts organischen

Kohlenstoff durch C02-Aufnahme. Eine entscheidende Rolle spielen hierbei die Chlorophyll­

a-Moleklile, die Sonnenenergie in chemisch gebundene Energie transformieren konnen. Bei

dieser Reaktion wird Sauerstoff an die Atmosphare abgegeben und nachfolgend bei der

Respiration durch die Konsumenten, wie Pflanzen, Menschen und Tiere verbraucht. Die

biologische Aktivitat in der euphotischen Zone ist aufgrund ihrer kurzen Reaktionszeit

maBgebend fUr die Pufferwirkung der Ozeane in bezug auf steigende CO2-Konzentrationen in

der Atmosphare. Wie im vorhergehenden Abschnitt bereits erlautert, beeinflussen u.a.

Produktion und Mineralisierung organischer Substanz den Kohlenstoffkreislauf in den

lichtdurchfluteten Schichten der oberen Wassersaule, aber auch an der Sediment-Wasser

Grenzschicht.

-180 -120 I

-60 o 120 I

180

60

-0

--60

PIT [g/(m2 yr)]

0-50 50 - 100 100 - 150 150 - 200200 - 250 250 - 300 300 - 350 350 - 400 400 - 450 >400

Abbildung 2 (a): Primarproduktion (PPT) nach Antoine et al. (1996).

6 Einleitung und Fragestellung

-180 -120 -60 o 60 120 180 I

<0.25 0.5 - 0.75 1 - 1.25 1.5 - 1.75 2 - 2.25 2.5 - 2.75 3 - 3.5 4 - 4.5 5-6 8 -12 >20

Abbildung 2 (b): Vertikal eXpOliiertes Material zum Meeresboden (Jpoc), berechnet nach Antia et al. (2001), nach Jpoc =OlxPpl.77z-0.68; z: Wassertiefe (ETOP05). FUr die Berechnung wurde die

Primarproduktion nach Antoine et al. (1996) eingesetzt, verfiigbar als Rasterdatensatz. Die Wassertiefe geht als ETOP05 Raster ein (siehe Abschnitt 1.3.2).

Nach Antoine et al. (1996) werden ca. 36-45.6 GtCIJahr durch Phytoplankton im OberfHichenwasser produziert (Abb. 2a). Andere Schatzungen gehen von 40-50 GtCIJahr aus (Behrenfeld & Falkowski, 1997-a,b). Beide genannten Modelle beruhen auf flachendeckenden Langzeitbeo bachtungen der F arbverteilungsmuster im chlorophy 11-spezifischen WellenHingebereich (400-700 nm) der Meeresoberflache, die zwischen 1978-1986 mit Hilfe des CZCS (engl. Coastal Zone Calor §.canner) an Bard des Nimbus7-Sate11iten erhoben wurden. Die Modelle von Antoine et al. (1996) und Behrenfeld &

Falkowski (1997-a,b) berilcksichtigen in den Algorithmen zur Abschatzung der gesamten Menge an organisch gebundenem Kohlenstoff (PPT) aus der Phytoplanktonmasse die ilber die Tiefe integrierte euphotische Zone. Die wesentlichen Unterschiede beider Mode11e sind in den hOheren Abschatzungen nach Behrenfeld & Falkowski (1997 -a,b) zwischen 30° und 80° nordlicher Breite und 30° und 50° sildlicher Breite zu finden. Die Ursache unterschiedlich geschatzter lokaler Primarproduktionsraten liegt hauptsachlich in der eingesetzten photoadaptiven Variablen (P opD, welche ein MaB fUr die tagliche maximale Produktion bei Lichtsattigung ist (Behrenfeld & Falkowski, 1997-a).

Der groBte Teil des primar produzierten organischen Materials wird im kurzgeschlossenen Kohlenstoffkreislauf der lichtdurchfluteten Zone unter Bildung von ge16stem organischem Kohlenstoff (DOC) und C02 mineralisiert und steht den Organismen als regeneriertes organisches Material (PR) erneut zur VerfUgung. Ein Anteil von ca. 30 % in Schelfregionen

EinleitzlI1g 1I11d Fragestellzll1g 7

und 10-15 % im offenen Ozean wird in Form abgestorbener Organismen als Aggregate und

Kotpillen (engl. fecal pellets) exportiert (PE). Das partikuHire organische Material (POM)

unterliegt auf seinem Weg durch die Wassersaule der Mineralisation. Ein geringer Antei1 des

partikularen organischen Kohlenstoffs (POC) von 4-17 % der Primarproduktion entlang der

Kontinentalhange und ca. 1.5 % im Bereich der Tiefsee erreicht den Meeresboden (Lampitt &

Antia, 1997; Wollast, 1998; Schlitzer, 2002; WenzhOfer & Glud, 2002) (Abb. 2b; Abb. 3).

Nach Wollast (1998) wird nur ein kleiner Prozentsatz von 0.5-3 % entlang der

Kontinentalhange und weniger als 0.01 % in der Tiefsee dem Kohlenstoffkreislauf langfristig

entzogen und in den Sedimenten und Gesteinen liber geologische Zeitraume gespeicheli.

Mehr als 80 % des gesamten in den Weltozeanen exportielien organischen Materials wird auf

nur 20 % ihrer Gesamtflache in den Schelfregionen und am oberen Kontinentalhang

akkumuliert (Wollast, 1998). Den Kontinentalrandern kommt daher bei Betrachtungen der

COrBilanz eine bedeutende Rolle zu.

Eine Vielzahl empirischer Gleichungen beschreibt die exponentielle POC-Abnahme in

Abhangigkeit der Wassertiefe auf der Basis der Primarproduktion (Suess, 1980; Betzer et aI.,

1984; Maliin et aI., 1987; Pace et aI., 1987; Berger et aI., 1989; Antia et aI., 2001). Diese

Ansiitze setzen einen veliikalen Transport von Kohlenstoff durch die Wassersaule voraus. In

marinen Systemen spielen jedoch nicht nur vertikale, sondern auch laterale Transportprozesse

eine wichtige Rolle (Abb. 3). Nach Wollast (1998) werden so ca. 2 GtCIJahr aus den

Schelfgebieten in die Tiefsee transpOliieli. In Suspension gehaltener organischer Kohlenstoff

in der oft mehrere hundert Meter machtigen sogenannten ,,Nepheloid-Layer" wird durch

intensive Wasserzirkulation liber den Schelf zum oberen Kontinentalhang verlageli (z.B.

Freudenthal et aI., 2001). Ein weiterer Anteil des organischen Materials wird vor all em auf

dem Schelf und am oberen Kontinentalhang durch Bodenstromungen lateral verdriftet und

erneut abgelageli. Aus Sedimentfallenuntersuchungen ist bekannt, dass ca. 10 % der

klistennahen Produktion durch hangabwarts gerichtete, sogenannte ,,Downs lope" -Prozesse

verfrachtet wird (Garzoli, 1993; Rowe et aI., 1994; Peterson et aI., 1996; Wollast, 1998,

Arthur et aI., 1998; Hensen et aI., 2000; Giraudeau et aI., 2002; Mollenhauer et aI., 2002;

Hensen et aI., 2003). Der Eintrag durch Fllisse betragt nach Schllinz & Schneider (2000) ca.

0.4 GtCIJahr und wird liberwiegend auf dem Schelf abgelagert. Welcher Anteil hiervon durch

die genannten Prozesse liber die Schelfkante in tiefere Wasserbereiche gelangt, unterliegt

gegenwartiger Diskussion (Schlesinger and Melack, 1981; Ittekkot, 1988; Hedges, 1992;

Meybeck, 1993; Ludwig et aI., 1996; Wollast, 1998).

8

0.3 Gte!y!

40-50 GtClyr

Oberflachel1str6nHl.l1,g

B ellthi;c;che Rlic1doSUJlg -+-

Einleitllng lInd Fragestellllng

,.?'" ~"<'~'" ~ __ .,.d;' "

COo.

. 0.4 Gtc'!y!

Primar ~r6auktion (PPT);/ y/,

~~""'hW""~"'M''''''"'A''WN'''"N~~'''V'''''l

: i FluJ3eilrtl'ag I .. Regene:tierte Procluktioll -/ p /,/' R.

-+-

Abbildung 3: Schematische Darstellung des Flusses organischen Kohlenstoffs zum Sediment sowie Faktoren, die die TOC-Konzentration im OberfHichensediment beeinflussen (Romankevich, 1984; Behrenfeld & Falkowski, 1997-a,b; Wollast, 1998; SchlUnz & Schneider, 2000; Rullkotter, 2000).

Die genannten V Ol'gange machen deutlich, dass regional spezifische Prozesse stattfinden, welche die Einbettung organischen Kohlenstoffs in das Sediment und die TOC-Konzentration im Oberflachensediment maBgeblich mit beeinflussen kannen (Abb. 3). Diese regionalen Variationen betreffen jedoch nicht nur die vertikalen und lateralen Eintragspfade, sondem auch physikalisch gesteuelie TranspOliprozesse und chemische Mineralisations- und Lasungsprozesse. So hangt die TranspOligeschwindigkeit eines Partikels oder Aggregats von seinem spezifischen Gewicht ab. Je haher dieses ist, desto schneller erfolgt der Transport durch die Wassersaule und mehr labile organische Substanz elTeicht den Meeresboden (z.B. Armstrong et aI., 2002; Klaas & Archer, in press). Dies wird durch sogenmmte Ballastminerale wie Kalzit und lithogenes Material gesteueli, die mit der organischen Matrix assoziiert sind. Besonders in kiistennahen Gebieten, in denen hohe Sedimentationsraten infolge hoher Massenflusse biogenen und siliziklastischen Materials auftreten, wird die

Einleitllng 1I11d Frageste1111l1g 9

Einbettungseffizienz organischen Materials erh6ht. Umgekehli bewirkt die Abnahme verdunnender nicht-reaktiver Antei1e am Partike1fluss eine Abnahme der Einbettungsraten (Henrichs, 1992; lahnke, 1996; Zabe1 et aI., 2000). le Hinger ein Pariike11abi1en organischen Materials mit Oxidantien wie z.B. Sauerstoff in Kontakt b1eibt, desto gr6Ber ist der Anteil, der abgebaut werden kann (Reimers, 1989; Harinett et aI., 1998). Der TOC-Gehalt im Sediment ist in erster Linie abhangig von der Gesamtsedimentation und ihrem Antei1 an organischer Substanz sowie der Ruck16sung der unterschiedlichen Komponenten durch fruhdiagenetische Prozesse.

1.1.4 Benthische Remineralisierungsprozesse

Die Bedeutung der Sedimente fUr den Kohlenstoffl<Teis1auf liegt darin, dass eingetragenes organisches Material 1angfristig dem Kreislauf entzogen werden kann. Die Freisetzung der Hauptnahrstoffe Nitrat, Phosphat und Silizium wahrend der Fruhdiagenese ist direkt oder indirekt an den Abbau organischen Materials gekoppelt, so dass die Betrachtung der benthischen Stoffflusse uber die Sediment-Wasser Grenze Rucksch1usse auf die Menge des eingetragenen organischen Materials zulasst.

MaBgeblich fUr die Menge organischen Materials, das den Meeresboden erreicht und die Menge, die schlie13lich eingebettet wird, ist daher auch seine Qualitat, d.h. seine moleku1are Zusammensetzung und die Menge verfUgbarer Oxidantien in Bodenwasser, Porenwasser und Sediment. Das eingetragene organische Material stellt fUr die benthische Lebensgemeinschaft den wesentlichen Energie- und Nahrstofflieferanten dar. Die Reihenfolge der zum Abbau der Organik verwendeten Oxidantien und die damit verbundene Redoxzonierung richtet sich daher nach dem jeweiligen Energiegewinn fUr die beteiligten Mikroorganismen. Sauerstoff dient generell a1s bevorzugter termina1er E1ektronenakzeptor, danach fo1gen Nitrat, Mangan­und Eisenoxyhydrate, sowie Su1fat und der Umsatz organischer Substanz durch Fermentation, bzw. CO2-Reduktion (Froelich et aI., 1979). Nach Canfie1d (1993) zeigt sich jedoch, dass die Obergange zwischen den Zonen flie13end sind und es zu Oberlagerungen kommt. Fur die meisten ozeanographischen Regionen gilt, dass der groBte Tei1 der organischen Substanz durch Sauerstoff abgebaut wird (z.B. Bender & Heggie, 1984; Henrichs & Reeburgh, 1987; Canfie1d, 1989, Canfie1d, 1993). In Tiefseesedimenten wird mehr a1s 90 % des eingetragenen Materials aerob, d.h. in der oxischen Zone abgebaut, da aufgrund oligotropher Verha1tnisse und der 1angen Transportwege durch die Wassersau1e nur wenig reaktives organisches Material den Meeresboden erreicht (Canfie1d, 1989; Midde1burg et al., 1993). Nach der Grundg1eichung von Froelich et al. (1979) wird eine organische Mode1substanz mit dem molaren C:N:P Verhaltnis 106:16:1 (Redfie1d, 1958) mit 138 Mol O2

umgesetzt und ist mit der Freisetzung von CO2, N03 und P043

- assoziiert (Gl. 2).

6G 0 = -3190 kJ mar!, 6G 0 = gewannene freie Energie

Obwoh1 Gleichung 2 im Sinne von Redfie1d (1958) die am weitesten verbreitete Grundg1eichung darstellt, weisen andere Untersuchungen an Phytop1ankton-Gemeinschaften auf regional abweichende mo1eku1are Zusammensetzungen hin. Daraus ergeben sich andere

10 Einleitllng lInd Fragestellzmg

Verhaltnisse von Corg:02 beim Umsatz des organischen Materials (Takahashi et aI., 1985;

Anderson and Sarmiento, 1994).

Die Konzentration gelOsten Sauerstoffs in der Wassersaule und im Porenwasser hangt von

den Wechselwirkungen zwischen Atmosphare und Wasseroberflache, von der

Tiefenzirkulation, von der Sauerstoffproduktion des Phytoplanktons, von der

Sauerstoffzehrung durch den mikrobiellen Abbau des organischen Materials sowie den

physikalischen Eigenschaften des Sedimentes ab. In Hochproduktionsgebieten flihrt die

erhahte Sauerstoffzehrung unterhalb del' euphotischen Zone haufig zur Ausbildung von

Sauerstoffminimumzonen, die bis in 1000 m Wassertiefe beobachtet werden. Mit

zunehmender Akkumulation reaktiven organischen Materials In Schelf- und

Kontinentalhangsedimenten steigt daher der Anteil sub- und anoxischer Diagenese

(J0rgensen, 1982; Cai & Reimers, 1995).

Der diffusive StofftranspOli von Sauerstoff in das Sediment nach dem 1. Fick'schen Gesetz

stellt eine Vereinfachung des zu erwalienden Sauerstoffflusses tiber die Sediment-Wasser

Grenzflache dar. Bioirrigation in ktistennahen Gebieten flihli zu einem aktiven Transport von

Sauerstoff durch (makro-) benthische Organismen in das Sediment und kann eine deutliche

Steigerung des Gesamt-Sauerstoffflusses gegentiber dem diffusiven Fluss bedeuten.

Bioturbation flihrt zu einer Erhahung del' Durchmischung des Sedimentes und daher zu einer

langeren Kontaktzeit der Partikel mit Sauerstoff.

1.1.5 Der SiliziumkreislauJ

Der Siliziumkreislauf ist eng an den Kohlenstofflueislauf gebunden. Derzeit stellen

Coccolithophoriden und Diatomeen die wichtigsten Gruppen innerhalb der

Primarproduzenten dar. Diatomeen bilden 35-75 % der Phytoplanktonspezies (Nelson et aI.,

1995). Sie bauen ihre Skelettsubstanz aus Silizium auf, so dass dieser Nahrstoff ein

limitierender Faktor der Primarproduktion ist. Silizum wird in Form von biogenem Opal oder

als lithogenes Silikat (Feldspate, Hornblende, etc.) durch Fltisse oder Erosion eingetragen.

Die glob ale jahrliche Produktion betragt nach Nelson et al. (1995) und Treguer et al. (1995)

etwa 240 Tmol Silizium. In gelOster Form liegt es als undissoziierte Kieselsaure vor. Das

Bodenwasser ist gewahnlich an biogenem Silikat untersattigt. Im Sediment diffundiert das

gelOste Silizium aus dem Porenwasser in das tiberlagernde Bodenwasser. Die Fltisse von

Sauerstoff und Silizium verlaufen demnach in entgegengesetzter Richtung. Die Faktoren, die

die Lasung biogenen Opals beeinflussen, sind nicht nur chemisch-pysikalische Parameter wie

Druck, Temperatur, Untersattigung des Bodenwassers und die spezifische Oberflache der

Skelettteile. So kannen sich um Opal-Aggregate schtitzende organische Htillen ausbilden

(engl. organic coatings) oder die Skelettteile in Kotktigelchen (engl. Jecal pellets) eingeschlossen werden, die den direkten Kontakt mit dem umgebenden untersattigten Wasser

verringern (Treguer et aI., 1995; Bidle & Azam, 1999; Smetaczek, 1999; Ragueneau et aI.,

2000; Bidle et aI., 2003). Nach Bidle & Azam (1999) und Bidle et al. (2003) bewirken

enzymatisch katalysierte mikrobielle Abbaureaktionen an den coatings die beschleunigte

Lasung biogenen Opals. Die Lasungskinetik wird jedoch ebenfalls durch in das Skelett

inkorporierte Schwermetallionen oder litho gene Alumo-Eisen Verbindungen beeinflusst (z.B.

EinleitzlI1g zl11d Fragestelllll1g 11

van Cappellen & Qui, 1997, van Bennekom, 1988). Die Vielfalt der Einflussfaktoren auf die

Silikatriick16sung ist eine entscheidende Ursache fUr die schlechte allgemeine KOlTelation

zwischen der Konzentration von Kieselsaure im Porenwasser und dem Opal-Gehalt im

Sediment (van Cappellen & Qui, 1997).

Die Bedeutung des Siliziumkreislaufes und seiner Kopplung an den Kohlenstoffkreislauf liegt

in der Freisetzung und Bereitstellung von Nahrstoffen. Nach Treguer et aI. (1990) und Nelson

et aI. (1995) finden 90 % der Opal relevanten Losungsvorgange in der Wassersaule statt und

10 % im Oberflachensediment. Die Opalerhaltung im Sediment betragt nach Angaben von

Nelson et aI. (1995) und Treguer et aI. (1995) im Mittel 3 % und libersteigt damit die

Einbettung organischen Kohlenstoffs. Der hohe Anteil benthischer Rlick16sung ins

Bodenwasser verdeutlicht die Bedeutung benthischer Mineralisationsprozesse und der damit

verbundenen Stofffllisse liber die Sediment-Wasser Grenze (Zabel et aI., 1998; Hensen et aI.,

1998; Hensen et aI., 2000). Da sowohl Opalproduktion als auch Kieselsaure Konzentration

des Bodenwassers regionalen Schwankungen unterliegen, ist auch die Freisetzung in das

Bodenwasser variabeI. Dies wird vor all em in stark eisenlimitierten sogenannten HNLC­

Regionen (engI. High Nutrient L.ow Chlorophyll) deutlich, in den en Diatomeen hohe

Silizium-Kohlenstoff bzw. Silizum-Stickstoff-Verhaltnisse aufweisen (Timmennans et aI.,

1994). Nach Dugdale et aI. (1995), Dugdale & Wilkerson (1998) und Brzezinski et aI. (2003)

steigert sich so im Verhaltnis zur organischen Kohlenstoff Pumpe die Effizienz der Silizium

Pumpe.

1.2 Erfassung regionaler Verteilungsmuster benthischer Mineralisationsprozesse

Die Erfassung regionaler Verteilungsmuster benthischer Stofffllisse und des Flusses

partikularen organischen Materials durch Interpolation der Messdaten ist haufig durch die

verfUgbare Datenbasis sowie deren regional inhomogene Verteilung erschweli (Zabel, 1994).

Nahrstoffflusskmien des ostlichen S-Atlantik von Zabel et aI. (1998) und des gesamten S­

Atlantik von Hensen et aI. (1998) zeigen daher regionale Abweichungen der Schatzwerte von

den Messwerten, die auf engraumige Variationen und die gewahlte Auflosung des

Datemasters zuriickzufUhren sind. Die Ergebnisse zeigen ebenfalls, dass die prinzipiellen

Velieilungsmuster von Primarproduktion und benthischen Stofffllissen korrelieren, jedoch

auch hier regionale Abweichungen auftreten (Zabel et aI., 1998; Hensen et aI., 1998). Globale

Abschatzungen zum Partikelfluss organischen Materials (Jpoe) zum Sediment auf der Basis

von Korrelationen zwischen benthischer Sauerstoffzehmng und Primarproduktion (z.B.

Christensen 2000; Wenzhofer & Glud, 2002), sowie empirische Abschatzungen des Jpoc

durch Primarproduktion und Wassertiefe werden daher kontrovers diskutieli (Hensen et aI.,

1998; Zabel et aI., 1998; Hensen et aI., 2000; Zabel et aI., 2000; Schlliter et aI., 2000). Aus

einer Vielzahl von Ergebnissen aus Sedimentfallen wird deutlich, dass sowohl Prozesse der

oberen Wassersaule, als auch bodennahe Prozesse, wie der advektive Transport suspendielien

organischen Materials durch bodennahe Nepheloid Layer, zu einer Entkopplung der Prozesse

der vertikalen Wasserschichten verschiedener Wassennassen fUhren (Jahnke et aI., 1990;

Wefer & Fischer, 1993; Pefia et aI., 1996; Lampitt & Antia, 1997; Neuer et aI., 1997;

Freudenthal et aI., 2001; Kawahata, 2002).

12 Einleitll71g und Fragestellllng

Da sehr viele fruhdiagenetische Mineralisationsprozesse durch den Umsatz labiler organischer

Substanz mit Sauerstoff in den obersten Sedimentabschnitten beeinflusst werden, liegt es

nahe, zur Abschatzung benthischer Stoffflusse den flir Oberflachensedimente in hoher

Datendichte verfligbaren Parameter "organischer Kohlenstoff' (TOe) als Proxy- bzw.

Kontrollparameter zu verwenden. Hierflir existieren bereits eine Reihe von Studien, die die

regionale Abhangigkeit zwischen benthischen Stoffflussen bzw. Einbettungsraten belegen

(z.B. Hensen et aI., 2000) und diese nutzen, um globale Velieilungsmuster des Flusses

pmiikularen organischen Materials zum Sediment zu erstellen (z.B. lahnke, 1996). lahnke

(1996) erstellte auf der Basis von 68 Messdaten eine empirische Gleichung, we1che die

KOlTelation zwischen benthischer Sauerstoffzehrung und Einbettungsrate organischen

Kohlenstoffs beschreibt. Regionale Unterschiede in der Sedimentationsrate, die zu annahernd

regionalspezifischen Einbettungsraten flihren, wurden durch eine Kalzit-Konektur der

Einbettungsraten berucksichtigt. Das Problem regional auftretender Unterschiede in bezug auf

partikulare Eintragspfade und Prozesse, die das Verhaltnis von Sauerstoffzehrung zu

Einbettung bestimmen, bleibenjedoch weitestgehend ungelOst (Abb. 3) und es zeigt sich, dass

eine Diskretisierung in benthische Provinzen sinnvoll erscheint.

Sathyendranath et aI. (1995) und Longhurst et aI. (1995) unterteilten den Weltozean in 57

biogeochemische Provinzen, die auf ozeanographische Randbedingungen und Chlorophyll-a­

Verteilungsmuster zuruckzuflihren sind. Die Chlorophyll-a Muster basieren auf CZCS (engI.

Coastal Zone Calor S.canners) Satelliten Daten (1979-1987). lm wesentlichen reflektiel'en

diese Provinzen Prozesse der oberen Wassersaule und regional charakteristische

ozeanographische Kreislaufe. Die Betrachtung benthischer Mineralisations- und

Freisetzungsraten erfordert dem gegenuber eine Regionalisierung, die bodennahe Prozesse

beliicksichtigt. In del' vorliegenden Arbeit wurde die hohe verfligbare Datendichte von TOC

im Obel'flachensediment genutzt, um benthische TOC-Provinzen auf der Basis von

Variogramm-Analysen (Kriging-Veliahren) zu definieren und eme globale TOC

Verteilungskarte (TOC-Raster, 1° x 1°) zu erstellen (vgI. Kapitel 2.1). Ausgehend von der

Charakterisierung 33 benthischer TOC-Provinzen wurden regional-spezifische

Transferfunktionen zwischen benthischen diffusiven Sauerstoffflussen (DOU) und TOC­

Konzentrationen im Oberflachensediment erstellt, um den minimalen Fluss partikularen

organischen Materials (]Poc) zum Sediment abzuschatzen (vgI. Kapitel 2.2). Zu diesem

Zweck wurde von einem Gleichgewicht zwischen DOU und dem minimal en Fluss an POC

(JPOCa) zum Sediment ausgegangen. Die resultierenden globalen Velieilungskarten

benthischer diffusiver Sauerstoffzehrung, bzw. POCs stellen die Gnmdlage flir weitere

Abschatzungen und Verteilungskarten zur benthischen Nahrstofffreisetzung, wie der

Siliziumfreisetzung dar (vg I. Kapitel 2.3). Zur Datendarstellung und Verwaltung, sowie del'

Transforn1ation des Basis-TOC-Rasters durch provinzspezifische Transferfunktionen in

regionale DOU- bzw. l poca-Verteilungskarten wurde ein Geo-Infonnations-System (GIS)

genutzt (vgI. Kapitel 1.3).

Eil11eitlll1g und Fragestelll1l1g 13

1.3 Nutzung eines Geo-!nformations-§.ystems

Geo-Informations-.s.ysteme (GIS) bieten die Moglichkeit raumbezogene Daten in hoher Anzahl einzulesen und in Fonn einer Datenbank zu verwalten und zu verarbeiten. Um das Ziel del' Erstellung globaler Stoffflusskarten und del' regionalen Bilanzierung del' Stoff­Umsatze zu realisieren, stellt die Verwendung eines GIS unter Einbeziehung del' internen und einer externen Datenbank das geeignete Mittel dar. Als Geo-Infonnations-System wurde das Programm Arc Vievv'@ 3.2 del' Finna ESRI (.E:nvironmental .s.ystems Research Institute, Inc.) gewahlt. Um die Funktionalitat des Programms zu erhohen, wurden zahlreiche Erweiterungen del' Firma ESRI und externe Erweiterungen implementiert.

Zur Dateneingabe und Verwaltung wurde u.a. das externe relationale Datenbanksystem Access 2000 genutzt. Del' V Olieil einer relationalen Datenbank besteht darin, dass del' Datenbestand, im Gegensatz zu hierarchischen Modellen, tabellarisch nach strukturellen Gesichtspunkten gegliedeli und gespeichert wird. Zwischen den Daten, sogenannten Entitaten, besteht keine Rangordnung, so dass redundanzfreie Relationen zwischen jeder Entitat oder einer Entitatsmenge (Tabellen) erstellt werden konnen. Durch diese TabellenstruktuI' eI'gibt sich eine hohe Flexibilitat des Datenmanagements, da jede Entitat odeI' Entitatsmenge als AbfI'ageschlussel eingesetzt werden kann.

Der DatentransfeI' wird durch ODBC-Schnittstellen (engl. Qpen Data !l.ase Connectivity) und die Anwendung del' international genormten Sprache SQL (engl. S,tructured Query Language)

gewahrleistet. SQL ennoglicht unabhangig von der Datenbank das Abfragen und Einspeisen von Daten.

Das Kemsruck eines GIS sind Daten, die durch ein Datenbankverwaltungssystem (engl. gata

f2.ase management rystem, DBMS) zu einer Datenbank angeordnet werden. Um diese Datenbank gruppieren sich eine Reihe von Werkzeugen zur DaI'stellung und VeI'schneidung von Daten, Karten odeI' Bildmaterial, zur Interpolation, Analyse und Transfonnation von Daten sowie Abfrage- und Auskunftsmodule (Abb. 4).

Geo-Infonnations-Systeme vereinen in ihrer Vielfalt die Anforderungen, die an Daten mit unterschiedlichen Eigenschaften gestellt weI'den konnen. Neben den geometrischen Eigenschaften (metrische und topologische) werden strukturelle und thematische Eigenschaften gespeichert, verwaltet und zur Abfrage bereitgestellt. Die Vorteile eines GIS liegen darin, Gemeinsamkeiten del' Daten und ihrer Eigenschaften zu eI'kennen und gegebenenfalls zu neuen StruktuI'en mit neuen Eigenschaften zu transfonnieren.

Um aus Rohdaten eine gute und aussagekraftige Ergebniskarie zu erhalten, sind oft viele zeitaufwendige Zwischenschritte notwendig. Die Rohdaten mussen in ein GIS bzw. Datenbank kompatibles Fonnat konvertieli weI'den, wobei sich die Anlage del' Datenbank in Abhangigkeit von der Fragestellung nach individuellen Gesichtspunkten I'ichtet. Vektordatenmodelle oder diskontinuierliche Punktdaten werden durch Anwendung verschiedener Interpolationsverfahren zu kontinuierlichen oder kategorischen Rasterdatenmodellen konveliiert.

14

Dateneingabe

Punhdaten/ Vektordaten Rasterdaten Bilddatcn

Einleitllng lInd FragestellzlI1g

GI S = Geo-lnfor1nations-Systen1 Inillrmationssystem fiir raumbezogene Daten

Verwaltullg m emer

Datcllhank

Datenausgabe

Punktdaten/ Vekturdaten Rasterdaten Bilcldaten

~7 l·· ~ Ers~.el1ung. IOgiSCherHi. - Erstel~ung von Rasterdatellln~del1ell. (RDl'\J) Abfragell. ,- raumhche Analyse und TrallSforIllatlOllen. (SQL) : - Verschneidungen, Projektionen

] [- Georeferellzieren ...

Abbildung 4: Fliessmodell zur Nutzung eines Geo-Informations-Systems zur Dateneinspeisung, Visualisierung und Verarbeitung von georeferenzierten Daten.

1.3.1 Datenmodelle

Digitale Datensatze werden in emern GIS in Form rnehrerer Modelle durch em DBMS verwaltet. Die Datenorganisation erfolgt liber raurnbezogene Koordinaten, d.h. alle Entitaten (=Datensatze) oder hoheren Einheiten werden raurnbezogen verwaltet.

In diesel' Arbeit wurde vorwiegend rnit zwel Datenrnodellen gearbeitet: dern Vektordatenrnodel (VDM) und dern Rasterdatenmodell (RDM). In einern Vektordatenrnodell werden Daten in Entitaten gleicher Eigenschaften zusarnrnengefasst und als Klassen definiert. Diese konnen bei geornetrischer Definition der Entitaten Punktdaten, Linienstiicke oder Flachen (=Polygonzlige) sein (Abb. 5). Irn folgenden werden Punktdaten, als Spezialfall des Vektordatenrnodells gesondert benannt. In einern GIS konnen den Entitaten bestirnmte sernantische Eigenschaften, wie Messstationen, Sauerstoffgehalt, Wassertiefe, TOe, Prirnarproduktion, benthische Stofffllisse oder Zonelllimnen zugeordnet und in einer Attributtabelle gespeichert werden.

Bei Punktdaten wird jede Lokation als einfaches xyz-Tripel behandelt. Linienstiicke sind Surnrnen von xyz-Tripeln und Flachen geschlossene Linienziige. Allen xyz-Tripeln werden Identifikationsschliissel zugeordnet. Parallel rnlissen jedoch auch zur Festlegung von Grenzen, Distanzen oder Uberschneidungen, Linienziige oder Flachen durch topologische, nicht raurnbezogene Attribute definiert werden. Zugewiesen werden Anfangs- und Endpunkte, linke und rechte Polygonflache und rnogliche charakterisierende Attribute wie Zonenname oder Flacheninhalt (ZonengroBe). Diese interne Datenbank ist der georeferenziert orientielien Datenbank untergeordnet. Der Zugriff auf die verschiedenen Datenbanken und deren Verknlipfung erfolgt intern liber die Identifikationsschllissel.

Einleitlll1g 1ll1d Fragestel1zll1g

Punktdaten z. B. TOC', BOC', DOlf

VeKtordaten z. B. ZOl1':l1,

Ivfeeresstn'il11ungen

R asterdaten z. B. TOC I

, ROC:, DOLT

x

15

.............................. ~._. SC'Nf5."'E

x x

7: X

x

Lage des lJrsprungs jeder Zell-Infonnation (x', y')

Abbildung 5: Informationsebenen in einem Geo-Info1111ations-System (GIS). Dargestellt sind Punktdaten­Vektordaten und Rasterdaten ( TOC=Total organic carbon, 2: BOC=Bottomwater oxygen content, 3: DOU=Diffusive oxygen uptake). Punktdaten werden reprasentieli durch z.B. unregelmaBig verteilte Messdaten. Vektordaten sind gerichtete Informationen wie Meeresstromungen oder benthische Provinzen. Rasterdaten sind Zell-Infonnationen, die in einem regelmaBigen Raster (=Matrix) in Fonn von Zeilen und Spalten angeordnet sind.

Ein weiteres Datenmodell stellt das Rasterdatenmodell (RDM) dar (Abb. 5). Die Welt wird als rechtwinklige Matrix aufgespannt und in Zeilen und Spalten unterteilt. J eder Zelle ist ein spezifischer Wert (Zell-Infonnation=Pixel) zugeordnet. Die Georeferenzierung jeder Zelle erfolgt uber den Ursprung des Rasters (P(xo,YO)) und ihrer relativen Lage zum Ursprung. Jede Zelle ist dabei sowohl geometrisch als auch topologisch definiert, da jede Zelle 8 Nachbam hat, die mit Hilfe der Spalten- und Zeilennummer identifiziert werden kannen. Die einfachste Form eines Rasterdatenmodels ist ein Satellitenbild oder ein Luftbild. Ahnlich wie ein Foto besteht es aus einer Serie von Punkten mit diskreten Farb- und Grauwerten. Aus Zellen mit gleichen Farb- und Grauwelien, bzw. Zellwerten als explizite Daten, werden Gruppen gebildet (Zonen-Grid), denen Attributdaten zugeordnet werden kannen. Datenmatrizen mit Gruppen gleicher Welie und zugeharigen Attributtabellen werden als kategorisches Grid (=Integer) gespeichert und erlauben nur ganzzahlige numerische Zellwelie del' Gruppen. Liegen Datenmatrizen mit reellen Zahlen vor, die sich nicht in Gruppen zusammenfassen lassen, so werden diese als kontinuierliches Grid (=Floating-Point) gespeichert. Fur ein GIS bedeutet das Erstellen eines kontinuierlichen Rasters einen enormen Speicheraufwand, da jede Zelle gesondert abgelegt, verwaltet und fUr die Datenabfrage bereitgestellt wird. Eine Attributtabelle steht fUr Floating-Point-Grids nicht zur VerfUgung. Je haher die Zeilen- und Spaltenzahl einer Matrix, desto haher die Aufiasung und kontinuierlicher ist das erstellte Bild. Del' erforderliche Arbeits- und Speicheraufwand fUr das GIS ist entsprechend haher.

16 Einleitung 1Ind Fragestellung

1.3.2 H ochaujl osende Rasterdatenmodell e

Die Datenquellen, die fUr ein Rasterdatenmodell zur VerfUgung stehen, sind vielfaltig. Im allgemeinen kann jede regelmaBige Punktdatenverteilung durch ein GIS m em Rasterdatenmodell konveliiert werden, da jeder Datenpunkt einer Zelle einer Matrix zugeordnet wird. Solche digitalen Punktdaten sind die als ASCII- oder Binar Daten im Internet l verfUgbaren ETOP05 Daten (Earth Topography - 5' Auflosung). Diese Daten wurden weltweit aus verschiedenen Quellen zusammengestellt und zu kategorischen Rasterdatensatzen mit diskreten (numerischen) Zellenwerten kompilieli. Die in der vorliegenden Arbeit verwendeten Daten zur Bathymetrie entstammen dem u.s. Naval

Oceanographic Office und sind tiber das National Geophysical Data Center (NGDC) abzufragen (http://www.ngdc.noaa.gov/mgg/global/seltopo.html). Eine hohere Auflosung bietet GTOP030. Die Rasterdaten werden dort in einer 2' Auflosung angeboten (http://edcdaac.usgs.gov/gtopo30/gtopo30.html). Globale Verteilungen von Primarproduktionsdaten werden tiber bestimmte Algorithmen (z.B. VGPM Modell, engl.: J!..ertically generalized 12.roduction model, Behrenfeld and Falkowski, 1997 -a,b) aus Farbverteilungsmuster der Meeresoberflache errechnet, die via Satellit aufgezeichnet werden (z.B. Nimbus 7). Mit Hilfe des CZCS (Coastal Zone Color S.canners) der NASA wurden Pigmentdaten des Oberflachenwassers erhoben und durch lineare Interpolation ein kontinuierliches Chlorophyll-a-Raster von 2048 x 1024 Datenpunkten (Pixel) erzeugt. Die Chlorophylldaten und die elTechneten Primarproduktionsmuster liegen als Binar-Daten vor und sind z.B. tiber die Oceanographic Productivity Database der Rutgers University (http://marine.rutgers.edu/opp/ProductionlVPGMRes.html) (Behrenfeld & Falkowski, 1997-a,b), bzw die JGOFS-France database (http://www.obs-vlfr.frljgofs2/modelisationlhome­p.htm) (Antoine et al., 1996) zuganglich.

Liegen Messdaten in hoher Auflosung in einem Arbeitsgebiet vor, konnen unter Berticksichtung bestimmter geostatistischer Bedingungen aus den Rohdaten kontinuierliche oder kategorische Rasterdaten erstellt werden. Diese Moglichkeit wurde in der vorliegenden Arbeit zur Erstellung emer Kmie des Gehaltes organischen Kohlenstoffs 11n Oberflachensediment (TOC) aus weltweit in hoher Auflosung vorliegenden TOC-Messdaten durch Interpolation genutzt. Daten verschiedener Informationsebenen Wle Sedimenteigenschaften, Oberflachenstromungen, Flusseintrag und Primarproduktiondaten wurden in das GIS eingespeist und in Verbindung mit geostatistischen Methoden (Semi­Variogramm-Analysen nach der Kriging-Methode) zur Charakterisierung benthischer Sediment-Provinzen eingesetzt (vgl. Kapite12.l).

Geo-InfoDnations-Systeme bieten die Moglichkeit verschiedene Rasterdatenmodelle gleicher Auflosung und Ausdehnung miteinander oder auch mit anderen Datenmodellen z.B. Vektordatenmodellen zu verkntipfen. Man kann sich die Datensatze als Informationsebenen gleicher Geokodierung (Lage des Ursprunges) vorstellen, die jeweils durch ein Datenmodell reprasentieli werden (vgl. Abb. 5). Durch die geographische Verkntipfung der Ebenen komlen die Beziehungen zwischen den einzelnen Infonnationsebenen analysiert und modellieli

I Stand 9.12.03

Einleitllllg 1Il1d Fragestel1zlI1g 17

werden. Zwischen verschiedenen Rasterdatenmodellen besteht zudem die Moglichkeit bei gleicher Ausdehnung, raumlicher Lage und Auflosung durch Anwendung von Algorithmen, die empirische Beziehungen zwischen den Ebenen abbilden, eine Matrix zu transformieren und dadurch neue Rasterdaten zu generieren. Diese Prozedur nennt man cell-by-cell processing, da die Transfonnation Zelle fUr Zelle vorgenommen wird. Maximale Informationserhaltung ist hierdurch gewahrleistet (Abb. 6).

2

3 4

--:. B. Kategrj~chc~ TOC '· Rastcr

z. B.: DOU'zuTOC' unci HOC'

z (x. yJ = ax- by+c

Erstdlen "in~~ ll~Uon Rosters C=.H. DOU) durch Anwendm del' Trambfi.mktion in sJle=i~sc:hell RegiOllen Raster 2= =(x, y)' Rast"r 1

Zur Berec1ulllng von Bilmcen und Budgds (=Raster 2*) =.B. Von Stofflliimn in spojiischen Regionen

Abbildung 6: Schematische Darstellung der Datenverarbeitung in einem GIS. Punktdaten werden zu Rasterdaten konvertiert (Raster 1) und empirische Relationen auf die Rasterdaten angewendet (Transferfunktionen). Durch cell-by cel1 processing wird dabei jede Zelle mit exakter geographischer Position neu berechnet (Raster 2). Einzelne Raster werden miteinander verschnitten und konnen zur Berechnung von Bilanzen und Budgets in eine flachentreue Projektion tranSf0l111iert werden. 1: TOC=Total organic carbon; 2: DOU=Diffusive oxygen uptake; 3: BOC=Bottomwater oxygen concentration.

In Abbildung 6 ist dargestellt Wle die einzelnen Infonnationsebenen unterschiedlicher Datenmodelle in der vorliegenden Arbeit miteinander verknlipft wurden. Die Schritte 1 bis 5 fassen die notwendigen Datentransfonnationen von den umegelmaBig verteilten TOC­Messdaten lib er empirische Relationen zu einem neuen Raster z.B. der diffusiven Sauerstoffaufnahme (DaU) in einer flachentreuen Projektion zusammen.

Zur Darstellung und Bearbeitung von Stoffflusskarten und der Bilanzierung des Stoffhaushaltes bieten Geo-Informations-Systeme die Moglichkeit zur flachentreuen Projektion des Rasters. Die Stofffllisse konnen so pro Flacheneinheit, d.h. Zelle fUr Zelle, exakt berechnet werden. Dies erfordert eine geometrische Transformation von einer geographischen Projektion in eine flachentreue (z.B. Lambert- Azimutal oder Mollweide), wobei sowohl die ZellengroBe als auch die Zellenanzahl neu berechnet werden mlissen (Resampling). Die Einheit Dezimalgrad CO) wird in ein metrisches System (m) umgerechnet.

18 Einleitltng ltl1d Frageste111111g

Der VOlieil der gewahlten Projektion liegt in der minimal en Verzerrung, auch bei groi3erer Entfernung vom Projektionsmittelpunkt.

1.4 Kurzfassung der Manuskripte

Im folgenden werden die Inhalte der VIer eingereichten Manuskripte in kurzer Form dargestellt. Sie sind im Rahmen del' vorliegenden Dissertationsschrift entstanden. Die Manuskripte 1-3 (vgl. Abschnitt 2.1-2.3) wurden von mir als Erst-Autorin erstellt. Das 4. Manuskript (Abschnitt 2.4) ist thematisch eng an diese Manuskripte angegliedert und umfai3t eine Obersicht libel' die im Rahmen des SFB 261 (an del' Universitat Bremen) entstandenen Arbeiten zur Ennittlung von Stofffllissen liber die Sediment-Wasser Grenzschicht. Mein Beitrag liegt in der Erstellung der Verteilungskmien zur supralysoklinaren KalzitlOsung mit einem GIS und beruht auf Ergebnissen von Pfeifer et al. (2002). Die Manuskripte 1-3 beruhen auf eigenen Untersuchungen und wurden von mir selbststandig verfasst.

Manuskript 1

Regionalization of the organic carbon content in surface sediments - defining a new approach of benthic TOe-based regional provinces

K. Seiter, C. Hensen, J. Schroter, M. Zabel

VerfUgbare Daten zur weltweiten Velieilung organischen Kohlenstoffs Im Oberflachensediment (TOC) wurden aus unterschiedlichen Quellen kompiliert und regional anhand verschiedener Randbedingungen wie ozeanographischer Einflussgroi3en, Sedimentparametern und Berlicksichtigung lateral er Eintragsquellen zunachst qualitativ und anschliei3end durch Anpassung von Semi-Variogrammen nach dem Kriging-Verfahren auf regionale, ortsabhangige Strukturen geostatistisch untersucht (Quantitative Analyse). Die qualitative und quantitative Analyse ermoglichte die Einteilung in 33 benthische Sediment­Provinzen. In Abhangigkeit der Sediment-Provinzen und der zugehorigen spezifischen Semi­Variogramme konnte eine qualitativ hochwertige globale TOC-Rasterdaten-Verteilung unter Berticksichtigung regional variabler ozeanographischer und sedimentspezifischer Einflussfaktoren in einer 10 x 10 Auflosung erstellt werden. Sowohl die resultierende Einteilung in benthische Sediment-Provinzen, als auch das TOC-Raster bildet eine wichtige Grundlage fUr die folgenden Manuskripte 2 und 3.

Manuskript 2

Benthic carbon mineralization on a global scale

K. Seiter, C. Hensen, M. Zabel

In dies em Manuskript wurden ausgehend von den im 1. Manuskript entwickelten 33 benthischen Sediment-Provinzen in 11 libergeordneten Regionen empirische Beziehungen zwischen dem diffusiven benthischen Sauerstofffluss (DOU), dem Gehalt an organischem Kohlenstoff im Oberflachensediment (TO C) und dem Sauerstoffgehalt im Bodenwasser (BOC) hergeleitet. Dabei wurde die hohe Datenanzahl der Kontrollparameter TOC und BOC

Einleitll11g 1ll1d Fragestel/zll1g 19

durch Anwendung der regionalen Transferfunktionen genutzt, um regionale Verteilungsmuster des Anteils partikuUiren organischen Materials (POM) zu erstellen, der den

Meeresboden erreicht. Da der Hauptumsatz organischen Materials des uberwiegend

oligotrophen Weltozeans mit Sauerstoff als Oxidationsmittel geschieht, und nur ein kleiner Anteil des akkumulielien organischen Materials im Sediment eingebettet wird, konnte davon ausgegangen werden, dass die diffusiven benthischen Sauerstoffzehrungsraten den

pariikuHiren Fluss organischen Materials (JPOCa) zum Sediment gut widerspiegeln. Zur Obeliragung der Transferfunktionen wurde das TOC-Raster aus Manuskript 1 und eines

ebenfalls im Rahmen des 2. Manuskriptes erstelltes BOC-Raster eingesetzt. Die an Provinzen mit guter verftigbarer Datenbasis gebundenen Transferfunktionen wurden auf weniger gut

beprobte Gebiete mit vergleichbaren ozeanographischen, biogeochemischen und sediment­spezifischen Randbedingungen ubertragen. Neben den verschiedenen Faktoren, die zu

provinzspezifischen Transferfunktionen ftihren, wurden daher sowohl regionale Unterschiede in den Velieilungsmustern des abgeschiitzten Partikelflusses zum Sediment, als auch ihr

Anteil am Gesamtbudget diskutiert. Ein besonderer Schwerpunkt liegt auf dem Vergleich zwischen den Verteilungsmustern, die von einem rein vertikalen Transport des partikuliiren

organischen Materials uber der Wassertiefe ausgehen und der in dieser Arbeit entwickelten

Methode, die lateralen Eintragswege zu berucksichtigen.

Manuskript 3

The benthic silica release and its implication for the estimation of the non-lithogenic particle fluxes to the sea floor

K. Seiter, J. M. Holstein, C. Hensen, M. Zabel

Die Untersuchung der Kopplung zwischen mikrobieller Aktivitiit und benthischen diffusiven

Siliziumflussen und die Herleitung einer empirischen Beziehung zwischen Silizium Ruckfluss

und diffusivem Sauerstofffluss uber die Sediment-Wasser Grenze stellen einen wesentlichen Bestandteil des 3. Manuskriptes dar. Desweiteren konnte der SiliziumrUckflusses durch den

aeroben Abbau organischer Substanz (Sauerstoffzehrung) quantifiziert werden. Die in diesem

Zusammenhang entstandene Verteilungskarte benthischer Siliziumfreisetzung bildete die

Grundlage des abgeschiitzten Flusses biogenen Opals zum Sediment.

Ergiinzend zu den abgeschiitzten Partikelflussen organischen Kohlenstoffs zum Sediment

(2. Manuskript) wurden die ermittelten empirischen Zusammenhiinge sowie bekannte Beziehungen zwischen Kalzit16sung und aerobem Abbau organischen Materials genutzt, um

biogene Partikelflusse zum Sediment abzuschiitzen. Die Addition der einzelnen Partikelflusse

ftihrte zur Abschiitzung eines minimalen biogenen Paliikelflusses zum Sediment. Die Ergebnisse zeigen, dass die abgeschiitzten Pmiikelflusse herkommliche Akkumulationsraten

oft um ein Vielfaches uberschreiten, da der Hauptteil des biogenen Materials, das den

Meeresboden eneicht, sehr schnell remineralisiert wird.

Die in dies em Manuskript entwickelten empirischen Zusammenhiinge sttitzen sich auf eigene

Arbeiten, sowie Ergebnisse von Pfeifer et al. (2002).

20 Ein/eitllng 1I11d Fragestellzll1g

Manuskript 4

Fluxes at the benthic boundary layer-a global view from the S-Atlantic

C. Hensen, K. Pfeifer, F. WenzhOfer, A. Volbers, S. Schulz, J. Holstein, O. Romero, K. Seiter

Das Manuskript enthalt eine umfassende Betrachtung und Zusammenfassung zu Arbeiten bezliglich der Quantifizierung benthischer Stofffllisse im Slid-Atlantik. Ein besonderer Schwerpunkt liegt auf der Abschatzung regionaler und globaler Massenbilanzen. Neben regionalen Zusammenhangen zwischen benthischen Zehrungs- und Freisetzungsraten und Primarproduktion wird auch der Einfluss von Abbauprozessen des organischen Materials auf die Kalzitlosung im Oberflachensediment quantifiziert und global abgeschatzt. Auf der Grundlage der von Pfeifer et al. (2002) aufgestellten Beziehung zwischen benthischer Mineralisierung organischer Substanz und KalzitlOsung im Oberflachensediment wurde von mir eine globale Verteilungskarte der Kalzitlosung oberhalb der Lysokline erstellt.

Kartenanhang

Erganzend zu den in dieser Arbeit behandelten globalen Kartendarstellungen, wurden einige Regionen zusatzlich in hoherer Auflosung bearbeitet. Es wurden jeweils drei TOC- und JpOCa- Verteilungskarten in 0.1 ° x 0.1 o-Auflosung (Basis: TOC) des SW-afrikanischen Kontinentalhanges (A-I, H), des nordlichen Nordatlantiks (A-IH, IV) und der Arabischen See (A-V, VI) angefeliigt; diese stellen erganzende Arbeiten dar, die im Zusammenhang mit den Ergebnissen des 1. und 2. Manuskriptes (Abschnitt 2.1 und 2.2) zu sehen sind.

Ebenfalls im Kartenanhang befinden sich drei im Rahmen des 1. und 2. Manuskriptes (Abschnitt 2.1 und 2.2) enstandene Darstellungen. Zur besseren Obersicht wurden diese in den Manuskripten enthaltenen Karten zusatzlich im Anhang dargestellt. Es handelt sich um die glob ale TOC-Verteilungskarte (1 ° x 1°) (B-I), die BOC-Verteilungskatie (B-H) und die globalen Velieilungsmuster zur Abschatzung des Flusses partikularen organischen Materials zum Sediment (JPOCa) (B-Ill).

Datenanhang

Die den einzelnen Manuskripten zugehorigen Datenanhange sind am Ende dieser Arbeit zusammengefasst (Appendix I-IV).

Einleitllng zl11d Fragestellzl11g 21

Literatur

Anderson, L.A., Sarmiento, J.L., 1994. Redfield ratios ofremineralization determined by nutrient data analysis.

Global Biogeochemical Cycles 8, 65-80.

Antia, A.N., Koeve, W., Fischer, G., Blanz, T., Schulz-Bull, D., Scholten, l, Neuer, S., Kremling, K., Kuss, J.,

Peinert, R., Hebbeln, D., Bathmann, U., Conte, M., Fehner, U., Zeitschel, B., 2001. Basin-wide

particulate carbon flux in the Atlantic Ocean: Regional export patterns and potential for atmospheric

CO2 sequestration. Global Biochemical Cycles 15, 845-862.

Antoine, D., Andre, l-M., Morel, A., 1996. Oceanic primary production; 2. Estimation at global scale from

satellite (coastal zone color scanner) chlorophyll. Global Biogechemical Cycles 10,57-69.

Armstrong, R.A., Lee, c., Hedges, l1., Honjo, S., Wakeham, S. G., 2002. A new mechanistic model for organic

fluxes in the ocean based on the quantitative association of POC with ballast minerals. Deep-Sea

Research Il49, 219-236.

Arthur, M.A., Dean, W.E., Laarkamp, K., 1998. Organic carbon accumulation and preservation in surface

sediments on the Peru margin. Chemical Geology 152,273-286.

Behrenfeld, M.l, Falkowski, P.G., 1997a. A consumer's guide to phytoplankton primary productivity models.

Limnology and Oceanography 42, 1479-1491.

Behrenfeld, MJ., Falkowski, P.G., 1997b. Photosynthetic rates derived from satellite-based chlorophyll

concentration. Limnology and Oceanography 42, 1-20.

Bender, M.L., Heggie, D.T., 1984. Fate of organic carbon reaching the deep-sea floor: a status report.

Geochimica et Cosmochimica Acta 48,977-986.

Berger, W.H., Smetacek, V.S., Wefer, G., 1989. Ocean productivity and paleoproductivity - an overview. In:

Berger, W. H., Smetacek, V. S., Wefer, G. (Eds.), Productivity of the ocean: present and past, 1989.

John Wiley & Sons Lim., pp. 1-34.

Betzer, P.R., Showers, W.l, Laws, E.A., Winn, C.D., DiTullio, G.R., Kroopnick, P.M., 1984. Primary

productivity and particle fluxes on a transect of the equator at 153°W in the Pacific Ocean. Deep-Sea

Research 31, 1-11.

Bidle, K.D., Azam, F., 1999. Accelerated dissolution of diatom silica by marine bacterial assemblages. Nature

397,508-512.

Bidle, K.D., Brzezinski, M.A., Long, R.A., Jones, lL., Azam, F., 2003. Diminished efficiency in the oceanic

slica pump caused by bacteria-mediated silica dissolution. Limnology and Oceanography 48, 1855-

1868.

Brzezinski, M.A., Dickson, M.L., Nelson, D.M., Sambrotto, R., 2003. Ratios of Si, C and N uptake by

microplancton in the Southern Ocean. Deep-Sea Research II 50, 619-633.

Broecker, W.S., Peng, T.-H., 1983. Tracers in the sea: Lamont-Dohery Geological Observatory, Columbia

University, Palisades, New York.

Cai, W.-J., Reimers, c.E., 1995. Benthic oxygen flux, bottom water oxygen concentration and core top organic

carbon content in the deep northeast Pacific Ocean. Deep-Sea Research 42 (l0), 1681-1699.

Canfield, D.E., 1989. Sulfate reduction and oxic respiration in marine sediments: implications for organic carbon

preservation in euxinic environments. Deep-Sea Research 36, 121-138.

Canfield, D.E., 1993. Organic matter oxidation in marine sediments: Interactions of C,N,P and S biogeochemical

cycles, 1993. Springer-Verlag, pp. 333-363.

Christensen, lP., 2000. A relationship between deep-sea benthic oxygen demand and oceanic primary

productivity. Oceanologica Acta 23, 65-82.

22 Einleitllng lI11d Fragestellung

Dugdale, R.e., Wilkerson, F.P., 1998. Silicate regulation of new production in the equatorial Pacific upwelling.

Nature 391, 270-273.

Dugdale, R.e., Wilkerson, F.P., Minas, HJ., 1995. The role ofa silicate pump in driving new production. Deep­

Sea Research 42,697-719.

Freudenthal, T., Neuer, S., Meggers, H., Davenport, R., Wefer, G., 2001. Influence of lateral particle advection

and organic matter degradation on sediment accumulation and stable nitrogen isotope ratios along a

productivity gradient in the Canary Island region. Marine Geology 177, 93-109.

Froelich, P.N., Klinkh31IDner, G.P., Bender, M.L., Luedtke, N.A., Heath, G.R., Cullen, D., Dauphin, P.,

Hammond, D., Hartman, B., Maynard, V., 1979. Early oxidation of organic matter in pelagic sediments

of the eastern equatorial Atlantic: suboxic diagenesis. Geochimica et Cosmochimica Acta 43, 107S-

1088.

Garzoli, S.R., 1993. Geostrophic velocity and transport variability in the Brazil-Malvinas Confluence. Deep-Sea

Research 140,1379-1403.

Giraudeau, l, Meyers, P.A., Christensen, B.A., 2002. Accumulation of organic and inorganic carbon 111

Pliocene-Pleistocene sediments along the SW African margin. Marine Geology 180,49-69.

Hartnett, H.E., Keil, RG., Hedges, J.1., Devol, A.H., 1998. Influence of oxygen exposure time on organic carbon

preservation in continental margin sediments. Nature 391, S72-S74.

Hedges, l1., 1992. Global biogeochemical cycles: progress and problems. Marine Chemistry 39,67-93.

Henrichs, S.M., 1992. Early diagenesis of organic matter in marine sediments: progress and perplexity. Marine

Chemistry 39, 119-149.

Henrichs, S.M., Reeburgh, W.S., 1987. Anaerobic mineralization of marine sediment organic matter: rates and

the role of anaerobic processes in the oceanic carbon economy. Geomicrobiological JournalS, 191-237.

Hensen, e., 1996. Friihdiagenetische Prozesse und Quantifizierung benthischer Stoff-Fliisse m

Oberflachensedimenten des Siidatlantiks. Berichte aus dem Fachbereich Geowissenschaften der

Universitat Bremen 88, Bremen, Deutschland.

Hensen, e., Landenberger, H., Zabel, M., Schulz, H.D., 1998. Quantification of diffusive benthic fluxes of

nitrate, phosphate and silicate in the Southern Atlantic Ocean. Global Biogeochemical Cycles 12, 193-

210.

Hensen, e., Zabel, M., Schulz, H.D., 2000. A comparison ofbenthic nutrient fluxes from deep-sea sediments of

Namibia and Argentinia. Deep-Sea Research Il47, 2029-20S0.

Hensen, e., Zabel, M., Pfeifer, K., Schwenk, T., Kasten, K., Riedinger, N., Schulz, H.D., Boetius, A., 2003.

Control of sui fate pore-water profiles by sedimentary events and the significance of anaerobic oxidation

of methan for the burial of sulfur in marine sediments. Geochimica et Cosmochimica Acta 67, 2631-

2647.

IPCC, 2001. Climate change 2000-the science of climate change, contribution of working group I to the third

Assessment Report, technical summary of the working group I report; intern et address (URL):

http//www.ipcc.ch: Interngovernmental Panel on Climate Change.

Ittekkot, V., Rixen, T., Suthhof, A., Unger, D., 2002. Der glob ale Kohlenstoffkreislauf. In Wefer, G. (Ed.),

Expedition Erde. Beitrage zum Jahr der Geowissenschaften 2002, Bremen, Deutschland.

Ittekkot, V., 1988. Global trends in the nature of organic matter in river suspensions. Nature 332, 436-438.

Jahnke, R.A., 1996. The global ocean flux of particulate organic carbon: Areal distribution and magnitude.

Global Biogeochemical Cycles 10, 71-88.

Jahnke, R.A., Reimers, e.E., Craven, D.B., 1990. Intensification of recycling of organic matter at the sea floor

near ocean margins. Nature 348, SO-S4.

Einleitllng lInd Fragestel1zlI1g 23

J0rgensen, B.B., 1982. Mineralization of organic matter in the sea bed - the role of sulfate reduction. Nature,

296,643-645.

Kawahata, H., 2002. Suspended and settling particles in the Pacific. Deep-Sea Research II 49,5647-5664.

Klaas, C, Archer, D.E., in press. Association of sinking organic matter with various types of mineral ballast in

the deep sea: Implications for the rain ratio. Global Biogeochemical Cycle.

Lampitt, R.S., Antia, AN., 1997. Particle flux in the deep seas: Regional characteristics and temporal variability.

Deep-Sea Research I 44, 1377-1403.

Longhurst, A, Sathyendranath, S., Platt, T., Caverhill, CM., 1995. An estimate of global primary production in

the ocean from satellite radiometer data. Journal of Plankton Research 17, 1245-1271.

Ludwig, W., Probst, J.-L., Kempe, S., 1996. Predicting the oceanic input of organic carbon by continental

erosion. Global Biogeochemical Cycles 10,23-41.

Martin, J.H., Knauer, G.A, Karl, M., Broenkow, W.W., 1987. VERTEX: carbon cycling in the nOliheast Pacific.

Deep-Sea Research 34, 267-285.

Meybeck, M., 1993. C, N, P and S in rivers: from sources to global inputs. In Wollast, R., Mackenzie, F. T.,

Chou, L. (Eds.), Interactions of C, N, P and S: biogeochemical cycles and global change, 1993.

Springer, pp. 163-193.

Middelburg, J.J., Vlug, T., Jaco, F., Van der Nat, W.A, 1993. Organic matter mineralization in marine systems.

Global and Planetary Changes 8,47-58.

Mollenhauer, G., Schneider, RR., MUller, PJ., SpieJ3, V., Wefer, G., 2002. GlaciaVInterglacial variability in the

Benguela upwelling system: Spatial distribution and budgets of organic carbon accumulation. Global

Biogeochemical Cycles 16, 1134.

Nelson, D.M., Treguer, P., Brzezinski, M.A, Leynaert, A., Queguiner, B., 1995. Production and dissolution of

biogenic silica in the ocean: Revised global estimates, comparison with regional data and relationship to

biogenic sedimentation. Global Biogeochemical Cycle 9,359-372.

Neuer, S., Ratmeyer, V., Davenport, R., Fischer, G., Wefer, G., 1997. Deep water particle flux in the Canary

Island region: seasonal trends in relation to long-term satellite derived data and lateral sources. Deep­

Sea Research I 44,1451-1466.

Pace, M.L.G.A., Knauer, D.M.K., Martin, J.H., 1987. Primary production, new production and vertical flux in

the eastern Pacific. Nature 325, 803-804.

Pefia, M.A., Denman, K.L., Forbes, J.R., Calvert, S.E., Thomsen, R.E., 1996. Sinking particle fluxes from the

euphotic zone over the continental slope of an eastern boundary current region. Journal of Marine

Research 54,1097-1122.

Peterson, RG., Johnson, CS., KJ'auss, W., Davis, R.E., 1996. Langrangian measurements in the Malvonas

Current. In Wefer, G., Siedler, G., Berger, W. H., Webb, D. J. (Eds.), The South Atlantic: Present and

past circulation, 1996. Springer, pp. 239-247.

Pfeifer, K., Hensen, C, Adler, M., Wenzhoefer, F., Weber, B., Schulz, H.D., 2002. Modeling of subsurface

calcite dissolution including the respiration and reoxidation processes of marine sediments in the region

of equatorial upwelling off Gabon. Geochimica et Cosmochimica Acta, 66(24), 4247 -4259.

Ragueneau, 0., Treguer, P., Leynaert, A., Anderson, RF., Brzezinski, M.A, DeMaster, DJ., Dugdale, R.e.,

Dymond, J., Fischer, G., Francois, R., Heinze, C, Maier-Reimer, E., Martin-Jezequel, V., Nelson,

D.M., Queguiner, B., 2000. A review of the Si cycle in the modern ocean: recent progress and missing

gaps in the application of biogenic opal as a paleoproductivity proxy. Global and planetary Change 26,

317-365.

24 Einleitzll1g 1ll1d Fragestel1zmg

Redfield, A.C., 1958. The biological control of chemical factors in the environment. American Scientist 46, 205-

22l.

Reimers, C.E., 1989. Control of benthic fluxes by particulate supply. In Berger, W.H., Smetacek, V.S., Wefer,

G. (Eds.), Productivity of the ocean: Present and past, 1989.1. Wiley & Sons Lim., pp. 217-233.

Romankevich, E.A., 1984. Geochemistry of organic matter in the ocean. Springer, Berlin - Heidelberg - New

York - Tokio.

Rowe, G.T., Boland, G.S., Phoel, W.c., Anderson, R.F., Biscaye, P.E., 1994. Deep-sea floor respiration as an

indication of lateral input of biogenic detritus from continental margins. Deep-Sea Research II 41, 657-

668.

Rullk6tter, 1., 2000. Organic Matter: The driving Force for Early Diagenesis. In Schulz, H.D., Zabel, M. (Eds.),

Marine Geochemistry, 2000. Springer.

Sarmiento, 1.L., Gmber, N., 2002. Anthropogenic Carbon Sinks. Physics Today 55, 30-36.

Sathyendranath, S., Longhurst, A., Caverhill, C.M., Platt, T., 1995. Regionally and seasonally differentiated

primary production in the North Atlantic. Deep-Sea Research 42, 1773-1802.

Schlesinger, W.H., Melack, J.M., 1981. Transport of organic carbon in the worlds rivers. Tellus, 172-187.

Schlitzer, R., 2002. Carbon export fluxes in the Southern Ocean: results from inverse modeling and comparison

with satellite-based estimates. Deep-Sea Research II 49,1623-1644.

Schliinz, B., Schneider, B., 2000. Transport of terrestrial organic carbon to the oceans by rivers: re-estimating

flux- and burial rates. International Journal of Earth Sciences 88, 599-606.

SchlUter, M., Sauter, EJ., Schiifer, A., Ritzau, W., 2000. Spatial budget of organic carbon flux to the seafloor of

the northern North Atlantic (600N-800N). Global Biogeochemical Cycles 14,329-340.

Siegenthaler, u., Sarmiento, J.L., 1993. Atmospheric carbon dioxide and the ocean. Nature 365,119-125.

Sigman, D.M., Boyle, E.A., 2000. Glaciallinterglacial variations in atmospheric carbon dioxide. Nature 407,

859-869,2000.

Smetacek, V., 1999. Bacteria and silica cycling. Nature 397, 475-476.

Suess, E., 1980. Particulate organic carbon flux in the oceans - surface productivity and oxygen utilization.

Nature 288, 260-263.

Takahashi, T., Broecker, W.S., Langer, S., 1985. Redfield ratio based on chemical data from isopycnal surfaces.

Journal of Geophysical Research 90, 6907-6924.

Timmennans, K.R., Stolte, W., de Baar, HJ.W., 1994. Iron mediated effects on nitrate reductase in marine

phytoplancton. Marine Biology 121,389-396.

Treguer, P., Nelson, D.M., Gueneley, S., Zeyons, c., MOl·van, 1., Buma, A., 1990. The distribution of biogenic

and litho genic silica and the composition of particulate organic matter in the Scotia Sea and the Drake

Passage during autumn 1987. Deep-Sea Research 37, 833-851.

Treguer, P., Nelson, D.M., van Bennekom, A.J., DeMaster, DJ., Leynaert, A., Queguiner, B., 1995. The silicate

balance in the world ocean: A reestimate. Science 268,375-379.

Wefer, G., Fischer, G., 1993. Seasonal patterns of vertical particle flux in equatorial and coastal upwelling areas

of the Eastern Atlantic. Deep-Sea Research 40, 1613-1645.

van Bennekom, AJ.V., 1988. Deep-water transit times in the eastern Indonesian basin, calculated from dissolved

silica in deep and interstitial waters. Netherlands Journal of Sea Research 22, 341-354.

van Cappellen, P.V., Qiu, L., 1997. Biogenic silica dissolution in sediments of the Southern Ocean. 1. Solubility.

Deep-Sea Research II 44, 1109-1128.

Einleitz/l1g zlI1d Fragestel1zmg 25

WenzhOfer, F., Glud, R., 2002. Benthic carbon mineralization in the Atlantic: a synthesis based on in situ data

from the last decade. Deep-Sea Research 149,1255-1279.

Wollast, R., 1998. Evaluation and comparison of the global carbon cycle in the coastal zone and in the open

ocean, In Brink, K.H., Robinson, A.R. (Eds.), The Sea, Vol. 10, 1998. University of Brussels, pp. 213-

252.

Zabel, M., 1994. FrUhdiagenetische Stoff-F1i.isse in OberfHichen-Sedil11enten des aquatorialen und ostlichen

SUdatlantik. Berichte aus dem Fachbereich Geowissenschaften der Universitat Bremen 50, Brel11en,

Deutschland.

Zabel, M., Dahmke, A., Schulz, H.D., 1998. Regional distribution of phosphate and silicon fluxes across the

sediment water interface in the eastern South Atlantic. Deep-Sea Research 45,277-300.

Zabel, M. Hensen, c., SchlUter, M., 2000. Back to the ocean cycles: Benthic fluxes and their distribution

patterns. In Schulz, H.D., Zabel., M. (Eds.), Marine Geochel11istry, 2000, Springer. pp. 373-394.

26 Regionalization of the organic carbon content in surface sediments

2. Ergebnisse

2.1 Regionalization of the organic carbon content in surface sediments -

defining a new approach of benthic TOe-based regional provinces

Katherina Seiter*' 1, Christian Hensen2, Jurgen Schroter1, Matthias Zabel 1

lDepartment of Geosciences, University of Bremen, Klagenfurter StraJ3e, D-28359 Bremen, Germany

2GEOMAR - Forschungszentrum fUr Marine Geowissenschaften, SFB 574, Wischhofstr. 1-3,24148 Kiel,

Germany

Abstract

Approaches to quantify the organic carbon accumulation on a global scale generally do not consider small-scale

variability of sedimentary and oceanographic boundary conditions along continental margins. In this study, we

present a new approach to regionalize the total organic carbon content (TOC) in surface sediments « 5 cm

sediment depth). It is based on a compilation of 5553 single measurements from various sources. The global

TOC distribution was determined by the application of a combined qualitative and quantitative-geostatistical

method. Overall, 33 benthic TOC-based provinces were defined and used to process the global distribution

pattern of the TOC content in surface sediments in a 10 x 10 grid resolution. Regional dependencies of data

points within each single province are expressed by modeled semi-variograms. Measured and estimated TOC

values are well correlated, emphasizing the reasonable applicability of the method. Since, the accumulation of

organic carbon in marine surface sediments is a key parameter in terms of controlling mineralization processes

and material exchange between the sediment and the ocean water, this approach will help to improve global

budgets of nutrient and carbon cycles.

Regionalization of the organic carbon content in sur/ace sedil71ents 27

1. Introduction

Marine sediments are the largest reservoir of organic carbon on earth and thus an important

factor in terms of climate change. The major part of organic carbon is generated by

photosynthetic fixation of inorganic carbon by tenestrial plants and marine phytoplankton

from atmospheric CO2. The marine primary production (PP), as the amount of organic

material available to suppOli consumers, non-photosynthetic protists, and decomposers is in

the order of 40 to 50 GtC yr-1 (Berger et. aI., 1989; Antoine et aI., 1996; Behrenfeld and

Falkowski, 1997, Sanniento and Gruber, 2002). Only a small amount of sinking particulate

organic carbon (1.5 % of PP in the open ocean and up to 17 % on the slopes) is deposited at

the benthic boundary layer, where the largest part of it is oxidized (Lampitt and Antia, 1997;

Wollast, 1998; Schlitzer, 2002; WenzhOfer and Glud, 2002). According to Wollast (1998),

0.5 -3 % of the primary production along the continental shelves and slopes and 0.014 % in

the open oceans is buried. Continental margins make up only a small portion of about 20 % of

the global ocean area. However, 80 % of the globally accumulated organic matter is buried

here (Wollast, 1998) and thus they are the major repositories of carbon in the present day

ocean (Premuzic et aI., 1982; Henrichs and Reeburgh, 1987; Walsh, 1991). Generally, the

organic carbon flux decreases rapidly with increasing water depth as the material is

remineralized and various relationships have repOlied to describe this process (Suess, 1980;

Betzer et aI., 1984; Mmiin et aI., 1987; Pace et aI., 1987; Berger et aI., 1989; Antia et aI.,

2001). On ocean-wide scales the accumulation and burial of organic matter (total organic

carbon content, TOC) basically reflects the distribution pattern of primary production.

Regionally, however, there exist small-scale variability in biogeochemical and sedimentary

processes, which decouple this simple relation as previously shown by geochemical

investigations at the sediment-water interface of the eastern south Atlantic

(e.g. Zabel et aI., 1998).

Most striking in this regard is that a large amount of approximately 2.2 GtC yr- 1 of the

primary production produced in coastal areas is expOlied from the coastal zone to the open

ocean (Wollast, 1998). The main responsible process is the intense lateral export of suspended

organic matter (SPM) across the shelf and slope due to the cunent patterns (Garzoli, 1993;

Rowe et aI., 1994; Peterson et aI., 1996; Wollast, 1998; Alihur et aI., 1998; Hensen et aI.,

2000; Giraudeau et aI., 2002; Mollenhauer et aI., 2002). An additional source of organic

carbon is supplied by riverine discharge and eolian input. The permanent global riverine

discharge of organic carbon is about 0.4 GtC yr-1 (Schlesinger and Melack, 1981; Ittekkot,

1988; Hedges, 1992; Meybeck, 1993; Ludwig et aI., 1996). The fate of this - particularly

refractory - material is widely unknown.

Usually, the input of organic matter is positively conelated with the sedimentation rate and

the burial rate of organic matter (Henrichs, 1992). In rapidly deposited, predominantly

siliziclastic slope and shelf facies, however, these parameters are negatively conelated

(Jahnke, 1996; Tyson, 2001). Accordingly, high mass fluxes of biogenic opal, calcite, and the

tenigeneous material have a variable impact on the sedimentation rate and thus, on the

organic carbon preservation and TOC in surface sediments (e.g. Jahnke, 1996; Tyson, 2001).

28 Regionalizatiol7 of the organic carbon content in s1Irface sedi111ents

The first extensive data compilations of TOC in surface sediments were provided by Bordovskiy (1965), Romankevich (1977) and Brezukov et al. (1977) and summarized by

Premuzic et al. (1982), Romankevich (1984), and Berger and Wefer (1992). Their maps

clearly reflect the distribution pattem of high organic carbon deposits (> 0.5 wt%) along the continental margins and low contents in basin areas « 0.5 wt%). Further large-scale data sets

were compiled by Cwienk (1986) for the N0l1h Pacific, the southem Atlantic Ocean (Mollenhauer et aI., in press) and by Vetrov et aI. (1997), and Vetrov and Romankevitch (1997) for the Indian Ocean. Apart from the regional studies of the latter, all these benthic

TOC distribution maps result from simple large-scale interpolations. Specific regional

interrelations between control parameters for the TOC content are not examined.

A promising approach in this regard is given by Longhurst et aI. (1995) and Sathyendranat

(1995), who subdivided the surface Ocean into 57 biogeochemical provinces.

Due to the decoupling of the primary production and the TOC content at the sea floor, their classification cannot be simply transferred to the benthic regime. In this study, we use the

calcium carbonate and opal content in surface sediments, differences in primary production, surface current pattems and local riverine input to describe discriminable benthic TOC-based

regional provinces in the sense of biogeochemical provinces. Based on this approach, we

present a new global map of the benthic TOC distribution in a lO x 10 grid resolution.

2. Material and Methods

Data base

For the processing of the benthic TOC distribution pattem, we compiled 5553 data points (Fig. 1). The major portion of the point-data was retrieved from the world data base Pangaea

(www.pangaea.de) or was adapted from the literature. The data base was quality controlled,

regarding redundancies and inconsistencies. A complete list of all sampling sites and

references compiled within this project can be queried from the data archive Pangaea

(www.pangaea.de). The TOC content is given in weight percent of dry, salt-free sediment.

Methods of measurements and sampling devices are described in the respective references.

For presenting a general global distribution pattem of calcite and opal in surface sediments, we used 5147 and 3413 single measurements of calcite and opal, respectively. The main

p0l1ion of the data was extracted from Archer (1996) - additional data sets were implemented

from the data base Pangaea (www.pangaea.de) and from the literature. The distribution pattem was recalculated by the application of the kriging method using a linear semi­

variogram model without discretization.

Regionalization of the organic carbon content in surface sediments 29

-180 -120 -60 o 60 120 180

-180 -120 -60 o 60 120 180

o TOe observations

Figure 1: Global distribution of TOe sampling sites in surface sediments « 5 cm sediment depth). Sampling sites are indicated as gray dots.

Generally, we considered sediment data within a maximum sediment depth of 5 cm. Within this interval the main mineralization processes take place (Reimers, 1987; Lohse et aI., 1998; J0fgensen, 2000; Wenzhofer and Glud, 2002). Exceptions were made for 113 sites: 0-10 cm, Prakash Babu et aI. (1999), Arabian Sea data; 4 sites 0-15 cm, Smith (1987) eastern Pacific Ocean.

The digital global bathymetry was taken from the ETOP05 global topography dataset, which was compiled by the U.S. Naval Oceanographic Office NOAA, National Geophysical Data base, Boulder as listed in Tab. 1.

For a comparison of the sedimentary distribution patterns with the primary production, we used the global annual estimation after Antoine et aI. (1996). The data set was queried from the JGOFS-France database Pigments and primary production fi'ol11 satellite ocean colour (Tab. 1).

A compilation of the groups of data sources and classification criteria as riverine and eolian input, or advective sediment transport are summarized in table 1.

Table I: Summary of control parameters and data sources.

Parameters Data base Primary Production

(Antoine et aI., 1996)

Bathymetry (ETOP05)

Riverine discharge Slope instability Dust transport

Antarctic and Arctic ice­edge

Sediment surface data Corg, Opal, CaC03

Providing of data

JGOFS-France data base

NGDCi WDC MGG (National Geophysical Data ('enter/

IForId Data Centerfor marine Geology and Geophysics. BOlllder)

USGS ( US Geological Surl'c))

Antarctic Sea Ice Archive data Sea Ice Atlas for the Arctic Ocean (trom .Iuly2001 and January 2001)

PANGAEA (Pmjekte: ADEP[), BfGSET, SFB 261, .JGOFS, SINOPS)

Source (http://www.obs-vlfi·.fTijgofs2imodelisation/home-p.htm ).

www.ngdc.noaa.gov/mgg/globaliseltopo.html

Hillier (1995), SchlUnz et al. (1999) Stow and Mayall (2000)

http://www.ldeo.columbia.edui"-bhuber Environmental Working Grollp- Joint U.S. Russian Sea Ice Atlas for the Arctic Ocean (CD Version

1.0, I. September 2000) The national Snow & Ice Data Cetlter: Digital SAR Mosaic and Elevation Map of the Greenland lee

Sheet www.nsidc.org

www.pangaea.de

Romankevich E.A., Vetrov A.A. (2001) Vetrov A.A., Romankevich E.A., Benensol1 M.A. (1997)

(Shirshov Institute o(Oceano!o,t,'Y, Russ. Academ); q(Sciences, ;'yfOSCOl-I')

P. Muller, T. Wagner (Univ. Bremen) Dierk Hebbeln (Univ. Bremen)

Tim Jenneljahn (Univ. Bremen) Gesine Mollenhauer (Univ. Bremen)

w o

:::0

~. s: 2. hi· ~ o· ::;

.~ '" ~ Cl

~ ~

~. r; ~

~ Cl ::; r; Cl ::; (i) ~

~ ~ ~.

r; (\)

'" (\)

~ i\: ::; r;;-

Regionalization of the organic carbon content in surface sediments 31

2.2 Geostatistical methods

2.2.1 Semi-variogram modeling

F or the evaluation of benthic provinces and processing of TOe distribution patterns, we used

the kriging method. The data were treated as regionalized variables and the variance of the

investigated data was defined as functions of space in the sense of Matheron (1965) and de

Marsily, 1986).

The stationarity of the data is a basic condition for geostatistical, especially variogram

analyses. Therefore, all regionalized variables z(x) at all points x within a region are assumed

as realizations of a certain random function Z(x). All values fluctuate in space around a given

mean m (Eq. 1). According to Matheron (1965), the intrinsic hypothesis (stationarity of

second order) has to be fulfilled in praxis.

E[Z(x)] = m '\Ix (1)

E[Z(x) - Z(x + h)] = m(h) (2)

The stationarity of second order is given for the first two moments of the random function

(mean and variance). The expectance value E between the increments of two realizations Z(x)

and Z(x+h) depends on the distance vector h (Eq.2). The variance of the increments is

expressed by a semi-variogram. It reflects half of the mean of squared differences between

two observations as a function of the distance (Eq. 3) (lones, 1999).

1 l1(h) { }2 Y = -- L Z( x J - Z( Xi + h)

n(h) i~1

Z=Random function of all regionalized variables z(x)

E=Expectance value of Z(x)

n=number of pairs

y=semi -variance

(3)

Since the geostatistical condition of nonnal distribution was fulfilled for the logarithm of the

organic carbon content, we transformed all data values for subsequent semi-variogram

analyses and interpolation procedure (Fig. 2). For visualization and comparison with

measured data, they were retransfonned SUbsequently.

The ordinary-kriging procedure requires the modeling and adjustment of an experimental

semi-variogram for the detennination of the spatial structure. The semi-variogram model

mathematically specifies the spatial variability of the data set. The derived weights, which are

applied to sample points during the grid node calculations, are functions of the semi­

variogram models. They are used to estimate values at the unknown locations (Davis, 1973).

32 Regionalization of the organic carbon content in swface sediments

Estimated and measured values show the following general linear relation in ordinary kriging:

n

Z * (x o) = L::!cjz(xJ j

z*: estimated value at the point Xo

Ai : weighting factor, depending on distance and direction

z(xJ data value at the point Xi

(4)

(5)

The estimation of the semI-vanance y is based on data clusters around a celiain location

described by the distance vector h (h=(hx , hy)). Experimental semi-variograms are fitted by

idealized models of the semi-variograms (cf. sect. 3.1, Fig. 3). The idealized theoretical semi­

variogram models are defined as simple mathematical functions, which relate (y) to (h) with

the basic assumption that the semi-variance increases with increasing distance. In general, a

spatial cOlTelation is defined up to a characteristic threshold called sill, which reflects the total

variance of the system. The distance (~h) up to the sill is the range of the spatial dependency.

In our study, four types of semi-variogram models were adapted to the domain data sets. The

most common types are the spherical model (Eg. 6) and the exponential model (Eg. 7), which

both have linear slopes in the origin. Whereas the spherical model reaches the sill at the range,

the exponential semi-variogram approaches the sill asymptotically. For the calculation of the

weighting, the practical range (distance where the semi-variogram value is 95% of the sill) is

used. For continuous variances in the nearest neighbour, the gaussian model was applied

(Eg. 8). The range is defined at 95 % of the sill as well, but it is the only semi-variogram with

an inflection point near the origin.

Models for spatial analyses involve two spatial scales. The first represents the spatial

variability of the observed sample points in an investigation area, whereas the second

represents the short scale variability (non-spatial elTor tenn Co; Eg. 9). Co is the

neighbourhood variability, the nugget-effect, between two observations nearby. It describes

the natural variability and the accuracy of the measurement. In case of a nugget-effect the

asymptotical threshold is defined as sil1=scale+ Co. For areas without a regional dependency,

the nugget effect model is used to express the discontinuity in space, where the kriging

estimator is simply the arithmetic mean of the neighbouring values (Eg. 9).

Regionalizatiol1 o/the organic carbon content in s1Ilface sedimel1ts 33

The theoretical seml-vanances are calculated by the following equations after Pannatier

(1996):

lc. [1.5~ -0.5[~)3] y(h) = a a

c

spherical

{ [

31h

l]

y(h) = c· 1- e a exponential

{I 3(1hl)' l

y(h) ~ C f -e (·f gausslan

nugget

c=sill value, or in the case of a nugget-effect in the origin the scale

a=range

h=distance vector

2.2.2 Direction controlled modeling

If 0< h < a (6)

otherwise

If 0< h < a (7)

If 0< h (8)

Ifh = 0 (9)

otherwise

Natural processes may have prefeITed directional characteristics in the investigated areas. In

case of directional spatial conelations as processes in prefened directions, we specified an

angular window with a direction tolerance (Udir). If necessary, we defined an anisotropy ratio,

which considers that observations lying further apaIi show a higher regional dependency in a

prefened direction. The relative weighting in the grid node procedure is defined by this

anisotropy ratio. The maximum range is given by the eigenvector in the prefeITed direction.

Thus, the semi-variogram is a three dimensional function with two independent variables

(direction and distance, h) and the dependent variable (semi-variance, y). In accordance with

the standard notation the direction is 0° along the abscise, and 90° along the ordinate.

2.2.3 Interpolation procedures

Interpolation procedures were carried out by applying the modeled semi-variograms 111 a

variable sector search. For each procedure, the search area was limited individually by

subdivision in 1-4 sectors. The number of sectors was chosen by the data distribution and the

geometry of the area. For example, a narrow area along the coast, with sampling sites

perpendicular to the coast was kriged by applying an anisotropy factor as well as a 2 sector

search with a bisecting line perpendicular to the coast. The arrangement of the sectors was

defined by the angle of the search ellipse. In case of clustered data, the four-sector search was

applied. For evenly distributed data, this separation was inelevant.

34 Regionalization of the organic carbon content in surface sediments

Although the weighting and thus the influence on grid node procedure for observations further

apart is well limited through the semi-variograms itself the search radii were set to the range

or the double range to avoid blanked areas. If no semi-variogram of equation 6-8 could be

modeled, the spatial extension of the investigation area and the available data were chosen as

the limiting criterion. A minimum of 2 data points per sector was obligate.

The significance of a map is expressed by the standard elTor of the regionalized area. In

kriging procedures the unbiasedness is guaranteed, if the coefficients are solved with a

minimal elTor variance (minimum square error, ( 2), which signifies kriging as an exact

interpolation method.

The difference between an estimated value Z*(x) and the true value Z(x) will be zero if the

hypothesis of stationarity holds true and no trend is assumed (Eq. 10).

E[Z * (x) - Z(x)] = 0 (l0)

In our study the calculation and modeling of the semi-variograms was done with the computer softwares Variowin@ 2.2 (Pmmatier, 1996) and Surfer@ 8.0 (Golden Software,

Inc., V8.0, 2002). For the kriging procedure and the grid processing the variogram models

were transformed and adopted by Surfer. FUliher calculations, overlays of different data sets

and merging of individual grids were done by the Geographical Information System Arc

View@ 3.2a (Esri, 2001).

2.2.4 Projection

Normally, kriging and semi-variogram analyses are used in areas, where the earth's curvature

has not to be taken into account. When projecting a global map, image, or other data, there are

certain tradeoffs. There is no single projection that best suits all purposes and distOliion

increases with increasing distance from the center of projection. For cOlTecting the distOliion

to the higher latitudes (6S0-8SoS; 6so-83.S0N), coordinates were recalculated to an equidistant

azimuthal projection with the projection center in the middle of the investigation area before

applying semi-variogram analyses and gridding. Finally, the resulting data grids where

reprojected into a cartesian coordinate system. The influence of distortion within 6soN and

6SoS for the semi-variogram analyses was tested, but is negligible. For the reprojection the

missing cells were resampled by using bilinear interpolation along the latitudes. The

resampling of the cells by the given new projection was done by bilinear interpolation of the

cell numbers.

3. Results and Discussion

The data density is much higher 111 the coastal zone compared to the wide basin areas.

Summing-up all basin areas versus all coastal areas a data density ratio of liS is given. Thus,

80 % of the TOe sites are distributed on ~ 12 % of the global ocean area (shelves and slopes,

< 4000 m water depth), whereas 88 % (basins, > 4000 m water depth) of the global ocean area

is covered by only 20 % of the data.

Regionalization of the organic carbon content ill sw/ace sedill1ents 35

The highest TOC contents are observed in the Atlantic Ocean along the West African continental margin, especially off Namibia, offshore the Congo mouth and along the West African margin. In the Pacific Ocean, TOC enriched areas are located along the continental margin of Peru and Chile and offshore Califomia. In the Indic, especially in the Arabian Sea, the sediments are highly enriched with TOC offshore Oman and India. The average TOC content is 0.5 wt% in the deep ocean and 2 wt% along the eastem margins. Whereas coastal data range between 0.05 - 21.2 wt% with a median of 1.5 wt%, basin data range between 0.01 - 4.8 wt% with a median of 0.4 wt%. Thus, in opposite to the TOC value distribution, which is significantly skewed, the distributions of the log-nonnal values are symmetric (Fig. 2a,b). The separate analysis of basin and west-coast data show conspicuous positive (Fig. 2c) and negative shifts (Fig. 2d).

Toe [\\1°,-;' 1 0.1 0.5 1 5 10

i I

f 1200 I

a I

1000 I

800 600 400 200 global area

0 1200 I

'l1 1.5 -1.0 -0.5 10.0 0.5 1.0 1.5 v 1000 I -' I or;;

b I

bf) 800 :=

600 ~ := 400 a global coastal areas if.' 200 U

0 ~ 600 0 500 >. C () 400 := g 300 er u 200 t.!::::

westcoasts (high productivity areas) 100 0

600 1.5 -1.0 -0.5 : 0.0 0.5 1.0 1.5 500 I

d I

400 I

300 200 100 global basin areas

0 -1.5 -1.0 -0.5 00 0.5 1.0 1.5

log TOe

Figure 2: Log-nom1al distribution of TOe sampling sites. Frequency and distribution on a global scale (a) for the global area, (b) for all coastal areas « 4000 m water depth), (c) for high productivity areas along the west coasts of the continental margins, and (d) basin areas (> 4000 m water depth). The dashed line marks the mean=median value of the global log-TOe value distribution. The arrow marks the appropriate mean TOe value, calculated from the non-transformed global TOe data compilation.

36 Regionalization of the organic carbon content in slII/ace sediments

3.1 Applicability of sel11i-variogram modeling over large distances

In a first attempt, spatial dependencies are investigated for two global zones, east and west of

the zero meridian [65°N, 65°S]. The isotropic semi-variograms of the log-normal distributed

surface data of the TOe content indicate a nugget-effect of 0.06 (East), respectively 0.l3

(West) Fig. 3a,b). Nevertheless, neither a significant range, nor a specific semi-variogram

model can be defined.

o.so < ~ ~

0.45 a v

0.40 g .~

0.35 l-; ;:::: v"

,!,. 0.30 a v

0.25 'f.i

0.20

0.15

0.10

O.OS

0.00 0

• • range: 2~o • •

•

•••

• •

r::mge: ?

•• ••• • • I. • " .... • .......... t. • ••••• ...

• •

IO 20 30 40 50 60 70 0 10 20 30 40 SO 60 70

Figure 3: Semi-variograms oflog-transformed TOe values (a) of the eastern and (b) of the westem hemisphere [65°N, 65°S]. Distance is given in [0] and is based on a 1.4° cluster for the log-transformed TOe values.

East of the zero meridian spatial dependencies up to 6.6°, 28° and 32° with variable sills can

be interpreted by different experimental semi-variograms. However, the increase up to the sill

value is not strictly monotonic (Fig. 3a). Hole effects, probably induced by the mainland or

islands and local underlying trends are responsible for the fluctuation around the sill and the

non-continuity in space. Thus, the variable TOe is assumed as unsteady. However, west of

the zero meridian, a spatial cOlTelation can be found up to a range of approximately 10°

(Fig. 3b). The progression of the variance function around the sill is more continuous due to

the large number of observations with a TOe content below 0.4 wt% within large distances

(appr. 40 %). It is evident that it is not possible to model a semi-variogram, which reflects the

spatial correlation sufficiently in these areas without further discretization. Additionally, data

points without any relations to spatial dependencies affect the grid calculation, which may

result in a more smoothed grid surface. Subdividing the global ocean into smaller benthic

provinces should result in semi-variograms with lower and more exact ranges, and individual

modeling parameters. Accordingly, further regionalization, depending on defined spatial

characteristics, will increase the accuracy of estimations.

Regionalization of the organic carbon content in slIrface sediments 37

3.2 Definition of benthic TOe-based regional provinces

The regionalization of the Toe content in surface sediments was calTied out by combined

qualitative and quantitative-geostatistical analyses of the spatial dependencies of the data

points. The qualitative analysis is used for the principle regionalization, and is based on main

oceanographic and sedimentary characteristics (=pre-provinces). The quantitative analysis is

the fine-tuning of the regionalization procedure, done by modeled semi-variograms of the

kriging method (=provinces). Thus, to find best boundaries of spatial cOlTelations and finally

separate the adjacent areas, semi-variogram models were fitted for each potential province.

The procedure is summarized in principal in figure 4.

Longllll.fst PrO\!1I1CtS-

Surface (Urrentf;;,

Primary productIOn

j"atentl tranqpOlt Riverine·!eolian inp\~

Bottom cmlentc:

Dowlldope process""

Qualitative Analysis ofTOC-data

Pelag,ic and continental

"

~ •.. ,. / DifferentmtiOn by ,,-at er depth ,/ /,''---~-----_h .. ",,,-,,_ J;l and (li8t~1(e fi:om the ''', contUl.e:nt

/Pl'e-Pl'O'illce~. \ Continentat Pelagial i i I

~-'-, I / ........... '--..~--~-·lh........-"'-~·

i !

~ Quantitative Analysis

..-----.····----1 Seilli-vru:io2.rrull ,.l,.milyrJ1')

Pl'Olinces

Sediment compo.,itioll

Opal

Calcite

Figure 4: Principle scheme of regionalization procedure as the qualitative approach with main influencing parameters marked as arrows. The quantitative analysis is done by semi-variogram modeling in each province (fine-tuning).

F or the general proVInce classification, which is based on the oceanographic surface

circulation (Fig. 5) and its effect on regional primary productivity biomes, we made a first

subdivision by using the scheme after Longhurst et aL (1995) and Sathyendranath et aL

(1995). A second classification was made by a simple separation between pelagic and

continental margin zones, because significant regional variabilities in expOli production are

only apparent down to certain water depths; in general, this limit is located at 4000 m water

38 Regionalizatiol1 0/ the organic carbon content il1 s1lr/ace sediment,

depth. The Antarctic-province was separated by the position of the annually averaged ice

coverage (Tab. 1). The northern higher latitudes were differentiated in their sub-seas as the

northern north Atlantic provinces (Greenland-Norwegian-Iceland Seas=GROE, Bm"ent and

Kara Sea=BARENTKARA) and the northern Kara Sea (KARA2). The polar regions,

permanently covered by ice, were excluded (Tab. 1). The province abbreviations used for pre­

selected and final benthic provinces are summarized in table 2.

Table 2: Abbreviations ofbenthic TOC-based regional provinces.

Abbreviation

BARENTKARA

KARA2

LAPTEVSEA

GROE

ANT

SOATL

NAMBCO

SWACO

ARGCO

RJOPLATA

BRAZCO

C;UBRACO

CUI

ETROPAT

NOATL

NEAMCO

WAFCO

CANAR

EURI

EUR2

NWAMCO

CHICO

NHl-/NEPAC

TROPAC/TROPAC2

SW-/SEPAC

PERCO

IND

EARAB

WARAB

SEAFCO

SOMALlCO

TANZACO

EfCO

Province description

Barents Sea and Kara Sea

Kara Sea

Laptev Sea

northern Nordic Sea

southern Polar Sea

southern Atlantic deep sea

cant. marg. off Namibia

cant. marg. off SW-Africa

cant. marg. off Argentina

Rio de la Plata mouth

cont. marg. off Brazil

cant. marg. off SE-America (e.g. Guyana)

Gulf of Guinea

eastern tropical Atlantic

northern Atlantic deep sea

cont. margin off NE-America

cont. margin offW-Africa

Canaries

N-European cant. margin

S-European cant. margin

cant. margin of NW-America

cant. marg. off Chile

NW-/NE-Pacific deep sea

eastern tropical Pacific/western tropical Pacific

SW-iSE-Pacific deep sea

cont. marg. off Peru

Indian Ocean deep sea

eastern Arabian Seat

western Arabian Sea

cont. margin offSE-Afi"ica

cont. marg. oft'Somalia

cont. marg. off Tanzania

cont. marg. off East India

East o/::.ero meridian Global area east of the zero meridian [65°N-65°S]

West o/zero meridian Global area west of the zero meridian [65°N-65°S]

Regionalization of the organic carbon content in sw/ace sedil71ents

60

o

-60-

-180 -120 -60 o

!\i Fquatortal

. n·· ____ ~··ll~I~I!~)~~<~I.~~:!jz\~:~ .. '\. Lqo:.llorl<ll

-180 -120 -60 o Riverine discharge of suspended sediment [Mt/yr]

o 10 0 50 0 100

o 250 0 500 0 1000 0 2000 0 3000

Riverine discharge of particulate organic matter [Mt/yr ]

<5 5-10 10-20 20-30 30-40

60 120 180

60

o

--60

60 120 180

~ Giant landslides

D Hydro-carbon exploration frontiers as slope instability zones

>40

Orientation of eolian dust input from the hinterland -----+ Surface currents ABF: Angola Benguela Front

39

Figure 5: Global distribution of major submarine landslides on continental margins and hydro-carbon exploration areas (modified after Stow and Mayall, (2000) adapted from Mienert et a1. (2003), discharge and orientation of riverine and eolian input of particulate matter after Schliinz et a1. (1999) and Hillier (1995), and principle surface currents.

In general, the accumulation of organic matter is correlated with the accumulation of other biogenic parameters as calcite or opal (e.g. Henrichs, 1992; Jahnke, 1996; Tyson, 2001). Based on the large differences in calcite contents, the nOlih-eastern Pacific Ocean and north­western Pacific Ocean (> ~lOoN, NE-INW-PAC) can be separated from the equatorial and southern pati (Fig. 6a; TROP AC, SEP AC). The entire southern and nOlihern Atlantic and the Indian Ocean are characterized by high calcite contents in surface sediments with maximum values for the Mid-Atlantic Ridge (> 90 wt%). The coastal area off the Rio de la Plata (RIOPLATA) could be differentiated from the adjacent areas off Argentina (ARGCO) and Brazil (BRAZCO) by its low calcite content (Fig. 6a). The global opal distribution is more heterogeneous and patchy (Fig. 6b). The coastal area off Peru (PERCO), the equatorial tropical Pacific Ocean (TROPAC), the equatorial eastern S-Atlantic upwelling area (ETROPAT) and the zonal belt along the Antarctic Circumpolar Current are characterized by diatom dominated phytoplankton growth and thus opal-rich sediments as shown in figure 6b (> 10 wt%; DeMaster, 1981; Treguer et aI., 1995; Ragueneau et aI., 2000). The entire deep­sea basin area of the northern Atlantic Ocean (NOATL, > 4000m water depth) shows an average opal content in surface sediments, queried from the grid and shown in figure 6b of approximately 3.5 wt%. The average sedimentary opal content of the deep-sea basin area of the Indian Ocean (IND) is approximately 15 wt% (Tab. 3).

40 Regionalization of the organic carbon content in swface sediments

-1S0 -120 -60 o 60 120 1S0 i , i

-1S0 -120 -60 o 60 120 1S0

Calcite [wt%]

<10 20 30 40 50 60 70 so 90 > 90

-1S0 -120 -60 o 60 120 180

60

-180 -120 -60 o 120 180

Opal [wt%]

<2.5 5 10 15 20 25 30 35 40 45 50 55 60 65 > 70

Figure 6: (a) Gridded map of calcite contents and (b) opal contents in surface sediments in [wt%].

Table 3: Zonal statistics of opal, calcite and TOe contents in surface sediments and primary production within the benthic provinces, extracted as the mean and max values ofthe regular grids and standard deviation (SD). Sources are summarized in table. I.

Provinces

GROE

KARA2

BARENTKARA

LAPTEV SEA

NWAMCO

CHICO

PERCO

TROPAC (east)

TROPAC2 (west)

SEPAC

SWPAC

NEPAC

NWPAC

NEAMCO

GUBRACO

BRAZeO

RIOPLATA

ARGeO

EUR2

EURI

NOATL

opal opal (max) (mean) [wt%] [wt%]

20.0

16.0

12.0

43.2

73.0

35.0

67.1

30.8

18.7

33.0

5.8

21.0

19_6

15.6

23.5

11.2

4.9

28.9

, " _._)

4.6

5.0

20.2

16.8

7.9

7.7

6.1

3.4

6.7

3.4

1.8

3.6

4.8

2.6

0.9

0.9

3.5

SO opal

[wt%]

2.0

2.9

3.3

13.1

9.0

4.6

10.3

4.7

2.4

7.1

1.3

2.2

3.0

3.6

3.1

1.4

0.6

3.1

calcite (max) [wt%]

62.3

46.4

58.6

10.9

96.6

92.5

54.4

92.8

88.9

92.3

66.8

75.4

83.6

12.9

68.6

85.8

47.8

93.9

calcite (mean) [wt%]

19.2

6.0

33.4

1.2

42.3

40.0

57.3

52.7

2.2

5.9

32.9

48.7

45.5

2.5

30.0

24.9

52.6

43.6

SO calcite [wt%]

14.6

8.2

16.2

1.9

30.9

26.2

.10.3 )" ., _J •. )

6.5

13.0

19.5

21.3

22.3

2.3

16.4

19.7

12.1

24.2

Toe (max) [wt%]

3

1.8

2.7

4.2

6.1

3.9

14.1

4.9

1.7

3.1

1.3

1.8

1.9

6.2

loO

1.1

3.4

0.96

0.83

1.9

5.7

Toe (mean) [wt'%]

0.7

1.2

1.1

0.95

1.6

1.4

4.9

1.2

0.8

0.49

0.8

0.44

0.6

0.9

0.4

0.5

0.8

0.3

0.3

0.8

0.4

SD TOe [wt%]

0.32

0.3

0.5

0.3

0.9

0.7

2.8

0.7

0.3

0.25

0.25

0.2

0.3

L1

0.1

0.2

0.6

0.2

0.2

0.4

0.2

ppl ppl SO (max) (mean) ppl

[g m-2 y(l] [g m-2 y(l] [g m-2 y(l]

448

6[9

510

629

781

770

379

534

342

897

[264

1280

381

12.10

557

626

508

605

112

183

196

225

129

1 [3

79

106

90

98

184

193

128

178

140

174

[46

110

52

78

59

78

28

30

27

28

22

25

107

182

37

104

63

45

55

", J_

>J

'" [Jq

o· ~ t::j' ~ o· :::s

~ ~ Cl

rKl !:::l

~.

" ~ Cl :::s

" Cl :::s

~ :::s ~

~ g Cl)

'"' Cl)

~ _. :::s ~ c;;-

~ >-'

~

Table 3: continued I N

opal opal SD calcite calcite SD TOC TOC SD ppl ppl SD Provinces (max) (mean) opal (max) (mean) calcite (max) (mean) TOC (max) (mean) ppl

[wt%] [wt%] [wt%] [wt%] [wt%] [wt%] [wt%] [\vl%] [wl%] [gm·2 y(l] [gm-2y(l] [gm-2 y(IJ

WAFCO 27.9 6.1 5.1 78.5 41.3 19.0 1.9 0.65 0.4 1262 302 195

CANAR 5.5 2.4 0.9 47.0 57.3 9.9 1.5 0.7 0.2 761 150 97

GUI 22.8 6.8 4.6 46.0 14.3 11.4 2.1 1.[ 0.5 819 189 63

ETROPAT 27.9 12.8 5.4 95.8 47.9 26.5 3.5 0.7 0.5 1159 182 75

SWAC02 41.1 6.8 6.3 93.1 38.9 26.9 8.2 1.5 1.2 1159 215 103

NAMBCO 41.1 3.7 4.4 85.2 67.3 16.9 8.2 2.7 2.0 562 256 92

SOATL 89.4 13.2 14.9 105.7 34.1 34.2 1.7 0.4 0.15 357 [01 36 >::I ~

WARAB 18.2 5.9 3.5 67.9 47.2 14.6 5.5 1.5 0.75 1163 266 95 C:;. g

EARAB 35.6 10.1 5.2 80.3 50.2 17.2 3.2 1.2 0.7 [489 299 2[3 ~ ~

SEAFCO 18.6 4.1 1.9 54.4 57.3 10.3 2.2 0.5 0.3 722 124 39 g' EICO 24.7 14.7 4.9 51.9 31.4 15.6 1.5 1.0 0.3 1045 313 208 ~ .....

;::,-

TND3 85.9 [5.6 14.9 97.0 45.8 30.5 1.9 0.4 0.17 954 97 43 '" Cl

ANT 41.3 8.3 8.3 68.9 5.9 7.5 1.45 0.12 0.2 558 75 64 ~ I:l ~

I: primary production alter Antoine et a!. (1996). ;:;. <:'l

2 : including NAMBCO. 2 3: including TANZACO, SOMALICO. c::J..

~ <:'l Cl ~ (\l ~ -. ~

f:! ~ <:'l

'" Co

'" g, ~ Co

Regionalization of the organic carbon content in slIIface sediments 43

Whereas the calcite and opal distribution patterns were mainly used for discretizating the pelagic from the coastal areas, the latter are strongly affected by the additional supply of tenigenous organic matter by riverine discharge (Fig. 5; Hillier, 1995; Sch1unz et aI., 1999). The most strongly affected provinces (GUBRACO, SW ACO, RIOPLATA, TROPAC) are located in the tropical areas, where erosion and sediment transpOli is high.

An additional criterion for the selection of provinces is the existence of submarine landslides along the continental margins (cf. Fig. 5; area> 1000 km2

). Landslides may be triggered, i.e. by seamount erosion (Ranero and von Huene, 2000) or by the dissociation of gas hydrates on the continental slopes (Stow and Mayall, 2000). Major gas hydrate exploration frontiers and downslope areas are located along the continental margin off Argentina (ARGCO), offshore the NE-American continental margin (NEAMCO) and off Namibia (NAMBCO), western­and eastern Arabian Sea (W ARAB, EARAB) and east of Greenland (GROE). Lateral advection of sediment through contour currents is well known along the SW-African Coast (5°N, 35°S), especially the Benguela region (NAMBCO; Walsh, 1991; Uenzelmann-Neben, 1998; Zabel and Schulz, 2001; Mollenhauer et al. 2002), the coast of West Africa, (CANAR, WAFCO; Weaver et aI., 2000) and along the continental margin off Argentina (ARGCO; Hensen et aI., 1998,2003).

For each pre-selected province specific seml-vanograms with changing propeliies were modeled in order to test existing regional dependencies. Borderlines between existing pre­provinces were shifted as long as the specific data sets reveal the best possible semi­variogram. Qualitative and quantitative analysis resulted in the definition of 33 benthic provinces (Fig. 7). However, semi-variograms are not defined for each potential pre-selected province, because of the limited availability of data. Out of 33 benthic provinces, 9 provinces had to be defined by the qualitative analysis only. The corresponding semi-variogram properties are summarized in Appendix 1. The semi-variograms of the BRAZCO-, GUI-, WAFCO-, SEAFCO and ANT- provinces did not reflect a spatial dependency of the TOC data (nugget effect model), thus we assume that existing global relations occur most likely on smaller scales. Data densities in these areas are, however, too low for further statistical analyses. For the SEAFCO-, EICO-, SE-/SWPAC-, SOMALICO- and TANZACO province no semi-variograms could be defined.

Despite the fact that the good approximations by the semi-variograms indicate strong spatial conelations within the defined 24 provinces, fmiher data, especially located at continental influenced provinces will celiainly refine the basic pattern ofbenthic provinces.

Oppositely, we assume that an increased data base would not lead to a significant improvement in pelagic areas (SEPAC/SWPAC, IND and SOATL, NOATL) due to the low spatial heterogeneity of mineralization processes.

44 Regionalization of the organic carbon content in swface sediments

0- o

-60-

I I

-180 -120 -60 o 60 120 180

Figure 7: Discretization in 33 TOe-based regional zones.

In Tab. 3, the zonal statistics in the different provinces are summarized. For each province the mean values of opal, calcite, TOC and primary production on the base of a projected 10 x 10

grid resolution was queried. The TOC [wt%] distribution pattern is gridded as described in Chapter 3.4.

In the following section, some representative regions of the Atlantic Ocean are discussed in order to clarify the process of semi-variogram modeling and parameter adjustment. The specific semi-variogram settings and kriging conditions can be derived from Appendix 1.

3.3 Case studies

The eastern South Atlantic

The comparison of three adjacent provinces (SOATL, SWACO, NAMBCO) of the southern Atlantic Ocean reflects the high regional variability of the data values. In figure 8 (a,b) semi­variograms of the total area are compared with those of separated provinces. One of the most obvious and striking observation is that the nugget-effect of the omnidirectional semi­variogram for the entire southern Atlantic Ocean area accounts for as much as 25 % of the total variance (0.2). This high nugget-effect (=increase of the local variance) is due to the sharp transition with strong gradients from basin to coastal areas. The discretization of coastal and basin areas minimizes the overall nugget-effect of the total variance and the total variance itself of the data set (scale + nugget-effect, Appendix. I; Fig. 8a,b). The widest regional dependency can be observed for the basin area of the southern Atlantic Ocean (SOATL) with 8.6 0 (258 observations). The low variability of the data values, the homogeneity and continuity of the processes in the basin area is clearly reflected by the low total variance of 0.065 and a nugget-effect of only15 % of the total variance (Eq. 11-11.3).

Regionalization of the organic carbon content ill swjace sediments

Southem Atlantic Ocean (total area):

h h 3

r(h) = 0.05 + 0.15[1.5· - - 0.5 ._] 8.55 8.55

Coast off Namibia (NAMBCO):

h h 3

y(h) = 0.005 + 0.125[ 1.5· - - 0.5·-] 3.66 3.66

SW- African coast (SWACO):

h h 3

y(h) = 0.02 + 0.1 0[l.5 . - - 0.5·-] 5 5

S- Atlantic basin area (SOATL):

~ r(h) = 0.01 + 0.055[1- e 8.6 ]

h: distance, [0]

y(h): semi-variance

0.30 --:-SOATL

,.=: "-

0.25 ~. ~

U :J ~ e:; .- 0.20 ~ >

I .-S 1.> 0.15 C/).

0.10

0.05

a

0.00 4---.--.----,.----,.--.----,------,--.----.---.

o 2 6 8 10

I I I

"

/

/ /'

// ,/

~/ /'

(lI)

45

(11)

(1l.1)

(11.2)

(1l.3)

b

Figure 8: (a) Comparison of experimental and theoretical semi-variograms of the southern Atlantic areas: Southern Atlantic Ocean and adjacent coastal zones, coast off Namibia, SW- African coast, basin areas. (b) Fitted theoretical semi-variogram models of the southern Atlantic areas: Southern Atlantic Ocean and adjacent coastal zones (1); coast off Namibia, NAMBCO (H); SW- African coast, SWACO with semi-variogram direction of 119° (Ill) and basin areas, SOATL (IV).

Along the West African continental margin, from the Bight of Biafra to the Cape Agulhas (SON - 400 S) we defined the coastal zone, SWACO, with 285 observations. The boundary to the basin area is defined by the water depth with less than 5000 m north of Walvis Bight and

46 Regiol1alizatiol1 of the organic carbon content in sw/ace sediments

less than 4000 m south of the Walvis Bight. The range of the directional semi-variogram (dir. 119°) with an angular window of 45° shows a maximum regional dependency along the main axis of 5°. The SWACO province is a generic province, including different intersecting sub-regions. Although we modeled an isotropic semi-variogram with a prefened direction along the coast, we did not apply an anisotropy-ratio in the higing procedure. River induced upwelling and downslope transport of sediments (e.g. Walsh, 1991; Uenzelmann-Neben, 1998; Zabel and Schulz., 2001; Holtvoeth et aI., 2001) - with the prefened direction perpendicular to the coast - characterizes the area around the Congo mouth. Moreover, the SW ACO province and the ETROP AT province overlap, making the SW ACO province also pati of the equatorial upwelling. The ETROP AT province is characterized by seasonally varying diatom blooms and thus, opal enrichment in surface sediments (ETROP AT, 170 observations, Tab. 3).

By refined semi-variogram mode ling the sub-province NAMBCO could be separated from SWACO. The semi-variograms of both provinces are similar, because of the high data density in the NAMBCO province. Regional dependencies are comparable and could be defined between 4° to 5°. Nevertheless, the high productivity area off Namibia and the nOlihern part of South Africa (NAMBCO) between the Angola Benguela Front (ABF) and Ltideritz (16°S -28°S) had to be treated separately. Whereas the tenigeneous input of organic matter is insignificant, lateral sediment transpOli processes play an important role. Thus, two main processes control the distribution pattern of TOC in surface sediments offshore Namibia: the high productivity coupled to surface currents along the continental margin and the lateral transport of sediment by strong bottom currents. The ABF separates the Angolan from the northern Benguela ecosystem (ABF, ~14°S-17°S; Shannon and Nelson, 1996; Meeuwis and Lutjehenns, 1990). The prefened isotropy direction for the NAMBCO province is given by the coastal Benguela cunent. The strong spatial dependency of the TOC surface data - with a main anisotropy axis along the continental margin - is reflected by a high sill of a steady exponential model and a very low nugget-effect. The semi-variogram model with an anisotropy ratio of 3 (Apppendix I; angle 135°) is fitted with a range of 3.6° along the main axis, using a lag spacing of 0.2°. Applying the anisotropy ratio along the continental margin, makes it possible to exclude misinterpretations caused by the sampling strategy, e.g. sampling along transects perpendicular to the coast. Without considering this self-produced anisotropy perpendicular to the coast, the contour lines would cluster around the data points. In figure 8a,b the fitted models with the typical sequence of the low variability of the TOC content for SOATL are compared to the high variability in the NAMBCO regions with strong regional dependencies.

The western South Atlantic

The western part of the southern Atlantic Ocean is not affected by coastal upwelling but partly the coastal regime is characterized by high productivity, e.g. the confluence of the cold, nutrient rich Malvinas Current and the wann, nutrient poor Brazil Cunent. The confluence zone is located off Rio de la Plata (~39°S). In the confluence zone complex frontal motions and mixing patterns of warm and cold water masses are formed. In comparison to the open ocean (~101 gC m-2 yr- 1

), phytoplankton growth is enhanced up to 1200 gC m-2yr-1 (Tab. 3)

Regiona/izatiol1 of the organic carbon content in surface sediments 47

Antoine et aI., 1996; Gordon, 1989; Peterson and Stramma, 1991). The interaction of the

Brazil Current and the Malvinas Current causes the differentiation of the ARGCO-zone

offshore Argentina and Uruguay and the BRAZCO-province offshore Brazil in front of the

Rio de la Plata. The transition area is defined as the RIOPLATA-province.

The ARGCO province can be fitted by a spherical semi-variogram model, with a low nugget­

effect of 8 % of the total variance and a range of 10° (Fig. 9a; Eq. 12). Observations from

water depths > 4000 m are attributed to the SOA TL province. The Malvinas Current

determines the direction of the variogram with 60° and an angular window of 45° without a

prefened anisotropy weighting along the continental margin. Since the ARGCO-zone is a

complex zone one simple anisotropy direction cannot be specified. The ARGCO area is

affected by strong surface (Malvinas Cunent) and bottom water flows, resulting in highly

dynamic sedimentary processes and gravity controlled mass-flows (Garzoli, 1993; Peterson et

aI., 1996; Hensen et aI., 2000; Hensen et aI., 2003). Strong currents extending to the sea floor

reduce or even prevent the accumulation of fine-grained predominantly terrigenous sediments

on the shelf and upper-slope. The major part of the riverine discharge of the Rio de la Plata or

tributaries is not accumulated on the delta but transported far into the basin. Below 4000 m,

the Antarctic Bottom Water (AABW) fonns a strong contour current along the Argentine

continental margin, carrying along fine-grained sediments into the central basin (Ewing et aI.,

1964, Garzoli, 1993; Peterson et aI., 1996; Hensen et aI., 2000). The spherical semi-variogram

of the RIOPLATA-zone (50 observations; Fig. 9b, Eq. 12.1) as a possible sub-region shows a

lower absolute variance of 0.15 within a lower range of 2° than the whole ARGCO-zone. The

semi-variogram is modeled with a directional view of the semi-variogram of 153°

perpendicular to the coast and an angular window of 60°.

0,30

~

f§ 0,25 C1J

.;2 f.3 f (),20

B V>

0,15

O]()

(}05

Coast of Argentina

•• ..

(),OO -f--...--,---,--.,.-.,..--,---.-.,---...----.

o 2 10

Distance, h [0]

Rio de la Piata

• • • • • • h

Figure 9: Experimental and theoretical semi-variogram analysis of (a) the Argentine zone (ARGCO) and (b) the sub-region RIOPLATA.

A completely different situation is reflected by the semi-variogram of the area off Brazil. The

BRAZCO province is strongly affected by discharge of several rivers, which transpOli large

48 Regionalizatiol1 of the organic carbon content in slIrface sedil71ents

amounts of sediments as well as organic matter from the semi-arid tropical Hinterland (Tintelnot, 1995). The variable input and sedimentary regime is expressed by the high

variability of the semi-variogram fitted as a nugget effect model. Thus, no regional

dependency could be defined by the semi-variogram.

Coast of Argentina:

h h 3

y(h) = 0.02 + 0.23[1.5· - - 0.5·-] 10 10

Rio de la Plata mouth:

h h 3

y(h) = 0.15[1.5· - - 0.5·-] 2 2

h: distance, [oJ

y(h): semi-variance

The northern North Atlantic

(12)

(12.1)

For the northern North Atlantic, east of Greenland, a semi-variogram was modeled by the use

of the projected data (Fig. 10; Eq. 13). For the semi-variogram in figure 10-1 all observations

were considered. The high scatter of the data is expressed by the high nugget-effect (~50 % of the sill). The semi-variogram plotted without data < 550 m water depth (Fig. 10-H) shows a

continuous shape and a low nugget effect of 18 % of the total variance (0.009).

0.10

"-' ~ 0.08 8

fa ';'

.;::

& 0.0()

0.04

0.02 )

0.0 8.0:-.:10'

Distance. h [111]

Figure 10: Experimental and theoretical semi-variogram of the Greenland Sea, including all data (1) and excluding water depths < 550 m (ll).

Regionalization of the organic carbon content in sw/ace sediments 49

The general flow directions of the East Greenland Current and the Norwegian Current (West Spitzbergen Current) were considered by a directional semi-variogram analysis (observation angle 85°, angular window 65°). The East Greenland Current (EGC) flows southwards along

the eastern coast of Greenland from Fram Strait (79°N) to Cape Farewell (600 N) (Woodgate et aI., 1999). The EGC carries sea ice - associated with detritus (Hebbeln, 2000) - and polar water from the Arctic Ocean through the Fram Strait. The warmer wedge-shaped Norwegian Coastal Current flows northward along the west coast of Norway (Ikeda et aI., 1989). At the central Norwegian shelf, the bottom topography is complex involving shallow banks separated by deep trenches. The complex hydrographical and sedimentary system (erosion by undercurrents along the continental slopes, lateral drift, and ice-drifted transport of organic matter) leads to a generally high spatial variability of observations. Thus, we presume that the discretization in a western and eastern sub-region would considerably improve the semi­variogram modeling. Nevertheless, the low data density did not allow a further discretization.

Greenland-Iceland-Norwegian Sea:

h h3

y(h) = 0.009 + 0.05[1.5 . - 0.5· ] 750000 750000

h: distance, [m]

y(h): semi-variance

3.4 Processing the Toe distribution pattern on a global scale

(13)

The global total organic carbon distribution in a 1 ° x 1 ° grid resolution was generated by the application of the kriging method for each benthic province as defined above (Fig. 11). Interpolation was done separately for each of the provinces [83.5°N, 85°S]. Continents were set as borderlines to avoid the expansion of the interpolated grid. To create a global map, the basin areas were gridded using additional data of the adjacent coastal provinces and combined subsequently by overlapping corresponding zones. In the overlapping area the transition is calculated by the arithmetic mean, which slightly smoothes the grid. However, the error is

small since overlaps are restricted to areas with low concentration gradients.

50 Regionalization o/the organic carbon content in swface sediments

-60

-0

, , , -180 -120 -60 o 60 120 180

TOe [wt"Io]

<0.25 0.5-0.75 1-1.25 1.5-2 2.5-3 3.54 4.5-5 5.5-6 6.5-7 > 7

Figure 11: Global distribution pattern of the total organic carbon content in surface sediments « 5 cm sediment depth).

During grid node procedures the settings of the semi-variogram, especially the ranges and the

direction of the semi-variograms are the spatially limiting factor. The choice of the number of

data points within one sector was defined by the analysed spatial correlations of the semi­

variograms and is higher in the basin areas (4-6 observations) and lower in the coastal areas

(2-4 observations). The search radius was set to the range or the double range of the semi­variogram, to avoid blanked areas in the case of an exponential and gaussian semi-variogram.

However, those TOC distribution patterns, which were defined by the qualitative analysis

only, were gridded with an assumed linear semi-variogram model (nugget effect model: BRAZCO, OUI, WAFCO, ANT). The TOC distribution pattern of the pre-selected provinces

with a low data resolution, were processed with the data of the adjacent provinces (SEAFCO,

ElCO, SE-/SWPAC, SOMALlCO, TANZACO).

3.5 Accuracy ofTGe estimations

The quality of the semi-variogram models was tested by the cross-validation method, which

means a systematic removal of data points from the population. During this procedure, the model was applied to the remaining samples in order to estimate a new value. In figure 12a

we show the resulting histogram for the error value, which means the difference of the

logarithmic interpolated values and the logarithmic observed values for the entire world ocean [65°N, 65°S]1. The error value follows a statistical normal distribution, with the mean error value close to zero, indicating that the model is a good representation of the spatial variability.

1 cross-validation was calculated using data within [65°N, 65°S], only for provinces defined by semi-variogram analysis.

Regionalization of the organic carbon content in s/Il/ace sediments

25

20

>-. (.)

15

~ ~ v

t!:< <5 10 t c..

N: 3546 5 - STD: 0.28

Mean: 0

_r-1iInnn ~nmn. -1.5 -l.0 -0.5 0.0 0.5 l.0 l.5 2.0

elTor value (log interpolated value-log observed value)

Figure 12 (a): Percentage frequency of the error value as log interpolated value-log observed value.

i=: '-l <':)

(5 u -.,::: J:: '-l

-.:: « E <':)

-5 ;:l 0 rtl

"'0 .2 '-l

.§ rtl <':)

u 0 r Of)

..9

2.0

1.5

1.0

a.)

0.0

-0.)

-1. 0

-1.5

R: 0.89 SD:0.23 11: 1101 P:<O.OOOI slope: 1

•

• • ••

•

-2. 0 +----r---,-..,..-_r_---r-._~--,-_.___r____r-._--r-_r-.___, -2.0 -1.5 -1. 0 -0.5 0.0 0.) 1.0 1.5 2.0

log TOC measured southem Atlantic Ocean

51

Figure 12 (b): Results of cross-validation after discretization done with all semi-variogram proofed areas of the southem Atlantic Ocean.

52 Regionalization of the organic carbon content in s1trface sediments

The re-estimated values obtained by cross-validation are related to the measured values (R=0.84, standard deviation (SD)=0.3). The good reproducibility of the spatial dependency by the chosen semi-variogram models implies that the mean estimated values are in the order of the measured values. The cross-validation results of the southern Atlantic Ocean area are in the same order as the results of the global cross-validation with a regression coefficient of R=0.89 (SD=0.23, p< 0.0001), thus the southern Atlantic Ocean can be regarded as a representative area (Fig. 12b). Nevertheless, cross-checking is a method for re-estimating TOe contents with the modeled semi-variogram in each province and thus a method for testing the reliability of semi-variograms in the defined provinces.

To calculate the deviations of the estimated values from the real measurements and the trend of fail interpolation, we retrieved the residuals for each interpolated value from the grid for the southern Atlantic Ocean (Fig. 13a). The regression curve has a slope of 0.80 (R=0.87, SD=O.65, p< 0.0001) implying an overall underestimation. However, if a log-transfonnation is applied to the sample values, the back-transformation introduces a tendency of underestimation by allocating too much weight to lower sample values (Houlding, 2000). Generally, the values scatter around the regression curve, being a result of the relatively rough resolution of 1 ° x 1 0. The regression curve of interpolated values versus measured TOe contents of the entire global ocean [65°N, 65°S] are in the same order as the results of the southern Atlantic Ocean with a regression coefficient of R=0.84 (SD: 0.57).

14

/

U /

0 12 /

f-< / /

Of) /

..2 /

'~ 10 R: 0.87 /

0 /

,..Z3 /

c;j SD: 0.65 /

;::l /

'V 11=1101 /

"en 8 /

v Slope: 0.80 / :- • /

8 p<O.OOOl • /

0 .,./. • • ~ ••

6 • • / • • ~

/ • • • • • i'::"' • • ~, • • 'V 4 • • v c;j r

.S ~

2 v U 0 f-<

0 0 2 4 6 8 10 12 14

TOC mens [w1%]

Figure 13 (a): Regression analysis of the southem Atlantic Ocean after discretization. Solid line: real regression curve, dashed line: bisecting line. The estimated data values are retrieved from the grid.

Regionalization of the organic carbon content in surface sediments

.. CD '''';-

... , " .. .. .. " .. .. .. ..

@

.. ..

..

.. ..

~. .. .. ..

" .. 50° 40° 300 20° 10° 0°

.. It

.. .. ..

.. 10° 20° 30°

'400

:1 ~50° !

53

Percentage error of estimation

0-20 .. .. 20-40

40-60

60 - 80

.. 80 - 100

100-120

.. 120-140

140-160 1160° .. 160-180

.. > 180 -....-.;!

Figure 13 (b): Percentage estimation accuracy for the southern Atlantic Ocean calculated as the percentage error, 100 % is the exact consistency of estimated and measured value.

Expressed as the frequency distribution of estimation accuracy, 90 % of the data ranges in the

interval of the simple standard deviation of 80 % ([fl-0'; fl +0']) with a median of 99.5 (fl). This reflects the high estimation accuracy of our interpolation method. The frequency distribution

is significantly skewed to the lower values. Thus, after splitting the global data set into lower

and higher data values than the global median value of 0.6 wt%; the lower data are potentially

overestimated by a factor of 1.3 (basin data) and the higher values have a tendency to

underestimation by a factor of 0.8 (coastal data). In figure 13b the percentage deviations of the estimated Toe values from the measured Toe values are plotted for the southern Atlantic

Ocean.

4. Conclusions

A new global distribution map (1 ° x 1°) of the TOe content in surface sediments, which is

based on 5553 single data points, is presented. By conducting a qualitative pre-selection of the

data and a quantitative-geostatistical approach, a total number of 33 benthic TOe-based

regional provinces could be subdivided. Thus, complex and regionally variable interactions of

boundary parameters such as ocean currents, downslope processes, riverine input, and

sedimentation rates of biogenic and non-biogenic components could be considered and

summarized by semi-variogram analyses.

In comparison to existing global Toe distribution patterns, the regionalization in

discriminable provinces leads to a better accuracy of the estimation in general. This is

reflected i.e. by absolutely higher values along the continental margins. The estimation

54 Regionalization of the organic carbon content in surface sediments

accuracy is demonstrated by the reasonable regression fit (R=0.84) of the estimated vs.

measured TOe values and the fact that more than 90 % of the estimations range within the

simple standard deviation of ~80 %. Since the variability of regional dependencies may be

best described by a continuous spectrum of semi variograms (Higdon, 1998), a future effort could be to obtain non-stationary semi-variogram models in areas with high sample resolution. Furthermore, increasing the grid-resolution in areas with a high data density will

provide regional distribution patterns with much higher accuracy.

Since the Toe content in surface sediments is a sedimentary signal that unifies a large

number of sedimentary processes it can be used as a basic proxy parameter for mineralization

processes of organic matter and fluxes of dissolved components across the sediment-water interface. The main problem in estimating global distribution patterns of benthic fluxes (i.e.

oxygen uptake or nutrient exchange) is that their database is comparatively low (Jahnke, 1996; WenzMfer and Glud, 2002). However, the improved knowledge of empirical relationships and refined geostatistical approaches will enable the transfer of regression fits

from high to low sampled regions with similar boundary conditions. Thus, the presented

approach ofregionalization and evaluation of the TOe content in surface sediments is suitable

to enhance ongoing investigations on the marine cycle of carbon and other key elements in

future studies.

Acknowledgements

We would like to thank R. Sieger and M. Diepenbroeck for their helpful comments and

support during the data query of the data base Pangaea (www.pangaea.de. University of Bremen-Marum, A WI-Bremerhaven). Many thanks also to L. Gerullis for her patient

assistance during the work with Pangaea. We gratefully thanks H. Burger, M. Kolling and H.

Hecht for their helpful comments on this manuscript. G. Mollenhauer, D. Hebbeln, E.A Romankevich and A.A. Vetrov, P. MUller, T. Wagner and T. lennerjahn are thanked for their

providing of data. This work was funded by the Deutsche Forschungsgemeinschaft (ZA 19911- 1).

Regionalizatiol1 of the organic carbon content in slIlface sedil71ents 55

References

Antia, A.N., Koeve, W., Fischer, G., Blanz, T., Schulz-Bull, D., Scholten, J., Neuer, S., Kremling, K., Kuss, 1.,

Peinert, R., Hebbeln, D., Bathmann, U., Conte, M., Fehner, U., Zeitschel, B., 2001. Basin-wide

particulate carbon flux in the Atlantic Ocean: Regional export pattems and potential for atmospheric

CO 2 sequestration. Gobal Biogeochemical Cycles 15 (4), 845-862.

Antoine, D., Andre, l-M., Morel, A., 1996. Oceanic primary production; 2. Estimation at global scale from

satellite (coastal zone color scanner) chlorophyll. Global Biogeochemical Cycles 10 (1), 57-69.

Archer, D.E., 1996. An atlas of the distribution of calcium carbonate in sediments of the deep sea. Global

Biogeochemical Cycles 10 (1), 159-174.

Arthur, M.A., Dean, W.E., Laarkamp, K., 1998. Organic carbon accumulation and preservation in surface

sediments on the Peru margin. Chemical Geology 152,273-286.

Behrenfeld, M.l, Falkowski, P.G., 1997. A consumer's guide to phytoplankton primary productivity models.

Limnology and Oceanography 42 (7), 1479-1491.

Berger, W.H., Wefer, G., 1992. Klimageschichte aus Tiefseesedimenten. Naturwissenschaften 79, 541-550.

Berger, W.H., Smetacek, V.S., Wefer, G., 1989. Ocean productivity and paleoproductivity - an overview. In:

Berger, W.H., Smetacek, V.S. and Wefer, G. (Eds.), Productivity of the ocean: present and past, 1989.

John Wiley & Sons Lim., pp. 1-34.

Betzer, P.R., Showers WJ., Laws, E.A., Winn, C.D., DiTullio, G.R., Kroopnick, P.M., 1984. Primary

productivity and particle fluxes on a transect of the equator at 153°W in the Pacific Ocean. Deep-Sea

Research 31, 1-11.

Bordovskiy, O.K., 1965. Accumulation and transformation of organic substances in marine sediment. Marine

Geology 3, 3-82.

Brezukov, P.L., Yemelyanov, Y., Lisitzin, A.P., 1977. Organic carbon in the upper sediment layer of the world

ocean. Oceanology 17 (5), 561-564.

Cwienk, D.S., 1986. Recent and glacial age organic carbon and biogenic silica accumulation III marine

sediments. Master of Science Thesis in Oceanography, University of Rhode Island.

De Marsily, G., 1986. Quantitative Hydrogeolog. Academic Press, Inc., Harcourt Brace Javanovich.

Davis, J.c., 1973. Statistics and data analysis in geology. John Wiley Int. Edition, New York - London - Sydney

- Toronto.

DeMaster, DJ., 1981. The supply and accumulation of silica III the marine environment. Geochimica et

Cosmochimica Acta 45,1715-1732.

Ewing, M., Ludwig, W.J., Ewing, l1., 1964. Sediment distribution in the Oceans: The Argentine Basin. Joumal

of Geophysical Research 69, 2003-2033.

Garzoli, S.R., 1993. Geostrophic velocity and transport variability in the Brazil-Malvinas Confluence. Deep-Sea

Research 140,1379-1403.

Giraudeau, 1., Meyers, P.A., Christensen, B.A., 2002. Accumulation of organic and inorganic carbon m

Pliocene-Pleistocene sediments along the SW African margin. Marine Geology 180,49-69.

Gordon, A.L., 1989. Brazil-Malvinas Confluence-1984. Deep-Sea Research II 36, 359-384.

Hebbeln, D., 2000. Flux of ice-rafted detritus from sea ice in the Fram Strait. Deep-Sea Research I 47, 1773-

1790.

Hedges,1. 1., 1992. Global biogeochemical cycles: progress and problems. Marine Chemistry 39, 67-93.

56 Regionalization of the organic carbon content in swjace sediments

Henrichs, S.M., 1992. Early diagenesis of organic matter in marine sediments: progress and perplexity. Marine

Chemistry 39,119-149.

Henrichs, S.M., Reeburgh, W.S., 1987. Anaerobic mineralization of marine sediment organic matter: rates and

the role of anaerobic processes in the oceanic carbon economy. Geomicrobiological Journal 5 (3/4),

191-237.

Hensen, C, Landenberger, H., Zabel, M., Schulz, H.D., 1998. Quantification of diffusive benthic fluxes of

nitrate, phosphate and silicate in the Southern Atlantic Ocean. Global Biogeochemical Cycles 12, 193-

2lO.

Hensen, C, Zabel, M., Schulz, H.D., 2000. A comparison of benthic nutrient fluxes from deep-sea sediments of

Namibia and Argentinia. Deep-Sea Research II 47, 2029-2050.

Hensen, C, Zabel, M., Pfeifer, K., Schwenk, T., Kasten, S., Schulz, H., 2003. Control of sulfate pore-water

profiles by sedimentary events and the significance of anaerobic oxidation of methane for the burial of

sulfur in marine sediments. Geochimica et Cosmochimica Acta 67, 2631-2647.

Higdon, D., 1998. A process-convolution approach to modelling temperatures in the North Atlantic Ocean.

Environmental and Ecological Statistics 5 (2), 173-190.

Hillier, 1995. Erosion, sedimentation and sedimentology origin of clays. In: Velde, B. (Ed.), Origin and

mineralogy of clays. Springer, Berlin - Heidelberg - New York, pp. 162-219.

Holtvoeth, J., Wagner, T., Horsfield, B. Schubert, C.J., Wand, U., 200l. Late-Quaternary supply ofterrigenous

organic matter to the Congo deep-sea fan (ODP site 1075): Implications for equatorial African

paleoclimate. Geo-Marine Letters 21, 23-33.

Houlding, S.W., 2000. Practical Geostatistics: Modelling and spatial analysis. Springer.

Ikeda, M., Johannessen, J.A., Sandven, S., 1989. A process study of mesoscale meanders and eddies in the

Norwegian Coastal Current. Journal of Physical Oceanography 21,893-902.

lttekkot, V., 1988. Global trends in the nature of organic matter in river suspensions. Nature 332, 436-438.

Jahnke, R.A., 1996. The global ocean flux of particulate organic carbon: Areal distribution and magnitude.

Global Biogeochemical Cycles 10,71-88.

Jorgensen, B.B., 2000. Bacteria and marine biogeochemistry. In: Schulz, H. D. and Zabel, M. Marine

Geochemistry (Eds.), Marine Geochemistry, 2000. Springer.

Jones, C, 1999. Geographical Infonnation Systems and Computer Cartography. In: Jones, C (Ed.), The

Geographical Information Systems and Computer Cartography shape of the earth, 1999.

Lampitt, R.S., Antia, A.N., 1997. Particle flux in the deep seas: Regional characteristics and temporal variability.

Deep-Sea Research 144, 1377-1403.

Lohse, L., Helder, W., Epping, E.H.G., BaizeI', W., 1998. Recycling of organic matter along a shelf-slope

transect across the NW-european continental margin (Goban Spur). Progress in Oceanography 42, 77-

110.

Longhurst, A., Sathyendranath, S., Platt, T., Caverhill, CM., 1995. An estimate of global primary production in

the ocean from satellite radiometer data. Journal of Plankton Research 17, 1245-1271.

Ludwig, W., Probst, J.-L., Kempe, S., 1996. Predicting the oceanic input of organic carbon by continental

erosion. Global Biogeochemica1 Cycles 10 (1), 23-41.

Martin, J.H., Knauer, G.A., Karl, D.M, Broenkow, W.W., 1987. VERTEX: carbon cycling in the northeast

Pacific. Deep-Sea Research 34 (2), 267-285.

Matheron, G., 1965. Les variables regionalisees et leur estimation-Edition Masson et Cie. Paris.

Regionalizatiol1 of the organic carbon content in slllface sediments 57

Meeuwis, J.M., Lutjehanns, ] .R.E., 1990. Surface Thermal Characteristics of the Angola-Benguela Front. South

African Journal of marine Science 9, 261-79.

Meybeck, M., 1993. C, N, P and S in rivers: from sources to global inputs. In: Wollast, R., Mackenzie, F. T. and

Chou, L. (Eds.), Interactions of C, N, P and S: biogeochemical cycles and global change, 1993.

Springer, Berlin - Heidelberg - New York, pp. 163-193.

Mienert, J., Berndt, c., Laberg, J.S., Vorren, T.O., 2003. Slope instability of continental margins. In: Wefer, G.,

Billet, D., Hebbeln, D., ]orgensen, B.B., Schliiter, M., van Weering, T.CE. (Eds.), Ocean Margin

Systems, 2003. Springer-Verlag, Berlin - Heidelberg, pp. 179-193.

Mollenhauer, G., Schneider, R.R., MUller, P.J., SpieJ3, V., Wefer, G., 2002. Glaciallinterglacial variability in the

Benguela upwelling system: Spatial distribution and budgets of organic carbon accumulation. Global

Biogeochemical Cycles 16, 1134.

Mollenhauer G., Schneider R.R., ]enneljahn T., MUller PJ., Wefer G., in press. Organic carbon accumulation in

the South Atlantic Ocean: Its modern, mid-Holocene and last glacial basin-wide distribution. Global and

Planetary Change.

Pace, M.L., Knauer, G.A., Karl, D.M., Martin, ].H., 1987. Primary production, new production and vertical flux

in the eastern Pacific. Nature 325, 803-804.

Pannatier, Y., 1996. VarioWin-Software for spatial data analysis in 2D - Statistics and computing. Springer.

Peterson, R.G., StraImna, L., 1991. Upper-level circulation in the South Atlantic Ocean. Progress 111

Oceanography 26,1-73.

Peterson, R.G., ]ohnson, CS., Krauss, W., Davis, R.E., 1996. Langrangian measurements in the Malvinas

Current. In: Wefer, G., Siedler, G., Berger, W.H., Webb, DJ (Eds.), The South Atlantic: Present and

Past Circulation, 1996. Springer, pp. 239-247.

Prakash Babu, C, Brumsack, H.-J., Schnetger, B., 1999. Distribution of organic carbon in surface sediments

along the eastern Arabian Sea: a revisit. Marine Geology 162, 91-103.

Premuzic, E.T., Benkovitz, CM., Gaffney, J.S., Walsh, ].]., 1982. The nature and distribution of organic matter

in the surface sediments of world oceans and seas. Organic Geochemistry 4, 63-77.

Ragueneau, 0., Treguer, P., Leynaert, A., Anderson, R.F., Brzezinski, M.A., DeMaster, D.]., Dugdale, R.C,

Dymond, ]., Fischer, G., Francois, R., Heinze, C, Maier-Reimer, E., Martin-]ezequel, V., Nelson,

D.M., Queguiner, B., 2000. A review of the Si cycle in the modern ocean: recent progress and missing

gaps in the application of biogenic opal as a paleoproductivity proxy. Global and planetary Change 26,

317-365.

Ranero, CR., von Huene, R., 2000. Subduction erosion along the Middle America convergent margin. Nature

404,748-752.

Reimers, CE., 1987. An 111 situ microprofiling instrument for measuring interfacial pore water gradients:

methods and oxygen profiles from the North Pacific Ocean. Deep-Sea Research 34 (12), 2019-2035.

Rowe, G.T., Boland, G.S., Phoel, W.C, Anderson, R.F., Biscaye, P.E., 1994. Deep-sea floor respiration as an

indication oflateral input of biogenic detritus from continental margins. Deep-Sea Research II 41, 657-

668.

Romankevich, E.A., 1977. Geochemistry of organic matter in the ocean. Nauka (Moskow), 11-256.

Romankevich, E.A., 1984. Geochemistry of organic matter in the ocean. Springer Verlag, Berlin - Heidelberg -

New York - Tokio.

Romankevich, E.A., Vetrov, A.A., 2001. Cycle of carbon in the Russian Arctic seas: Nauka Press (in Russian).

Sarmiento, J.L., Gruber, N., 2002. Anthropogenic carbon sinks. Physics Today 55 (8),30-36.

58 Regionalization oj'the organic carbon content in surj'ace sediments

Sathyendranath, S., Longhurst, A, Caverhill, C.M., Platt, T., 1995. Regionally and seasonally differentiated

primary production in the North Atlantic. Deep Sea Reseearch 42 (l0), 1773-1802.

Shannon, L.V., Nelson, G., 1996. The Benguela: Large scale features and processes and system variability. In:

Wefer, G., Berger, W.H., Siedler, G. (Eds.), The South Atlantic: Present and past circulation, 1996.

Springer Verlag, Berlin - Heidelberg, pp. 163-210.

Schlesinger, W.H., Melack, l.M., 1981. Transport of organic carbon in the worlds rivers. Tellus, 172-187.

Schlitzer, R., 2002. Carbon export fluxes in the Southern Ocean: results from inverse modeling and comparison

with satellite-based estimates. Deep-Sea Research II 49, 1623-1644.

Schliinz, B., Schneider, R.R., MUller, PJ., Showers, WJ., Wefer, G., 1999. TelTestrial organic carbon

accumulation on the Amazonas deep sea fan during the last glacial sea level low stand. Chemical

Geology 159,263-281.

Smith, K.L.j., 1987. Food energy supply and demand: A discrepancy between particulate organic carbon flux

and sediment cOlID11Unity oxygen consumption in the deep ocean. Limnology and Oceanography 32 (1),

201-220.

Stow, D.A.V. and Mayall, M., 2000. Deep-water sedimentary systems - new models for the 21st century. Marine

and Petroleum Geology 17, 125-135.

Suess, E., 1980. Particulate organic carbon flux in the oceans - surface productivity and oxygen utilization.

Nature 288,260-263.

Tintelnot, M., 1995. Transport and deposition of fine-grained sediments on the Brazilian continental shelf as

revealed by clay mineral distribution. PhD Thesis, Ruprecht-Karls-Universitat.

Treguer, P., Nelson, D.M., van Bennekom, Al., DeMaster, D.l., Leynaert, A. and Queguiner, B., 1995. The

silicate balance in the world ocean: a reestimate. Science 268, 375-379.

Tyson, R.V., 2001. Sedimentation rate, dilution, preservation and total organic carbon: some results of a

modelling study. Organic Geochemistry 32, 333-339.

Uenzelmann-Neben, G., 1998. Neogene sedimentation history of the Congo Fan. Marine and Petroleum Geology

15 (7), 635-650.

Vetrov, AA., Romankevitch, E.A, 1997. New charts of distributions of organic carbon and coefficients of its

fossilisation in the bottom sediments of the ocean. Oceanology 37 (6), 769-775.

Vetrov, AA., Romankevitch, E.A, Benenson, M.A., 1997. Computer compilation of organic carbon distribution

maps for bottom sediments. Oceanology 37 (5), 651-656.

Walsh, J.J., 1991. Importance of continental margins in the marine biogeochemical cycling of carbon and

nitrogen. Nature 350 (6313), 53-55.

Weaver, P.E., Wynn, R.B., Kenyon, N.H., Evans, l., 2000. Continental margin sedimentation, with special

reference to the north-east Atlantic margin. Sedimentology 47 (Suppl.1), 239-256.

WenzhOfer, F., Glud, R., 2002. Benthic carbon mineralization in the Atlantic: a synthesis based on in situ data

from the last decade. Deep-Sea Research I 49, 1255-1279.

Wollast, R., 1998. Evaluation and comparison of the global carbon cycle in the coastal zone and in the open

ocean. In: Brink, K.H., Robinson, A.R. (Eds.), The Sea, 1998. University of Brussels, pp. 213-252.

Woodgate, R.A., Fahrbach, E., Rohardt, G., 1999. Structure and transport of the East Greenland Current at 75°N

from moored current meters. Journal of Geophysical Research 104, 18059-18072.

Zabel, M., Dahmke, A, Schulz, H.D, 1998. Regional distribution of diffusive phosphate and silicate fluxes

through the sediment-water interface: the eastern South Atlantic. Deep-Sea Research I 45,277-300.

Regionalizati077 of the organic carbon content in surface sediments 59

Zabel, M., Schulz, H.D., 2001. Importance of submarine landslides for non-steady state conditions in pore water

systems -lower Zaire (Congo) deep-sea fan. Marine Geology 176, 87-99.

60 The benthic carbon mineralization on a global scale

2.2 The benthic carbon mineralization on a global scale

Katherina Seiter l, Christian Hensen", Matthias Zabel l

IDepartment of Geosciences, University of Br em en, Klagenfurter Stral3e, D-28359 Bremen, Germany

"GEOMAR - Forschungszentrum fUr Marine Geowissenschaften, SFB 574, Wischhofstr. 1-3,24148 Kiel,

Germany

Abstract

In this study we present a global distribution pattern and budget of the minimum flux of particulate organic

carbon to the sea floor (JPOCa). The estimations base on regional specific characteristics in the relation between

the diffusive oxygen flux across the sediment-water interface, the total organic carbon content in surface

sediments, and the oxygen concentration in bottom waters. For this, we modified the principle equation after Cai

and Reimers (1995), as a basic monod reaction rate, within 11 regions where in situ measurements of diffusive

oxygen uptake exist. By application of the resulting transfer functions to other regions with similar sedimentary

conditions and areal interpolation, we calculated a minimum global budget of particulate organic carbon that

actually reaches the sea floor of ~0.5 GtC yr- I (>1000 m wd), whereas approximately 0.002-0.12 GtC yr- I is

actually buried in the sediments (0.01-0.4 % of surface primary production). Despite the fact that our global

budget is in a good agreement with previous studies, we found conspicuous differences between the distribution

patterns of primary production, sediment trap based calculations of the POC flux, and JPOCa of this study. These

deviations, especially located at the south-eastern and south-western Atlantic Ocean, the Greenland and

Norwegian Sea and the entire equatorial Pacific Ocean, strongly indicate a considerable influence of lateral

particle transport on the vertical link between surface waters and underlying sediments. This observation is

supported by sediment trap data. Furthermore, local differences in the availability and quality of the organic

matter as well as in transport mechanisms through the water column are reflected.

The benthic carbon mineralization on a global scale 61

1. Introduction

From a very simplistic view the marine carbon cycle is driven by carbon fixation and release within the water column and geochemical reactions of equilibration across the two borders, atmosphere-water interface and benthic boundary layer. In contrast to the first, the sea floor represents a net sink, where carbon is storaged and removed from the cycle for long periods of time. Thus, (bio )geochemical processes in marine sediments play an important key role to understand the budget of the marine and in the end also of the global carbon cycle (Jahnke, 1996; Christensen, 2000; Wenzhofer and Glud, 2002; Schlitzer, 2002; Lasaga and Ohmoto, 2002; Archer, 1996). Frequently discussed general questions are: How much organic carbon reaches the sea floor? Where are the main depot-centers, what are the driving processes of accumulating organic carbon and which pOliion is recycled by early diagenetic processes?

Only a small fraction of the primary production settles on the sediment surface. According to Wollast (1998), this pOliion amounts to 1-1.5 % of the primary production (PP) for the open ocean and up to 17 % on the upper slopes. Due to intense microbial degradation in surface sediments only 0.014 % and 0.5-3 % of PP are finally buried, respectively. On ocean-wide scales the accumulation and burial of organic matter principally reflects the distribution pattern of the primary production. Therefore, in several studies empirical equations have been derived, which describe the decreasing rain rate of patiiculate organic carbon (POC) with water depth by using local estimates of surface productivity and trap data (e.g. Suess et aI., 1980; Betzer et aI., 1984; Berger et aI., 1989; Pace et aI., 1987; Schlilter et aI., 2000; Antia et aI., 2001; WenzhOfer and Glud, 2002). Regionally, however, this approach of a simple veliical transport may not be sufficient, since intense water circulation and strong bottom water CUlTents can cause considerable lateral transport of suspended matter, especially across the shelf breaks and down the slopes (Jahnke et aI., 1990; Reimers et aI., 1992; Garzoli, 1993; Rowe et aI., 1994; Peterson et aI., 1996; Wollast, 1998; Arthur et aI., 1998; Hensen et aI., 2000; Giraudeau et aI., 2002; Mollenhauer et aI., 2002; Hensen et aI., 2003). After Wollast (1998) approximately 2.2 GtC of the annual primary production at coastal areas are expOlied to the deep ocean in that way. Fmihermore, common empirical equations also do not take other effects like potential riverine input of organic matter into consideration, which can significantly influence the accumulation and burial of organic carbon (Schlesinger and Melack, 1981; Ittekkot, 1988; Hedges, 1992; Meybeck, 1993; Hedges and Keil, 1995; Ludwig et aI., 1996; Wagner et aI., in press). Nevertheless, there are several additional factors, which are governing the organic carbon rain through the water column and thus decouple the empirical relation between surface primary production and organic carbon flux to the sea floor. Recent studies have cOlToborated former observations that also the composition of the paliicle rain is of particular importance. After Armstrong et aI. (2002) and Klaas and Archer (in press) POC fluxes are tightly linked quantitatively to the fluxes of opal and calcite, which serve together with lithogenic minerals as ballast calTiers and, therefore, may influence the degradation efficiency of POC in the water column. Furthermore, an increasing flux of non­reactive particles to the sea floor results in the reduction of the oxygen exposure time, which enhance the burial of organic matter (e.g. Jahnke, 1996; Hartnett et aI., 1998). Last but not least, also the availability of oxygen in bottom water can affect the preservation of organic

62 The benthic carbon mineralization on a global scale

matter (e.g. Betts and Holland, 1991; Calvert and Pedersen, 1992; Archer and Devol, 1992; Cai and Reimers, 1995).

However, in general two approaches are in use to quantify the POC flux to the sea floor (Jpoe). The direct way is based on particle collections with sediment traps. Like mentioned before, this way is susceptible to misinterpretation due to some effects named above. Additionally, the number of such measurements is rather limited. The indirect approach uses i.e. correlations between different proxy parameters. For example, Jahnke (1996) has specified and estimated global Jpoe via the diffusive benthic oxygen uptake (DOU) and carbonate cOlTected organic carbon burial rates. In any case, all approaches on a global scale disregard the regional variability of environmental conditions so far.

In our study, we emphasize the key role of the total organic carbon content (TOC) in surface sediments, the benthic oxygen availability, and the microbial oxygen respiration for calculation of Jpoe. Furthermore, we try to consider regional variabilities in oceanographic and sedimentary conditions. With an extended data compilation of 122 sites where DOU measurements exist, we use the indirect approach of developing specific transfer functions by multiple regression analyses on TOC, DOU, and the bottom water concentration (BOC). The characterization of typical relationships between these parameters in different benthic provinces enables us to extrapolate these observations to the global ocean. Finally, the resulting distribution pattern of Jpoe is discussed with regard to other studies, especially sediment trap data, and a rough estimation of the global TOC burial rate is given.

2. Methods

2.1 Global data base and compilation

2.1.1 Raster data

Total organic carbon content in sUliace sediments (TOe, < 5 cm water depth)

As a basis for the TOC distribution at the deep-sea floor, we used the global map of Seiter et al. (subm.) as depicted in figure 1 a.

The benthic carbon mineralization on a global scale 63

-180 I

-120 -qO o 60 120 180

-60

-0

--60

I I

-180 -120 -60 o 60 120 180

<0.25 0.5-0.75 1-1.25 1.5-2 2.5-3 3.5-4 4.5-5 5.5-6 6.5-7 > 7

Figure 1 (a): Global distribution pattern of the total organic carbon content in [wt%] in surface sediments « 5 cm sediment depth) (Seiter et aI., subm.).

The consideration of regional variablities in intercorrelations between the total organic carbon content (TOC), the diffusive benthic oxygen uptake (DOU), and the bottom water oxygen concentration (BOC) is based on the discretization of the benthic regime into 33 TOe-based regional provinces (Fig. 1 b; Seiter et al. subm.).

Figure 1 (b): Discretization of the global ocean in 33 benthic TOC-based provinces (TROPAC/TROPAC2; NEPAC/NWPAC; SEPAC/SWPAC: each pair is characterized as one province) (Seiter et aI., subm.).

64 The benthic carbon mineralization on a global scale

Primary production (PP)

The calculation of primary production (PP) was taken from Antoine et al. (1996). This model is based on chlorophyll-a data derived by the Coastal Zonar Colour Scanner (CZCS, NASA),

giving an annual global budget of 36.5-45.6 GtC yr- 1• Model results are comparable to

estimations after Behrenfeld and Falkowski (1997-b) in principle. Local differences are

caused by higher coastal values on average of Behrenfeld and Falkowski (1997 -b) and higher estimates along the latitudinal bands between 30o-50o S and 30o-80oN. The basin wide estimates derived after Behrenfeld and Falkowski (1997 -a,b) are lower, whereas the estimates

for the higher latitudes (above 600 N) are significantly higher than the correspondent mean

values after Antoine et al. (1996). Although there is a general agreement among the different models for particular stations, the estimates after Antoine et al. (1996) show a better

agreement with measured data (Anti a at aI., 2001). Thus, we processed the primary production estimates after Antoine et al. (1996) for further calculations and for comparisons

the Behrenfeld and Falkowski (1997-b) estimates. Mean values and budgets were queried

from the grids and summarized in Appendix II for the appropriate provinces.

Empirically determined benthic particular organic carbonfluxes (hoc)

For the estimation of particular organic carbon fluxes (Jpoe) , we applied the empirical algorithm after Antia et al. (2001) for the latitudes between [65°N, 85°S], which is based on

sediment trap data (Eq. 1). For the higher latitudes (> 65°N) we used the algorithm after

Schlliter et al. (2000) for comparisons with our estimations. The pattern was processed with

the basic grids of ETOP05-bathymetry (z) and primary production (PP) after Antoine et al. (1996) (Tab. 1).

J poe = c X ppa X Zb

c=O.l; a=l.77; b=-0.68 (Anti a et aI., 2001)

c=1; a= l.873; b=-l.l72 (SchlUter et aI., 2000)

Bottom water oxygen concentration (BOC)

(1)

For the data compilation and interpolation of the global distribution pattern of the bottom water oxygen concentration in a 1 ° x 1 ° grid resolution (cf. sect. 3.1, Fig. 2) we combined

6788 individual bottle data. The main part is queried from the Electronic Atlas of WOCE Data (Schlitzer, 2000), which provides global ocean observations for the decade 1988 until 1998. Additionally, we used the global hydro graphic station data set from the WHP (World Hydrographic Program), available at the National Oceanographic Data Centre. 239 individual observations were used from the Geochemical Ocean Sections study, available at the Climate Data Library (GEOSECS; IRI/LDEO). The data compilation was completed by

single data from individual sites, which were taken from the literature. The distribution

pattern was calculated by dividing the world oceans into the southern- and northern Atlantic Ocean, Pacific Ocean, and Indian Ocean applying the kriging method. To avoid artefacts due

The benthic carbon mineralization 011 a global scale 65

to the very irregular data distribution, we chose a random selection by 10 % of all data for

semi-variogram analyses. In each Ocean, a linear semi-variogram could be modeled.

2.1.2 Point data at individual sites

Benthic oxygen consumption (DOU, TOU)

The data of the benthic oxygen consumption were taken from the network for geological and environmental data PANGAEA (www.pangaea.de) and from the literature. With the majority,

it concems in situ diffusive oxygen uptake measurements (DOU, 122 sites). Because in situ

total consumption rates (TOU, 78 sites) are far less available and concentrate at some few

regions, for ongoing multiple regression analyses and fmiher investigations we used the in

situ DOU measurements only (cf. Fig. 2, Tab. 1).

In situ TOU data were predominantly used for comparative estimations along the continental

margin areas. Only at the higher latitudes, for the polar regions, our study depends as well on

ex situ DOU rates (29 sites). The alteration due to temperature increase during core sampling

might be of minor importance compared to decompression, since the temperature differences

between sea floor (-1, 1.3°C) and deck temperatures « O°C) are generally low (Sauter et al.,

2001). Decompression effects increase with water depth and are low for the shallow polar

seas (Sauter et al., 2001).

Trap data

A compiled data set with measurements from 61 locations was used

(s. Fig. lOb; Tab. 1; Appendix IV). We only considered trap data with a maximum distance to

the sea floor of 550 m. At individual sites data were integrated over periods between 61 days

and annually integrated long time series.

Sedimentations rates (SR)

For the estimation of mean burial rates 433 holocene sedimentation rates (SR) were compiled.

They were calculated by linear age interpolation between the core top and the first tie point

derived from oxygen stable isotope stratigraphy, radiocarbon ages, or as described in the

respective literature. We used data sets and data compilations, respectively, from Muller and

Suess (1979), Berelson et al. (1990), Morse and Emeis (1990), Murray et al. (1993), Wagner

(1993), Kreutz (1994); Cai and Reimers (1995), McManus et al. (1995), Murray and Leinen

(1996), Haese et al. (1997), van der Weijden et al. (1999), Lyle et al. (2000), Reimers et al.

(1992), Wollenburg and Kuhnt (2000); Mollenhauer et al. (in press), etc.

The complete data sets of all used data of this study can be queried from

www.pangaea.de. An overview is given in Tab. 1.

Table 1: Parameters and data sources, references for data collections, compiled for this study, can be gripped from www.pangaea.de.

Parameters

Primary production Behrcnfeld and Falkowski

(1997 -a,b)

Antoine ct al. (1996)

BathYllletry ETO[,05

Bottom water oxygen concentration

(BOC)

Total organic carbon contcnt in surface sediments

(TOe, <' 5 cm sediment depth)

Organic carbon flux to the sea 11o(\r (empirical relation based on

Betzcr et al. ( 1(84), modifkd afta Antia et al. (200 I))

Trap sites

DOlJ- diffusive oxygcnllptakc TOU - total oxygen uptake (> 1000 m; > 550 m wd

Cl reenland -I ce land-N orwegian Sea)

Sedimentation rates (SR), dry bulk density (DBD)

Data base

OPD (Oceanographic PrOdllCfivify Dafabase)

KiOFSJ'rance database

NGDC/ woe MGG (.iVafional Geophysical Dala Cenferi World

Data Celller for marine Geology and Geophysics. BOlllder)

USGS ((/5' Gwlogical SUl1'e},)

GEOSECS Data I Rffl.DEO (Climate Data Librarv

WOCE Data WOD/WO/\. ( World OceailVatabase and

World Ocean Atlas 1998)

Pangaea- Network for Geological and Environmcntal Data

Bathymetry and primal) production

This study, compilation of globally available sites

www.pangaea.de

Data siks/grid resolution

2048 :< 1024 binary matrix

204R x 1024 binary matrix (both: exclude ice-cover)

5' x 5' matrix

678R sites, compiled to I C x I" matrix

5553 surface data sites

[83YN,WS] I C x 10 matrix, rccalculated

61 sites

Number of sites, 122: DO{J in situ 29: DOll ex situ 78: TOU in situ

433 siles

pers, contacts

P,(J. Falko\\'ski (i?lIlgers

Univasif)) D, Kolbcr (Rllfgers

UniFersilJ

Source

Bchrcnfcld and FalkolVski ( 19978,\:1) Data compilation of primary prnduction

( http://marinc. rlltgers, cd u'opp:'I'rodllct i on VPGMRcs,html)

Antoine ct al. (1996) (http: .iwww.obs-vlfr.fr/jgofs2!ll1odcl isation)

http: . Iwww.obs-vlfr.fr!jgolVhtmlihtmliacccs_base.htm I

www.ngdc.noaa.gov/mgg'gloilal!scitopo,html

http:;iingrid,ldgo,colul11bia,cdu! http://\vww, node, noaa,gov:()C 5!dala_ \\'oa, hIm I

http://www-ocean.tamu.edlrWOCEuswoce.html

Rcgionalization of the organic carbon conte nl in surface sediments defining a new approach of benthic biog~ochcmical pwvinces­

(Seitcr ct al. subm ,)

Antia et aL (1001)

www.pangaea.de litenlture data (Appendix Ill)

w\v\v. pangaea. de literature data

www.pangaea.de literature data

0\ 0\

::J ~ c­~

~ :::;-. ;:::;.

" ~ c:i-~

::: ~

2.. t:::i. ~ 0' :::

~ !::)

IJQ

Cl c-~ r..,

" !::)

--­('i)

The benthic carbon mineralization on a global scale 67

2.2 Multi regression analyses

For regression analyses, we cOlTelated DOD with BOC and TOC. The basic approach was

adapted from Cai and Reimers (1995). To avoid steep gradients between basin and continental

margin areas, we included the data of the southern and nOlihern Atlantic and the NE-Pacific

basin province respectively for regression analyses. All multiple regression calculations were

done with the statistic software SPSS® 10.3.

The general equation follows a monod kinetic and is a first order in organic matter

concentration and hyperbolic in the concentration oxidant (Eq. 2; Boudreau, 1997).

-3[Ox] =-kx([TOC]x [Ox] ) at (Kox + [Ox D

Where - 3 [Ox] is the time dependent decay of organic matter per area. at

(2)

This tenn can be replaced by the oxygen consumption in [mmol m-2 day-I] as representing the

most important electron acceptor being utilized. It is also used as an approximation of the

total benthic mineralization rate in [mmol m-2 day-I], calculated with a given organic matter

composition by the Redfield ratio of (C:N:P 106/16/1) and the aerob degradation of organic

matter in a molar ratio of 106 mol C/138mol O2 after Froelich (1979). [Ox] is the

concentration of oxygen in overlaying bottom water (BOC) in [mmol m-3] and [TOC] the

concentration of organic carbon in surface sediments in [wt%]. Kox is a saturation constant of

BOC in [mmol m-3] and k is the rate constant (units of inverse time).

We modified equation (2) by the best fit to an empirically derived half logarithmic relation,

which couples DOD with TOC in surface sediments and BOC (Eq. 3):

_ 3 [Ox]:::::; DOD = _ [[In(TOC + K3)]X K, + K2]X [BOC]

at Kox + [BOC] (3)

Despite the fact that most regression models could be fitted sufficiently with a linear approach

for DOD vs. TOC and a rational approach of first order for DOD vs. BOC and TOC (Cai and

Reimers, 1995), provinces with enhanced productivity or reduced DOD were fitted best with a

logarithmic approach. Additionally, the best regression fits were found, if we consider 4

constants, KI, K2, K3 and Kox. For all regions with Kox ~ 0, the influence of reduced oxygen

content in overlaying bottom water is negligible. Thus, the reaction is essentially independent

ofBOC, ifBOC» Kax and the reaction follows a first order kinetic.

DOD has the unit [mmol m-2 day-I] as well as KI and K2, TOC and K3 have the units [wt%],

Koxhas the unit ofBOC as [mmol m-3].

In the relation between TOC and DOD the degradable fraction of the sum parameter TOC

contributes to the benthic depletion of organic matter, only (Eq. 3). The utilization of the

common TOC content will cause the intersection of the abscissae (TOCf3). Correspondingly,

68 The benthic carbon mineralization on a global scale

low TOC contents would con-elate with oxygen release into the bottom water, which is in

contradiction to common observations. To avoid this effect of an increasing pOliion of

refractory organic carbon (TOCp) with a decreasing TOC content we define the following case differentiation. Up to a certain TOC value (TOC1im), we assume a linear correlation

between DOU and TOe. At higher TOC contents, the relation can be described by equation (3). The transition between these two cases is given by the osculation point of a zero

fixed tangent on the function of equation (3). This approach results in the assumption of a linear con-elation up to TOC1im (Tab. 2, Eq. 4.2, cf. sect. 3.1). TOClim ranges between

0.32 wt% for the northern and southern Atlantic Ocean and 0.83 wt% for the Pacific Ocean (median global TOC values: 0.6 wt%). The linear relation between DOU and TOC is

equivalent to the relation after Cai and Reimers (1995) if Kox ~ 0 and is applicable for the

basin areas, where BOC is of minor impOliance and TOC is generally low (Eq. 4.1).

[TOC]xK !\ K > 0, TOC < TOC lim (4.1)

DOU=

[[In(TOC + KJ]x K, + K2]X [BOC] !\ K > 0 K > 0 TOC > TOC Kox + [BOC] ox 'I' hm

(4.2)

Estimations of TOU along the continental margin of SW-Africa, the Arabian Sea, the

Greenland-Norwegian-Iceland Sea, the continental margin off Chile and the continental margin off Namibia were derived by an exponential fit. Equation (5) relates the average TOU

to the average TOC in the particular provinces.

(5)

The mean refractory organic carbon content (TOC p) can be calculated for DOU=O and Kox=O after equation (6 and 6.1).

DOU = 0 = [In(TOC p + K 3 ) X K, + K 2] (6)

K2

Toe -e KI_K p - 3 (6.1)

The mean burial rate (~in [gC m-2 yr-1]) was calculated by using TOCp in [g g-l], the

sedimentation rate (SR in [cm kyr-1]) and the dry bulk density (DBD in [g cm-3

]).

~ = 0.1 x TOCp x SR x DBD (7)

For further data processing and budgeting the mean values of DBD and SR for the different provinces (cf. Fig. 1 b, Tab. 4) were calculated.

2.3 Mapping and budgeting carbanjluxes ta the seajlaar

The regression coefficients, determined by multiple regression analyses were applied on the

input data grids in the sedimentary provinces respectively (Tab. 2). After merging the

The benthic carbon mineralization on a global scale 69

particular grids the transition of adjacent provinces was calculated by the moving average of overlapping cells and was thus smoothed. Nevertheless, the error is assumed as small due to avoid overlapping in areas with high concentration gradients.

For calculating the global budget of the export of organic carbon to the sea floor, we projected the global distribution pattern of the processed benthic POC fluxes of this study as well as the reference distribution pattern of primary production, and export fluxes using the pseudocylindrical equal-area Mollweide projection. Since a projection causes the change of the map area and its geometry, cells have to be recalculated. The resampling of the cells by the given new projection was done by bilinear interpolation of the cell values (Arc View® 3.2 a). For all calculations and budgeting, the grids were resampled to the grid resolutions of the reference grids to a grid extend of 2048 x 1024 of a rectangular matrix.

3. Results and Discussion

3.1 Regression analyses and budgeting

Here, we present detailed regression analyses of TOC, in situ diffusive and total benthic oxygen flux rates (DOU, TOU), and the bottom water oxygen concentration (BOC). While huge data sets of TOC and BOC values are available on a global scale, information from oxygen flux measurements are very sparse. Furthermore, much more values for DOU exist than for TOU. Therefore, at first we will focus on DOU and refer to TOU only at local descriptions. Consequences resulting from this restrictive view will be discussed later.

Influence of Boe on DOU and TOe

The preservation of organic carbon ultimately depends on the time period that organic matter in surface sediments is exposed to oxic conditions before burial (Reimers, 1989, Hartnett et aI., 1998). Thus, along the continental margins increasing sedimentation rates and decreasing oxygen concentrations in bottom water favour the preservation of TOC. The global distribution of oxygen in bottom water in figure 2 clearly illustrates the formation of oxygen rich deep water mainly in the northern N-Atlantic Ocean and the thermohalin circulation pathways. Remarkable are the oxygen depleted continental margins, especially located along the north-western continental margin of America, Arabia and west India.

A significant influence of oxygen depleted bottom water on benthic respiration rates and TOC contents below 1000 m water depth was only determined for the NE-Pacific region. Nevertheless, the local importance of BOC can be identified by a comparison between the SE­Atlantic Ocean and the NE-Pacific Ocean (Fig. 3a,b). While already the low regression coefficient indicates the province specific relation between TOC and BOC, Kox -+ 0 indicates a negligible influence of BOC on DOU and TOC within the SE-Atlantic Ocean (Fig. 3a). The limited availability of oxygen in NE-Pacific bottom water shows a strong effect on the reduced organic carbon decay as expressed by an increased Kox (Fig. 3b).

70 The benthic carbon mineralization on a global scale

-180 -120 -60 o 60 120 180

< 20 40-60 80-100 120-140 160-180 200-220 240-260 280-300 >300

o in situ DOU data 0 in situ TOU data ~ ex situ DOU data

Figure 2: Distribution pattern of the bottom water oxygen concentration (BOC) in [mmol m-3] based on 6788

bottle data (WOCE, GEOSECS). Gray circles: in situ DOU data, white circles: in situ TOU data, triangles: ex situ DOU data. Only data used for regression analyses (wd > 1000 m) are displayed.

o in ~itu DOU data N-SWACO e In Situ [X)T J dal<l NOA'l1'/SOATL '{( In Situ rX)U (bt3 NAMRU)

[X)U

[mlllolnl' da~ 0

1]

o ill situ rx ){ J tlnl;'l l'-'EPAClNWAMCO

.g 8

I '"

40

35

3.0

25

2.0

1.5

III

0.5'

Figure 3: (a) Multiple regression analysis of the SE-Atlantic Ocean comprises the SWACO, NOATLlSOATL and NAMBCO provinces. (b) Multiple regression analysis ofthe NE-Pacific Ocean and the continental margin of NW-America (NWAMCO).

As documented by Cai and Reimers (1995), along the continental slope and rise of NW­America the TOC content decreases sharply with increasing BOC until an oxygen concentration of less than 35 mmol m-3 is reached. Between 35-125 mmol m-3 the TOC content remains relatively constant. Above 125 mmol m-3

, the organic carbon content decreases again with increasing BOC (Fig. 3b). This means, ifDOU is seen as fixed and BOC

The benthic carbon mineralization on a global scale 71

increases, the cOlTelative TOC content decreases. Like to be expected, highest TOC contents cOlTespond to lowest BOC values. If fresh water exchange provides high oxygen concentrations at the sediment-water interface constantly, high TOC concentrations cOlTespond to high BOC and no influence of BOC can be observed. Accordingly, the decay of organic matter for the most provinces could be derived from DOU and TOC by a simple 2D log-linear regression in these specific provinces.

Toe versus DOU (TOU)

From a first comparison, it is evident that a simple global correlation between TOC and DOU/TOU does not exist (Jahnke, 1996) (Fig. 4). Regional variabilities require more detailed regression analyses for specific areas. For this, we use the definition of 33 TOC-based benthic zones after Seiter et al. (subm.). While doing so, however, a very similar con'elation for single zones next to each other pennit their common view. For example, this applies to the western and eastern Arabian Sea (WARAB/EARAB) or the central northern and southern Atlantic Ocean (NOATLlSOATL). Results of regression analyses for zones where benthic DOU data are available are shown in figure (Sa I-XI) Coefficients are listed in table 2. For a better comparability, all regression fits are compiled in figure Sb.

7

6 ,.......,

>, 5 c:: --:::;

• t< r;l

,-< :::; 4 .,., • ~ -< ~ r-< .1 :::;

'---' • • • • ~ • ~

""" 2 '-'

• • 'W' • •

** r-., >---< * -5'*

* '* '* • Atlantic and Indic IXXJ data

o 2 ,.,

4 5 6 7 .1 * Pacific IX)U dat8

TOe [,\t(%]

Figure 4: Summary of all DOU data vs. TOe data used in this study. Filled squares indicate DOU data of the Atlantic Ocean (including Polar Seas) and Indian Ocean. Stars indicate data of the Pacific Ocean.

As depicted in figure Sa XII, no significant difference in the TOC/DOU relations was obtained between the European continental margin (EUR1) and the combination of the Rio de la Plata region and the continental margin off Argentina (RIOPLATA/ARGCO). The same applies to the NW-African continental margin (N-SWACO), the NE-American continental margin (NEAMCO) and the combinations of the continental margin off NW-America and Brazil (GUBRACO/BRAZCO). At the end 10 types of significant different TOC/DOU-

72 The benthic carbon mineralization 071 a global scale

relationships could be identified (Fig. 5a I-X, Tab. 2). They reflect the paiiicular

environmental boundary conditions, which can be interpreted as caused by differences in the

lability and availability of the paliiculate organic matter (POM). Both are closely coupled to the source of POM, as surface primary production or tenestrial riverine input, and the residence time of organic matter until burial. The last also depends on a combination of the

sedimentation rate and the total sedimentary composition (e.g. J ahnke, 1996). In this context, the transpOli of POM through the water column, and therefore the water depth itself, has an

additional effect of great significance. Accordingly, several recent studies, especially from the

eastern equatorial Pacific (TROP AC, Fig. 5a IX), give suppOli to the so called mineral association hypothesis, which means that the accelerated organic carbon export to the sea

floor is strongly determined by the presence of ballast minerals like calcite or lithogenic carriers (Berelson, 2002; Annstrong et aI., 2002; Klaas and Archer, in press). So, the steep

gradient for the sub-provinces of the Arabian Sea (W ARAB/EARAB; Figs. 5a I and 5b), indicating high benthic mineralization rates, may also be caused by the accelerated vertical

POM flux due to the well known intense eolian dust input from the deseli area of the Arabian peninsula and other Asian sources to this region (Ittekkot, 1993; Haake et aI., 1993; Schnetger

et aI., 2000, Siroko et aI., 2000). This rapid transpOli probably mediates the deposition of

more labile, easily degradable organic carbon.

Table 2: Coefficients for the functional relationship between DOU (TOU), TOC and BOC. Regression fits in bold were used for data processing.

No. Province Log- lincar decay of I" order Rational fUllction ofmonod kinetics

(Eq.4.2) (Eg.4.2)

DOU-R' 1\., K, 1\., DOU-R' 1\., K, K(,,,

I WARABi

0.72 11.99 -13.22 2.76 EARAB'

11 N-SWACO 0.73 4.10 -3.92 2.49

II NI,AMCO 0.:'5 1.46 0.73 0.5

11 G I J I3RACOiBRAZCO 0.86 4.87 -4.66 2.49

III CANARfWAITO 0.62 1.77 -0.34 1.23

IV RIO!'I.ATA/ARGCO 0.73 0.49 0.86 0.08

IV EURI' 0.63 1.43 -0.3 l.2

V NOATLiSOATI,' 0.72 0.28 0.83 -0.097 055' I 0.92 0.85

VI NAMBCO 0.92 0.95 1.21 0.11

VIf GROE' 0.88 7.91 -10.31 3.23

VIII LAPTEVSEA' 0.45 1.16 0.27 0.18

IX TROPAC 0.58 0.72 0.52 0.19

X CHICO/I'FRCO 0.89 2.86 -1.83 1.77

XI NEPAC/NWAMCO· 0.88 0.77 0.99 om 0.81 1.67 1.62 95.5

I: only moelelcel data based on in situ data were considered (EURI: Erping et aL 2002: W,\RAB/EARAB: Lutfct aI., 200() i. 2: regressioll fits cxcL oxygen consumption rates of: GeoB 1708-2 (Glud et al.. J 994) located at the Walvis Ridge.

K,

0.00

0.00

3: regression analyses based on ex situ anel in situ data (> 550 m wd), excluded data: Vering Plateau: VP6: Nonvegian continental margin: 36 jO 1 (Sauter ct aL, 200 I l, no tangents fit with NOATLlSOATL data. 4: NEPAC: 3D-regression applied. region includes the continental margin ofNW-;\ll1crica (NWAMCO), s. tigure .lb. 5: regression of the SE-Atlantic Ocean: NOATLlSOATI., NAMBCO. N-SWACO. s.figure .la. 6: EURI is not applicable on FUR2.

TangctHs exponential deca\ (Eq.S)

K TO(,,,,,, TOU-R' K1

- - 0.65 0.92

128 OJ2 0.32 134

1.49 OAg 0.94 0.90

- - - -

l.28 0.32 - -

1.28 0.32 - -

128 0.32 - -

l.28 0.32 DOlJ=TOll -

l.28 0.32 0.99 1.17

- - 0.94 0.20

- - - -

0.63 0.83 DOLJcTOLJ -

0.91 0.67 0,47 1.73

UlI 0.70 DOll=TOll -.. ,

transferred to pro\ inees

K,

1.3 I'leO

0.43 SEAFCO. SOfvl.AI .Ieo. S-

SWACO: TANZACO

0.85 Grid processed with

regression fil II

Grid processed with -regression lit IJ

- ETROPAL GUI

-

Oriel processed "ilh -regression fit IV

- IND. FIIR2"

0.57 Comparable to TROPAC

2.10

-

SEPAC/-SWPAC. -

TROPAC2

0.26

- NWPAC

::j ~ 0-

~ -.. :::s-r;'

~ Cl :::;

21

t N' s::, g. :::; Cl :::; l:l

Ckl c;-o­s::, -­c.., C') s::, ~

-.....l w

74

-;"

~. -:l " ~

~ r ;::; ~

:;:::

f-<

"0 :::: c::

-;" ;.-. .g

r. r

c 2 r ~

e-.. >-' r..

CS

R

6

-I

2 .0

0

-I I)

-' 2

0

r; ()

-I

2

()

()

-I

2

0

0

W.-\R.lJlE.o,RAB

I' /

" / "/ ,

/ .. " 2

RkJPL.-\ TA .-\RG('(J

2

2 CHIC(l·PERCO

3

considered in regression analvses • U1 SItl1 DI)U (lJtJ

-'

-:±1

-I

3

o ill ;1\U D,:>!.! doto N(IATL Sr.)A TL

A ex "tu DUll data lH iltll Telll d<'1i<1

The benthic carbon mineralization on a global scale

o 2 3 1.5

N('ATLS('ATL

10

0.5

().O -t-"'-r--r---,r-,--..,---,

o 2.0

0.5 1.0 L\PTEV Sf..\

15

075 1.00 125 1.50 NEPAC INWAMCO

2

o -t"'1Lr'=--r--,r-,--..,---,

o 2 -I (;

Ell ill "tu DOU dot" EIlRl

<> ill ;lnl D('U clota GUBR.-\C('BR-\zCO ..... III ;-:itn [hJU d<:lt8 NEA1J('()

leg-feSS1(lll tIt bil:-:ed (Ill IKfll dat<'l - - - Ie<J1e.:i'lOn tit hll-..!€,(\ Oil T(Jl T data

3

2

10 n.o 0.5. 10 1.5

8 NE.-\hICI) / ;' '"

(, -+' / / N.\"IBC<:> +/ I -+ //"

2

o 2.0

15 10

0.5

//Ii __ -r-""---

2 -+ TR':)PN:'

" 0-+-'14IjII.......,r--~-r--r---,

3

2 ",' /~~~~~-·-NE.--\r.I(--:Cl

.' .,." ___ .,_., EURI

excluded from regression analyses

~ E:X~ml D(-)U data IND (for comparison) ~ 111 ~'ltU IJ(IT) d~H{J S-S\\'.-\('I) (for comparison) [J In :,lhl DUll data GR()E

(excl. data from the VCil1ng Plateau and the cont. margin of Norway)

Figure 5 (a; I-XII): Regression plots in specific regions. Ordinate: DOU [tilinol m-2 day-I] and TOU [mmol m-2 day-I]; abscissae: TOC [wt%]. Black filled squares: in situ DOU data, cross: in situ TOU data. Regression in figure Sa-XI is not used for 2D-regression analyses (cf. Fig. 3b).

(i

5 ~.

-J.

.3

2

o t

I Tmlg:enl, ufNOATI.-·SOA1L fil

1 2 .3 4 5 6 Tor'. -mean tl)r NOA1L'SOA 11., fit ,Llta

lim

-:\ Lc:.!;end and r,lllp:e ofTIJC .!!lId ebln

I W_-\R'ill'E-\R'ill t) '.6-3.23 II. N-S\\:;l.CU 0.35-3.01

GUBR'..CO BR".ZC·O 0 19-1.15 NE.-\r\ICO 0.08-1.78

III CANARW."-F'CO. 0.09-1 9-1

IV RIOPLATlVARGCO 0.12- 3.36 EllRLO.23-L~3

\. NClATLSCIA1L 0.05-1.66 \ 1: NAI\IBO) 0.3 -4.&) \ 11 (TRUE (Ba'1l1) 0 (18-15-\1II L-\PTEY SEA 1l-\8-U3 IX: TROPAe: CI.12--I.--I X CHKDPERCO Cl 31-- 5S

:\1 NEPAC/NWAMCO 0.06-5.49

Figure 5 (b): Comparison of all regression fits with their validity by the range of TOC contents (grid-based). The numbers represent the respective regression types of table 2. The Black arrow marks TOC1im based on NOATLlSOATL fit data. The regression plot ofNEPACINWAMCO (fit XI) is not shown.

The benthic carbon mineralization on a global scale 7S

Other areas where intense terrestrial input seems to have a great influence may be the Polar Seas, especially off Svalbard and the region of the western Greenland Sea (Hulth et aI., 1996; Stein and Macdonald, 2003). The relation between DOU and TOC shows an exponential decrease of DOU related to increasing TOC (Fig. Sa VII*). This behaviour may indicate the transport of organic carbon across the slope with an increasing refractory fraction with distance to the coast. Ice-coverage and stratification of the water masses as well as high surface CUlTent velocities in the NE-Polyna reduce the vertical supply with fresh labile organic matter (Peinert et aI., 2001; Ritzrau et aI., 2001). Nevertheless, this area is of minor importance for the province budget of POCa and accordingly was not considered in further regression analysis.

For the basin area of the northern N-Atlantic and the Norwegian continental margin the positive TOC/DOU relation indicates that the overall degradation of organic matter is as high as in the Arabian Sea and freshly produced organic matter is rapidly transported from the euphotic zone by aggregate fonnation during bloom conditions or from lateral sources to the sea floor (Ritzrau et aI., 2001) (Figs. Sa I and Sa VII).

In contrast to the Arabian Sea, the Greenland-Norwegian-Iceland Sea (GROE) is not affected by lithogenic dust. Nevertheless, the importance oftelTestrial organic matter and therefore the refractory portion increases to the east (Thiede et aI., 1986; Hulth et aI., 1996; Dowdeswell et aI., 1998). However, high mineralization rates are caused by an intense seasonal phytoplankton production and rapid decay under well oxygenated benthic conditions (Ritzrau et aI., 2001). With our interpretation of TOC~ (cf. sect. 2.2, Eq. 6.1), highest amounts of refractory organic substance can be detected for the Laptev Sea (Tab. 4), which may correspond to the intense input of terrestrial material to this area by sediment-laden sea ice (Eicken, 2003) and riverine discharge of the Lena and Chatanga Rivers (Fig. Sa VIII).

The continental margin of SW-Africa (N-SWACO) has an intennediate position (Figs. Sa II and Sb). Both, high primary production - fostered by an intense riverine nutrient discharge - and the extensive input of lithogenic particles by the Ougoue River (O.SOS) and Congo River (6 0 S) strongly influences this benthic system. COlTespondingly, the terrigenous fraction of organic matter is estimated to S 8-7 6 % with significant amounts of low reactive mature organic matter (Wagner et aI., in press). Therefore, the high content ofless- and non­reactive particles results in enhanced preservation of organic matter and slightly lower DOU vs. TOC ratios in comparison with other provinces like the Arabian Sea or Polar Seas.

A system, which is absolutely dominated by the marine primary production, is the coastal upwelling off Namibia (NAMBCO; Figs. Sa VI and Sb). Here, the tremendous amount of POM leads to enhanced sedimentation rates, high TOC contents, and intense preservation of organic matter. As a result, the TOC/DOU ratio and the TOC concentration are positive correlated. Although sedimentary settings are quite different, this relationship is most extreme in the western Argentine Basin (RIOPLATA; Figs. Sa IV and Sb) and the central areas of the northern and southern Atlantic Ocean (NOATLlSOATL, Figs. Sa V and Sb). The first region is mainly characterized by intense lateral particle transport far from the south (Ewing et aI., 1964; Garzoli, 1993; Peterson et aI., 1996; Hensen et aI., 2000; Hensen et aI.,

76 The benthie carbon1l1ineralization on a global scale

2003) in combination with intense downslope transpOli along the continental margin. Thus, these processes result in the focus sing of rather less reactive and more lithogenic sediments in the Argentine Basin (e.g. Haese et aI., 1997; Romero and Hensen, 2002). Compared with this situation pelagic Oceans are generally characterized by low sedimentation rates and intense mineralization already within the water column. As a result, the benthic oxygen uptake in the deep sea is low.

The validity and accuracy of the regional fit functions can be verified by the calculation of DOU from TOC values (Fig. 6). The very good regression coefficient of R=0.93 gives evidence to our approach.

6.0

• ~I 4.0

'V '";,

e • RO.93 • • • ••

• 11=132 • 2.0 •• • •• ••• SD:0.34 • • • • • • P< 0.0001 •

slope: 1

• 0.0 ..,.....----r------r-----.---,..----r-----,

0.0 2.0 4.0 6.0

Figure 6: Estimation accuracy of all regression fits within 83.s oN-8SoS, given as estimated DOU (DOUest) vs. measured DOU (DOUmeas). Considered are all data (in situ DOU: 122 and ex situ DOU Greenland­Norwegian-Iceland Sea: 10). Regression fit=bisecting line.

Assuming that these 10 types of significant different TOC/DOU relationships and the special situation when limited BOC cannot be ignored are representative for the range of all benthic systems in the deep ocean, we are able to estimate benthic POC flux rates on a global scale. This was done by applying the results of regression analyses in the particular provinces and transferring them to those regions where DOU or TOU data do not exist. Infonnation on the decisive boundary conditions is a prerequisite for this. The main criteria used in this study to identify similar systems were the BOC, the appearance or absence of riverine input, the oceanographic settings (i.e. type of upwelling, current patterns, trophic situation), and the sedimentary settings (i.e. sedimentation rates, sedimentary composition; Fig. 7). Infonnation on the application of the single regression fits to other provinces are given in table 2.

The benthic carbon mineralization on a global scale 77

/ Organic carbon source ~

...... riverine Primary production eolian input input

1 1 1 1 ! ?~ c: Upwelling +

I Coastal I I Equatorial I I Confluence I

II riverioe aod e01iao I \ non-organic supply \ ! I Horizontal transpOlil+ Types Vertical transport . ..-

IAccelerate I I PreventIOn ----/ ~r----------'

Lateral sediment transport Depletion of

Erosion I I Deposition oxygen in bottom water

Figure 7: Transfer scheme for regression fits with main boundary conditions, which control the decoupling of TOe in surface sediments and the primary production.

3.2 Comparison of average rain rates

Comparisons of global budgets

Normally, the flux of particulate organic carbon to the sea floor (Jpoe) is calculated from sediment trap data (e.g. Betzer et aI., 1984; Berger et aI., 1989; Pace et aI., 1987; Antia et aI.,

2001). In contrast to this common method, information on benthic oxygen respiration rates

can be used to estimate the amount of organic carbon that accumulates at the sediment surface (e.g. Jahnke, 1996), because the oxygen uptake is directly correlated with the remineralization

of organic matter. Certainly, the portion that is preserved within the sediments cannot be

recorded by this approach. Therefore, our calculation represents a minimum-estimation of the

organic carbon rain rate (JPOCa). Since we only investigated water depths > 1000 m, the

influence of anaerobic respiration on JPOCa can be supposed as negligible.

The application of the specific transfer functions on the TOC and BOC grids results in DOU

estimates in a 10 x 10 grid-resolution. We used the molar ratio of aerob organic matter decay

(Redfield ratio) after Froelich (1979) of 1 06CI13 802 to convert these values into JPOCa (Fig. 8). In this way, we calculated a global JPOCa of ~0.50 GtC y{1 (Appendix II). This value

would correspond to ~ 1.6 % and 2.5 % of the primary production after Antoine et al. (1996)

in the deep-sea basins and along continental margins respectively.

78 The benthic carbon mineralization on a global scale

I I I I

-180 -120 -60 o 60 120 180

<0.25 0.75 1.25 1.75 2.25 2.75 3 3.5 4 4.5 5 6 8 12 20 > 20

Figure 8: Global distribution pattern of the calculated minimum flux of particulate organic carbon to the sea floor (JPOCa; 1° x 1°; > 1000 m wd).

For comparison, we recalculated the global Jpoc after Antia et al. (2001). Using the estimation of the primary production after Antoine et al. (1996) and the ETOP05 water depth grid Jpoc sums up to 0.44 GtC yr-1 (Appendix II). Additionally, the approach after Jahnke (1996), which is based on the correlation between the CaC03-corrected burial rate of organic matter and on the benthic oxygen consumption, was recalculated with 0.42 GtC yr-1 (600 S - 60oN; Appendix II, CI02 ratio 0.77). In principle, these three fundamentally different approaches give quite similar results. Figure 9 shows the comparison between areal integrated JPOCa estimations of this study with results after Antia et al. (2001). Only provinces with a significant influence of> 2'10-3 GtC yr- 1 on the global balance (83.5°N, 85°S) are depicted. Together, these represent about 95 % of the global ocean Jpoc. In Appendix II the calculated budgets within the benthic provinces are summarized.

The benthic carbon mineralization on a global scale

TAi\,ZACO ETROPAT

SEAFCO SWPAC SEPAC

TROPAC2 TROPAC

PERCO WARAB EARL1J3

NOATL SOATL

WAFCO "NWPAC

Il'm SWACO l\l£PAC

l\l\VA.MCO GROE

C=:J Jpocu add. this study

_ Jpoc after Antia et aL (2001)

~Tr~TrnT~nTrnTrnT~nT~~~rn~~Tn~~~rrn

-OJ) 1 0.00 0.01 0.D2 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.11

Flux ofpmticulate organic carbon to the sea floor [GtC yr -1]

79

Figure 9: Comparison between recalculated flux ofparticulate organic carbon to the sea floor (Jpoe) after Antia et al. (2001) in [Gt C yr- I

] (black bars) and calculated JPOCa (this study). GROE is calculated after Schliiter et al. (2000) and calculated JPOCa (this study). The gray fraction of the horizontal bar chart depicted the additional or less calculated portion of JpOCa. of this study. Deviations are depicted by gray bars. Provinces excluded « 0.002 GtC yr- 1

, NEAMCO, GUI, CHICO, GUBRACO, ARGCO, BRAZCO, CANAR, EURl, EUR2, RIOPLATA, NAMBCO, EICO, ANT, SOMALICO).

Most significant deviations in comparison with Antia et al. (2001) can be observed for the

Pacific Ocean. However, caused by their great extension the deep-sea basins of the northern

and southern Atlantic Ocean, the Pacific Ocean, and the Indian Ocean have the main impact

on global Jpoc bUdgeting. In total, they built up 84 % of the annual global amount of JpOCa,

which accumulates on 92 % of the global sea floor (> 1000 m wd, incl. equatorial areas). For

example, the total budget of the eastern tropical Pacific Ocean is in the order of the globally

integrated coastal budget of 14 % (0.07 GtC yr- I). Like to be expected, highest mean flux rates

are calculated for the main upwelling areas off Namibia, off Chile/Peru, and in the Arabian

Sea. A summary of mean flux rates in the specific provinces is shown in Appendix Ill.

However, apmi from the general good agreement of Jpoc estimates from sediment trap data,

primary production, and water depth with our results, conspicuous deviations have to be

discussed in more detail. For example, a relatively high JPOCa was calculated at the confluence

zone of the Malvinas and Brazil CUlTents. We mainly trace this feature to the current pattern in

this region, where intense flow velocities induce strong lateral transport processes (Hens en et

aI., 2000; Benthien and Muller, 2000; Frenz et aI., in press; Hensen et aI., 2003). This causes

winnowing (along the SW-slope off Argentine) and focus sing (at the slope off SE-Brazil and

within the Argentine Basin) of particles and therefore results in the decoupling of a vertical

80 The benthic carbon mineralization on a global scale

connection between primary production and sediment accumulation. Accordingly, Jpoc, which is based on primary production and trap data, may be dubious under such conditions. Another

difficulty for the simple assumption of a veliical link between surface water and the sea floor are variations in the intensity of mineralization processes. For example at the deep eastern

equatorial Atlantic JPOCa is lower than to be predicted from primary production by using a

common power model. The reason could be that the oxygen-rich North Atlantic Deep Water favours microbial degradation of organic matter already during settling. A similar high influence on organic carbon mineralization and accordingly a low preservation potential was

also proposed for the deep Guinea Basin by CUlTY and Lohmann (1990) and was recently confirmed by Mollenhauer et al. (in press). These observations support the idea to look on the

trap data in more detail.

Trap based hoc vs. JPOCa

Figure lOa contains direct comparisons between sediment trap data, hoc after Antia et al. (2001), and our estimations of JpOCa. For the most, there is a good accordance with our

estimations. The overall regression coefficient between measured particle fluxes and estimated JPOCa is given by R=0.69 (Figs. 10a,b). Considering the known biogeochemical and sedimentary processes that can affect the efficiency of trap collections, this correlation strikes

satisfactory on a first view. The estimations after Antia et al. (2001) and our estimations scatter around a regression curve with a slope of 1.1 and 0.96 respectively, which support our

result of a similar global budget of both approaches. Nevertheless, at single locations serious

differences appear.

8 Trap data vs. Jpoc recalculated after Antia et aL (2001) R0.43 0'

"75;., 7 SD: 1.8

slope: 1.1 6 p< 0.1

N:37 5

4

3 ~.

,-

'-'

i.. 2 • 'j

• • C. .~c C •

o

',.

• • iI •

Trap data vs. Jpc>c this study

RO.69 0, • SD:0.97 .~

0

.~

(J

0

slope: 0.96 p< 0.0001 ..

• • N:37 ,JI

.iIII 1,;', (:'

':j,

2 3 456 7 Trap data collected J

poc [g m-: yr-!]

< 550m above the sea floor

8

Figure 10 (a): Regression plot of JPOCa (this study) vs. hoc derived by traps (max. ~550 m above the sea floor) marked as black filled circles and Jpoc recalculations after Antia et al. (2001) vs. Jpoc derived by traps (marked as open circles). Trap sites excluded from regression analyses: MANOP _ H, MANOP _ S, MANOP _M; MSZ-11; epac; K12; PI; MW-I, M-S; EB-3, 6, 7 and all data from the Greenland­Norwegian-Iceland Sea. Solid line=bisecting line.

The benthic carbon mineralization on a global scale 81

One of the most striking deviations of J pOCu and trap data arises in the western tropical Pacific Ocean (sites Kl and K2, Fig. 10b No. 43 and 44). While sediment traps detected a flux of 5.15 gC m-2y(1 (Kl, s. Appendix IV) and 3.98 gC m-2yr-1 (K2, s. Appendix IV) respectively,

our results underestimate these values by a factor of ~0.5. The high Jpoc was mainly attributed

to the high nutrient supply by estuarine upwelling and deeper mixing of the upper water column (Kawahata, 2002). Furthermore, there is a tremendous addition of terrestrial riverine

suspensed load in this region (Kawahata, 1999; Milliman et aI., 1999; Kawahata and Murayama, 2000), which may favour a rapid downward transport associated with reduced

degradation within the water COlUlllil. Since we transferred regression fits for the entire tropical Pacific Ocean to this area, we are not able to consider local circumstances and may

underestimate the organic carbon flux to the sea floor in this area. In comparison, trap data from stations K7 (s. Appendix IV, Fig. lOb No. 47) and K5 (s. Appendix IV, Fig. lOb No. 50),

influenced by a high input of litho genic particles (Kawahata, 2002), nearly coincide with our estimations. The reason may be that for this province the influence of litho genic carrier minerals is considered in the transfer function.

60

o

·60

·180 ·120 -60

, ,

1- ',,,,V

-180 ·120 -60

• Locatiom of near bottom tnlp data

44 Number oftrap data Bar chart related to llllmber of near bottom trap data

o

o

o Jpo(' collected by particle traps « 550 m above the sea floor)

o J poc,. thi s stllCly 11 Jpac after Autia et al. (2001)

60 120 Greenland-N ofwegian-Ice!and Sea

180

47 -.50 I £i

.,.,.; Greenland-NoIwegian-Ice!and Sea

56 51".~ .. ~ ,,';;' '.:"

59_.58

, v

52 ·60

• 53

55 •

5\V

60

o

Figure 10 (b): Bar charts for comparisons of sediment trap data observations with Jpoca. of this study and flux reca1cu1ations after Antia et al. (2001). All trap data are numbered and summarized in Appendix IV.

82 The benthic carbon mineralization on a global scale

At sites LP] and ESTOC (not shown in Fig. lOb; s. Appendix IV), both located close to the Canary Islands, JPOCa vary from trap collections by a factor of ~3.4. Both sites are mainly characterized by intense atmospheric dust input and lateral pmiicle advection (Neuer et aI., 1997; We fer and Fischer, 1993; Freudenthal et aI., 2001). The latter is substantiated by time series data, which show pennanent higher flux rates in the lower traps (Lampitt and Antia, 1997; Neuer et aI., 1997). Therefore, the discrepancy between trap data and the JPOCa seems to be attributed to an additional lateral input of organic carbon within the benthic nepheloid layer, which is not recorded by the traps (Freudenthal et aI., 2001). A similar situation is known for the Arabian Sea (MS-4; MS-5, s. Appendix IV, Fig. lOb No. 41, 42), where sea­floor mineralization rates and thus JPOCa exceed the input rates, estimated from sediment traps (Lee et aI., 1998; Pfannlcuche and Lochte, 2000).

The trap data collections of trap stations EB-4 and EB-5 (s. Appendix IV, Fig. lOb No. 22, 23), deployed above the depot-center of organic carbon at the continental margin of NE­America, are well represented by our estimations. Certainly, far from the shore (EB-6 and EB-7; s. Appendix IV) trap based Jpoc is consistently higher as depicted from oxic mineralization rates (Biscaye et aI., 1988). This supports the idea of relatively rapid change of the benthic regimes to a higher portion of refractory organic matter towards the open ocean (Walsh, 1991; Walsh aI., 1991, Rowe et aI., 1994). However, at the near shore site EB-3 (s. Appendix IV) the underestimation of JPOCa against Jpoc could also be explained by our usage ofDOU instead ofTOU (Rowe et aI., 1994; cf. Fig. 5a VI).

In general, the main decoupling process between fluxes from trap data and our estimates may be the lateral transport of particles. At additional sites to those discussed above, such transpOli processes can be assumed. Therefore, JPOCa indicates the lateral supply of POC near the sea floor at sites DE, WA3 and WA] offshore Guyana and Brazil and P] of the NE-Pacific Ocean (s. Appendix IV and Fig. lOb in the above mentionend order No. 27,28,29,1).

3.3 The effect when using TOU instead of DOU

All calculations discussed so far are based on DOU measurements. But, like mentioned before, especially along continental margins the total oxygen uptake (TOU) can be several times higher due to bioirrigation of macrobenthic organisms (Glud et aI., 1994; Archer and Devol, 1992). In the following, we want to question the effect of this disregard for the calculation of JpOCa. Mean values of JPOCa for the specific provinces are diagrammed in figure 11 and summarized in table 3. For regions where sufficient in situ TOU data are available, we have also estimated the corresponding pmiiculate organic carbon flux to the sea floor (JPOCy, Tab. 3). Except using equation 5, the processing is the same as for JpOCa. Of course, generally JpOCy exceeds hoca at the continental margin provinces (Fig. 11). The greatest difference can be detected for the upwelling area off Namibia (NAMBCO), where a mean TOU of 5.1 mmol m-2d- 1 corresponds with an average rain rate of 17.2 gC m-2yr-1

.

Compared with the DOU based JPOCa of 5.6 gC m-2yr-1 a TOU:DOU ratio of 3.1 can be derived. For comparable regions, SWACO - adjacent to the north -, and the high productive areas off Chile (CHI CO) and the Arabian Sea, this ratio is 1.7, 2, and 1.6, respectively, which

The benthic carbon mineralization on a global scale 83

would go with the lower TOC contents of these sediments. But, it is not astonishing that zones

with other benthic and sedimentary conditions do not fit to this supposed simple trend.

" :" ~~ '- ~ -, =

; 2 'lJ ""

:2

20.0

19.0 Pc'flmean POC, based on DOlT in situ data

lKO EZ2Z3 mean POC, based on TOl T in situ data

17.0

16.0

15.0

14.0

13.0

12.0

1l.l) 10.0

9.0

KO

7.0

6.U

5.0

4.0

3.0

2.0

l.0

00 -ILI'~~~~~.lf::i\f2.\l~~

Figure 11: Mean JPOCa (queried from the grid, Fig. 8) to the sea floor (gray bars), and mean JpOCy

(white bars, Eq. 5). LAPTEV SEA is excluded.

So, TaU :DOU ratios of 2.6 and 1.5 seem to be characteristic for the NE-American

continental margin (NEAMCO) and the Greenland-Norwegian-Iceland Sea (GROE),

respectively, where the mean TOC contents are comparatively low (0.65 and 0.72 wt%). The

reason may be the same as discussed above (cf. sect. 3.1). A global relationship between JPOCa or hocy and TOC does not exist. However, due to reduced activity of benthic organisms, no

significant differences between JPOCa and hocy can be observed within the deep-sea basins of

the Atlantic Ocean and the NE-Pacific Ocean (TOU:DOU::::: 1). Because these constitute

~80 % of the global sea floor, the extrapolation of our JpOCy estimates to the entire ocean yield

a minimum organic carbon rain rate of approximately 0.60 GtC y(1. The arising factor of 1.2

when using DOU instead of TaU for the calculation of the global benthic carbon flux

correspond very well with results for the Atlantic Ocean only (WenzhOfer and Glud, 2002).

84 The benthic carbon mineralizatiol1 on a global scale

Table 3: Minimum, maximum, and mean values of JPOCa (queried from grid), standard deviation of JpOCCl (SD), mean Jpocr and ratio of JPOC/lroca (LAPTEV SEA excluded).

JPOCa JpOC<L Jpoc" SD2 Jpo('·/ Jl'oc/Jpocu

Province (min) (max) (mean) (Jpocu-mean) (mean) (mean) [g m-2 yr- I

] [g m-2 y(l] [g m-2 yr- I ] [g m-2 yr- I] [g m-2 y(l]

NWAMCO 0.87 7.64 3.77 1.14 Jpoc ({" JpOC'i

NEPAC 0.21 6.53 1.48 0.76 .1 pocu .- Jpoc.

NWPAC 0.23 5.60 1.80 0.90 no data

Cl-IlCO O.4B 9.93 4.41 ., -,' ..:....,_.) 8.63 1.96

PERCO 3.90 13.63 9.10 1.86 no data

TROPAC 0.27 HO 2.36 1.15 no data

(E-Equal. Pacitic) TROPAC2

0.40 3.20 1.59 0.67 no data (W-Equal. Pacific)

SEPAC 0.35 8.25 1.12 0.67 no data

SWPAC 0.63 2.72 1.69 0.51 no data

SOATL 0.17 5.99 1.49 0.49 JPOcu. - Jpou

SWACO' 1.26 10.70 4.40 1.92 7.63 1.73

NAMBCO 1.26 10.74 5.56 2.51 17.2 3.10

HROPAT 0.17 10.37 2.60 1.49 110 data

ARGCO 0.76 3.26 1.62 0.46 no data

RIOPLATA 1.10 4.94 2.62 0.69 no data

BRAZCO 0.54 2.88 1.78 0.47 no data

GU13RACO 1.02 2.36 1.66 0.22 110 data

NOATL 0.58 4.11 1.41 0.44 Jpo('(t "'- Jpoc"(

EURl 1.25 3.77 2.27 0.46 no data

EUR2 0.23 2.71 1.62 0.42 110 data

WAFCO 0.52 5.67 2.38 l.11 110 data

GUJ 0.62 5.22 3.78 1.20 no data

CANAR 1.20 3.70 2.40 0.46 no data

NEAMCO 0.05 4.97 2.01 1.16 5.16 2.57

IND 0.28 9.82 1.42 0.58 no data

SEAFCO 0.28 6.85 1.87 0.75 110 data

TANZACO 1.53 3.98 2.98 0.49 no data

SOMALlCO 2.19 4.22 2.87 0.44 110 data

LICO 1.27 14.12 9.21 2.60 no data

WARAB 2.55 28.59 11.93 5.81 18.41 J.54

EARAB 2.22 29.62 11.71 6,90 18.41 1.57

()ROE <0.25 7.97 1.86 1.12 2.86 1.54

1: incl. NAMBCO (upwding province off Namibia. Fig. I b) 2: standard deviation of Jpoc,<,-mcan

3.4 Estimation of mean burial rates

As described above, the mean refractory organic carbon content (TO CB) for single provinces was derived from the relation between DOU and TOC (cf. Fig. Sa I-XI). Thus, based on the

assumption that no oxygen uptake occurs further more, we interpret the intersection point of

the abscissae as an indicator for TOCB in wt%. Accordingly, mean burial rates were calculated by multiplying TOCB with sedimentation rates (SR) and the dry bulk density (DBD), (cf. sect.

2.2; Eq. 7.0, Tab. 4).

The benthic carbon mineralization on a global scale 85

For the range of TOC~ we calculated a lower and upper limit (Tab. 4). The lower limit is the

TOCr) value derived by the regression fits depicted in Figs. 5a I-XI and 5b and summarized in

table 4. In the calculation of these lower limits we considered the adjacent basin data, if a

province borders on the deep-sea basin to characterize the transition between the continental

margin and basin areas. The upper limit of TOCfl characterizes the refractory organic carbon

content within a continental margin province, calculated without the basin data during

regression analyses (cf. sect.2.2). In table 4 the mean SR, the TOC~ range, and the standard

deviation of the SR (SD) for appropriate benthic provinces are summarized. For the

calculation of the global budget of buried organic matter we calculated a range of the burial

rate by using the lower and upper limits of SR and TOC~, respectively. For the extrapolation

of the global ocean budget, the relative proportion of the specific province areas were

considered adequately.

As a result, the burial rates are low for the basin areas of the northern Pacific Ocean and the

southern and n0l1hern Atlantic Ocean and increase to the continental margins. For the eastern

tropical Pacific Ocean and the n0l1hern and southern Atlantic Ocean, we estimate an average

burial rate of 0.03-0.04 gC m-2yr-1. For the NE-Pacific Ocean, this value is slightly lower

(0.02 gC m-2yr-1). Highest rates are derived for the Arabian Sea and along the continental

margin of SW-Africa (SWACO) with 0.19-0.34gCm-2yr- 1 and 0.07-0.48gCm-2yr- 1,

respectively. However, all these values are based on mean TOC~ and mean SR in the

conesponding provinces. Thus, regional and seasonal variabilities are not considered and the

abbreviation of the mean values of specific areas might be tremendous in some exceptional

cases. Overestimation of mean burial rates can occur due to disregarding anaerobic

mineralization in the sediments and non-diffusive fauna-mediated oxygen uptake in

predominantly coastal sediments. Areal integration results in a global organic carbon burial

rate range of approximately 0.002-0.12 GtC yr- 1• Accordingly ~0.01-0.4 % of the primary

production is stored in the sediments of the predominantly oligotrophic open ocean areas and

withdrawn from the global cycle for a longer period of time. To the comparison, the global

lpoca budget is accounted for ~ 1.6 % of the surface primary production (Antoine et al., 1996).

If we adopt the organic carbon flux through the 1000 m depth horizon after

lahnke (1996; 0.86 GtC yr-1), ~40 % of this amount is degraded in the deeper water column.

With other estimates of the flux leaving the surface mixed layer

(Christens en, 2000: 2.3 GtC yr-1; Antia et al., 2001: 2.47 GtC yr- 1

) this value would double to

~80 %. Nevertheless, detailed re-evaluations, concerning the mineralization in the deeper

water column, are not addressed at this paper. However, undoubtedly, deduced from the

relationship between lpoca and TOC~ the benthic mineralization rate is immense.

Table 4: TOC!l calculations fixed and not fixed with basin data of the northern and southern Atlantic Ocean (TOCB=intersection point with the abscissae), mean sedimentation rate (SR) calculated for each province and standard deviation of SR (SDSR), calculated burial rate W) and range of TOC data.

TOCpl TOel\" SR' SDSR

Provinz (mean) [wt%] [wt%] [cm kyr- I

] [cm kyr- I]

WARABfEARAB 0.25 0.36 14.5 (4) ilIA (5) 5.8!5.3

N-SWACO 0.11 0.77 96 (35) 18.5

NEAMCO 0.11 6.0 (2) 0

GUBRACOfBRAZC( ) (l.II 5.8 (7) 130 (5) 4.0!l.89

CANAR!WA FCO

RIOPLAT Ai ARGCO O,C19 OJJ 23.1 (5) 13.9 (11 184!-

EURl 0.033 812 (1)

NOATLiSOAT[, 0.15 3.4 (113) 14.0 (95) 3.lIU

NAMBCO 0.17 17.3 (12) 29.7

GROE 0.21 0..15 4.6 (81) 10.8

LAPTEV SEA 0.61 19.0 en) 13.1

TROPAC 0.30 16 (32) 0.91

NEPACiNWAMCO 0.25 1 16 (5)/- lA

CHICO/PERCO (U-I 4.9 (I) 16.5 (2) -/3.5

1: mean refractory organic carbon based on regression Ilts with NOATLlSOATL fixed data~lo\Ycr limit. 2: mean refractor) organic carbon hased on regression fits "ithout NOATI JSOA rI .. data =upper limit.

~4

[g m -2 yr- l ]

0.19-0.27 I 0.24-0.34

O,()7 -0.48

0.04

0.14-0.47

0.016

0.03/0.04

0.19

(l.O6-0.12

0.76

0.03

0.02

0.003/0.004

3: mean sedimentation rate of each province, numbers in brackets: number ofdala points. SDsR standard deviation of SR. 4: organic carbon burial rate, based on mcan TOe I' (I _'l, and mean sedimentation rates adopted D'om the literaturc. 5: all calculations for water depths> 1000 m. min. max and mean values ofTOC are based on the projected grid 10 x 10 ailer Stilcr et al. (submitted), CiROE: water depths> 550 Ill.

TOC' TOe' TOC' (min) (max) (mean) [wt%] [wt%] [wt%]

04910.36 4.28/3.23 1.36/1.48

OJ5 3.01 1.21

0.08 1.75 (l.65

O.l9!O.18 l.OI/US O.43/0.5-l

O.32!OO9 104/1.94 O.62JO.64

0.26/0.12 3.36/0.77 0.941038

0.13 1.83 0.78

0.05/0.06 1.111.66 0.36/0.35

030 -1.86 1.87

0.08 1.57 o.n 0,48 153 0.97

0.12 4.7-1 1.22

0.06/0J 1.74/5.49 0.44/0.75

0.3 1/1.63 3.9/7.58 1.45/3.82

00 0\

:J ~ c::,­~

~ ::0-r;' r:, I::l ;; Cl ::::

~. :::: ~

~ hi' ~ o· :::: Cl :::: I::l

(}q

Cl c::,-~

'" r:, I::l (t

The benthic carbon mineralization on a global scale 87

4. Conclusions

In this study, we present a global distribution pattern of the minimum particulate organic

carbon amount that settles on the sea floor (lpOCa). The estimation bases on the relation between the diffusive oxygen flux across the sediment-water interface (DOU), the total

organic carbon content in surface sediments (TOC), and the oxygen concentration in bottom water (BOC). The great advantage and improvement of our method is the view from the

benthic side under consideration of different sedimentary and geochemical regimes in 11 benthic provinces. From this point of view, regional varying relations between the TOC, DOU

and BOC are considered. Differences in the availability and quality of organic matter or in transport mechanisms through the water column are reflected by regression fits. Thus,

regional lateral supply or winnowing of sedimentary organic carbon affect the vertical link

between surface water and sea floor and could be taken into account. This is chearly reflected by deviations of JPOCa estamations from common distribution patterns of pp and near bottom trap data. On a global view however, our estimations are in good accordance with former

investigations (Jahnke, 1996; Antia et aI., 2001). The global JPOCa budget was estimated to ~0.5 GtC y(l (> 1000 m wd), whereas approximately 0.002-0.12 GtC yr- 1 are buried in the

sediments (0.01-0.4 % of surface primary production). Nevertheless, our approach of

regionalization and transferring offers new opportunities in estimations of biogeochemical

budgets.

Acknowledgements

We would like to thank H. Hecht, K. Pfeifer and N. Riedinger for their valuable and critical

comments during the development of this manuscript. We gratefully appreciate the support

from M. Diepenbroek, L. Gerullis and R. Sieger (Pangaea - Network for Geological and Environmental Data). Many thanks also to all the unnamed colleagues who do not hesitate to

place their data to our disposal. This research was funded by the Deutsche

Forschungsgemeinschaft (DFG, ZA 19911- 1).

88 The benthic carbon mineralization on a global scale

References

Antia, A.N., W. Koeve, G. Fischer, T. Blanz, D. Schulz-Bull, 1. Scholten, S. Neuer, K. Kremling, 1. Kuss, R.

Peinert, D. Hebbeln, U. Bathmann, M. Conte, U. Felmer, and B. Zeitschel, Basin-wide particulate

carbon flux in the Atlantic Ocean: Regional export patterns and potential for atmospheric CO2

sequestration, Global Biogeochel11. Cycles, 15,845-862, doi: 200GB001376, 200l.

Antoine, D., J.-M. Andre, and A. Morel, Oceanic primary production; 2. Estimation at global scale from satellite

(coastal zone color scanner) chlorophyll, Global Biogeochem. Cycles, la, 57-69, doi: 95GB0283l,

1996.

Archer, D., and A Devol, Benthic oxygen fluxes on the Washington shelf and slope: A comparison of in situ

microelectrode and chamber flux measurements, Lil11nol. Oceanogr., 37(3),614-629, 1992.

Archer, D. E., An atlas of the distribution of calcium carbonate in sediments of the deep sea, Global

Biogeochem. Cycles, la, 159-174, doi: 95GB030l6, 1996.

Armstrong, R.A., C Lee, 1.1. Hedges, S. Honjo, and S.G. Wakeham, A new mechanistic model for organic

fluxes in the ocean based on the quantitative association of POC with ballast minerals, Deep Sea Res.,

Part 11,49,219-236,2002.

Arthur, M.A., W.E. Dean, and K. Laarkamp, Organic carbon accumulation and preservation in surface sediments

on the Peru margin, Chem. Geol., 152,273-286, 1998.

Behrenfeld, MJ., and P.G. Falkowski, A consumer's guide to phytoplankton primary productivity models,

Lil11nol. Oceanogr., 42,1479-1491, 1997a.

Behrenfeld, MJ., and P.G. Falkowski, Photosynthetic rates derived from satellite-based chlorophyll

concentration, Lil71nol. Oceanogr., 42, 1-20, 1997b.

Benthien, A., and PJ. MUller, Anomalously low alkenone temperatures caused by lateral particle and sediment

transport in the Malvinas CUlTent region, western Argentine Basin, Deep Sea Res., Part 1,47(12),2369-

2393,2000.

Berelson, W.M., D.E. Hammond, D. O'Neill, X.-M. Xu, C Chin, and ]. Zukin, Benthic fluxes and pore water

studies from sediments of the central equatorial north Pacific: Nutrient diagenesis, Geochim.

Cosmochil11. Acta, 54,3001-3012,1990.

Berelson, W., Particle settling rates increase with depth in the ocean, Deep Sea Res., Part 11,49,237-251,2002.

Berger, W.H., V.S. Smetacek, and G. Wefer, Ocean productivity and paleoproductivity - An overview, in

Pl'Oductivity o/the Ocean: Present and Past, edited by W.H. Berger et aI., pp. 1-34, John Wiley, New

York,1989.

Betts, 1.N., and H.D. Holland, The oxygen content of ocean bottom waters, the burial efficiency of organic

carbon, and the regulation of atmospheric oxygen, Palaeogeogr. Palaeoclimatol. Palaeoecol., 97, 5-18,

1991.

Betzer, P.R., WJ. Showers, E.A Laws, CD. Winn, G.R. DiTullio, and P.M. Kroopnick, Primary productivity

and particle fluxes on a transect of the equator at 153°W in the Pacific Ocean, Deep Sea Res., 31, 1-11,

1984.

Biscaye, P.E., R. Anderson, and B.L. Deck, An indication of lateral input of biogenic detritus, Deep Sea Res.,

Part 11, 41,657-668,1988.

Bory, A, C ]eandel, N. Leblond, A. Vangriesheim, A Khripounoff, L. Beaufort, C. Rabouille, E. Nicolas, K.

Tachikawa, H. Etcheber, and P. Buat-Menard, Downward particle fluxes within different productivity

regimes off the Mauretanien upwelling zone (EUMELI program), Deep Sea Res., Part 1,48,2251-2282,

200l.

The benthic carbon lI1ineralization on a global scale 89

Boudreau, B. P., Diagenetic models and their implication: Mode11ing transport and reactions in aquatic

sediments, pp. 142-159, Springer-Verlag, Berlin-Heidelberg, Germany, 1997.

Cai, W.-I., and C.E. Reimers, Benthic oxygen flux, bottom water oxygen concentration and core top organic

carbon content in the deep northeast Pacific Ocean, Deep Sea Res., 42(10), 1681-1699, 1995.

Calveli, S.E., and T.F. Pedersen, Organic carbon accumulation and preservation in marine sediments: How

important is anoxia?, in Organic matter: Productivity, accumulation, and preservation in recent and

ancient sediments, edited by lK. Whelan, and l W. Farrington, University press, 1992.

Christensen, I.P., A relationship between deep-sea benthic oxygen demand and oceanic primary productivity,

Oceanolog. Acta, 23,65-82,2000.

Cobler, R., and Dymond, l, Sediment trap experiment on the Galapagos, Equatorial Pacific, Science, 209, 801-

803.

Curry, W.B., and G.P. Lohmann, Reconstructing past particle fluxes lJ1 the tropical Atlantic Ocean,

Paleoceanography, 5,487-505,1990.

Dowdeswell, lA., A. Elverhoi, and R. Spielhagen, Glaciomarine sedimentary proceses and facies on the polar

North Atlantic Margins, Quatern. Sc. Rev., 17,243-272, 1998.

Dymond, l, and M. Lyle, Flux comparisons between sediments and sediment traps in the eastern tropical

Pacific: Implications for atmospheric CO2 variations during the Pleistocene, Limno1. Oceanogr., 30,

699-712,1985.

Dymond, l, and R. Collier, Biogenic paliicle fluxes in the equatorial Pacific: Evidence for both high and low

productivity during the 1982-1983 El Nino, Global Biogeochem. Cyc1es, 2, 129-137, 1988.

Dymond, J., and M. Lyle, Partic1e j!uxes in the ocean and implications for sources and preservation of ocean

sediments. Materialj!uxes on the slllface of the Earth, National Academy Press, Washington DC, 125-

142, 1994.

Dymond, l, E. Suess, and M. Lyle, Barium in deep-sea sediment: A geochemical proxy for paleoproductivity,

Paleoceanography, 7,163-181,1992.

Eicken, H., The role of arctic sea ice in transporting and cycling terrestrial organic matter, in The organic cabon

cyc1e in the Arctic Ocean, edited by R. Stein, and R.W. MacDonald, pp. 45-53, Springer-Verlag, Berlin,

Gennany, 2003.

Emerson, S., and M. Bender, Carbon fluxes at the sediment-water interface of the deep-sea: calcium carbonate

preservation,1. Mar. Res., 39, 139-162, 1981.

Emerson, S., K. Fischer, C. Reimers, and D. Heggie, Organic carbon dynamics and preservation in deep-sea

sediments, Deep Sea Res., 32, 1-21, 1985.

Epping, E.H.G., C. v.d. Zee, K. Soetaert, and W. Helder, On the oxidation and burial of organic carbon in

sediments of the Iberian Margin and Nazare Canyon (NE Atlantic), Progr. Oceanogr., 52, 399-431,

2002.

Ewing, M., WJ. Ludwig, and l1. Ewing, Sediment distribution in the oceans: The Argentine Basin, 1. Geoph.

Res., 69,2003-2033, 1964.

Fischer, G., G. Ratmeyer, and G. Wefer, Organic carbon fluxes in the Atlantic and Southern Ocean: relationship

to primary production compiled from satellite radiometer data, Deep Sea Res., Part 11, 47, 1961-1997,

2000.

Francois, R., S. Honjo, SJ. Manganini, and G.E. Ravizza, Biogenic barium fluxes to the deep sea: Implications

for paleoproductivity reconstruction, Global Biogeochem. Cyc1es, 9(2),289-303, 1995.

90 The benthic carbon mineralization on a global scale

Frenz, M., R. Heinrich, R. Hoppner, lB. Stuut, and T. Wagner, Surface sediment bulk geochemistry and grain

size composition related to oceanic circulation along the South American continental margin in the

Southwest Atlantic, in The S01lth Atlantic in the late Q1Iaternary: Reconstr1lction o/material b1ldget in

current systems, edited by G. Wefer et aI., Springer, Germany, in press.

Freudenthal, T., S. Neuer, H. Meggers, R. Davenport, and G. Wefer, Influence of lateral particle advection and

organic matter degradation on sediment accumulation and stable nitrogen isotope ratios along a

productivity gradient in the Canary Island region, Mar. Geo!., 177,93-109, 200l.

Froelich, P.N., G.P. Klinkhammer, M.L. Bender, N.A. Luedtke, G.R. Heath, D. Cullen, P. Dauphin, D.

Hammond, B., Hartman, V., Maynard, Early oxidation of organic matter in pelagic sediments of the

eastern equatorial Atlantic: suboxic diagenesis, Geochil11. Cosl11ochil11. Acta, 43, 1075-1088, 1979.

Garzoli, S.R., Geostrophic velocity and transport variability in the Brazil-Malvinas Confluence, Deep Sea Res.,

Part 1,40, l379-1403, 1993.

Giraudeau, l, P.A. Meyers, and B.A. Christensen, Accumulation of organic and inorganic carbon in Pliocene­

Pleistocene sediments along the southwestern African margin. Mar. Geol, 180,49-69,2002.

Glud, R.N., lK. Gundersen, N.P. Revsbech, and B.B. Jorgensen, Effects on the benthic diffusive boundary layer

imposed by microelectrodes, Lil11no!. Oceanogr., 39,462-467, 1994.

Haake, B., V. Ittekkot, T. Rixen, V. Ramaswamy, R.R. Nail', and W.B. Curry, Seasonality and interannual

variability of particle fluxes to the deep Arabian Sea, Deep Sea Res., Part I, 40, l323-l344, 1993.

Haese, R.R., K. Wallmann, A. Dahmke, U. Kretzmann, P.J. MUller, and H.D. Schulz, Iron species determination

to investigate early diagenetic reactivity in marine sediments, Geochim. Cosmochim. Acta, 61, 63-72,

1997.

Hartnett, H.E., RG. Keil, J.1. Hedges, and A.H. Devol, Influence of oxygen exposure time on organic carbon

preservation in continental margin sediments, Nat1lre, 391,572-574,1998.

Hebbeln, D., Flux of ice-rafted detritus from sea ice in the Fram Strait, Deep Sea Res., Part I, 47, 1773-1790,

2000.

Hedges, J.1., Global biogeochemical cycles: Progress and problems, Mar. Chem., 39, 67-93,1992.

Hedges, J.1., and RG. Keil, Sedimentary organic matter preservation: an assessment and speculative synthesis,

Mar. Chem., 49,81-115,1995.

Hensen, c., M. Zabel, and H.D. Schulz, A comparison of benthic nutrient fluxes from deep-sea sediments of

Namibia and Argentinia, Deep Sea Res., Part 11, 47,2029-2050,2000.

Hensen, c., M. Zabel, K. Pfeifer, T. Schwenk, K. Kasten, N. Riedinger, H.D. Schulz, and A. Boetius, Control of

sulfate pore-water profiles by sedimentmy events and the significance of anaerobic oxidation of methan

for the burial of sulfur in marine sediments, Geochim. Cosmochil71. Acta, 67, 2631-2647, 2003.

Honjo, S., S. Manganini, and J. Cole, Sedimentation of biogenic matter in the deep ocean, Deep Sea Res., 29,

609-625, 1982.

Honjo, S., J. Dymond, W. Prell, and V. Ittekkot, Monsoon-controlled export fluxes to the interior of the Arabian

Sea, Deep Sea Res., Part If, 46, 1859-1902, 1999.

Honjo, S., J. Dymond, R. Collier, and S.l Manganini, Export production of pmiicles to the interior of the

equatorial Pacific Ocean during the 1992 EqPac experiment, Deep Sea Res., Part JJ, 47 (2-3),831-870,

1995.

Hulth, S., P.O.J. Hall, T.H. Blackburn, and A. Landen, Arctic sediments (Svalbard): Pore water and solid phase

distributions ofC, N, P and Si, Polar Bio!., 16,447-462,1996.

Ittekkot, V., Global trends in the nature of organic matter in river suspensions, Nat1lre, 332,436-438, 1988.

The benthic carbon 111ineralization 011 a global scale 91

Ittekkot, V., The abiotically driven biological pump in the ocean and short-term fluctuations in atmospheric CO2

contents, Global Planet. Change, 8, 17-25, 1993.

lahnke, R.A., CE. Reimers, D.E. Craven, Intensification of recycling of organic matter at the sea floor near

ocean margins, Natllre, 348, 50-54, 1990.

Jahnke, RA., The global ocean flux of particulate organic carbon: Areal distribution and magnitude, Global

Biogeochem. Cycles, 10, 71-88, doi: 95GB03525, 1996.

Kawahata, H., Fluctuations in the ocean environment within the western Pacific wann pool during late

Pleistocene, Paleoceanography, 14,639-652, 1999.

Kawahata, H., Suspended and settling particles in the Pacific, Deep Sea Res., Part 11, 49,5647-5664,2002.

Kawahata, H., and M. Murayama, Radiacarbon of settling particles from hemipelagic region, Nuc!. Illstr. Meth.

B, 172,485-489,2000.

Klaas, C, and D.E. Archer, Association of sinking organic matter with various types of mineral ballast in the

deep sea: Implications for the rain ratio, Global Biogeochem. Cycles, in press.

Kreutz, M., Stofftransport durch die Bodengrenzschicht-Regionalisierung und Bilanzierung fUr den Nordatlantik

und das Europaische Nordmeer, Ph.D. dissertation, Kiel, GEOMAR / Christian-Albrechts Universitat,

1994.

Lampitt, RS., and A.N. Antia, Particle flux in the deep seas: Regional characteristics and temporal variability,

Deep Sea Res., Part 1,44,1377-1403,1997.

Lasaga, A.C, and H. Ohmoto, The oxygen geochemical cycle: Dynamics and stability, Geochil71. Cosmochil11.

Acta, 66,361-381,2002.

Lee, C, R.W. Murray, R.T. Barber, K.O. Buesseler, 1. Dymond, 1.1. Hedges, S. Honjo, S. Manganini, 1. MalTa,

C Moser, M.L. PetersOl1, W. Prell, and S.G. Wakeham, Particulate organic carbon fluxes: compilation

ofresults from the 1995 US JGOFS Arabian Sea Process Study, Deep Sea Res., Part 11, 45,2489-2501,

1998.

Ludwig, W., J.-L. Probst, and S. Kempe, Predicting the oceanic input of organic carbon by continental erosion,

Global Biogeochel11. Cycles, 10, 23-41, doi: 95GB02925, 1996.

Luff, R, K. Wallmann, S. Grandel, and M. SchliHer, Numerical modeling of benthic processes in the deep

Arabian Sea, Deep Sea Res., Part 11, 47, 3039-3072, 2000.

Lyle, M., A. Mix, C Ravelo, D. Andreasen, L. Heusser, and A. Olivarez, Millennial-scale CaC03 and C-org

events along the northern and central California margin: stratigraphy and origins, in Proceedings of the

Ocean Drilling Program, Sc.Res., 167, edited by M. Lyle, 1. Koizumi, T.C Moore, jr., and C Richter,

pp. 163-182, Ocean Drilling Program, College Station TX, 2000.

McManus, 1., D.E. Hmmnond, W.M. Berelson, T.E. Kilgore, DJ. Demaster, O.G. Ragueneau, and R.W. Collier,

Early diagenesis of biogenic opal: Dissolution rates, kinetics, and paleoceanographic implications, Deep

Sea Res., Part 11,42,871-903, 1995.

Meybeck, M., C, N, P and S in rivers: from sources to global inputs, in Interactions of C, N, P and S:

Biogeochemical cycles and global change, edited by R. Wollast et aI., 163-193, Springer, Berlin -

Heidelberg - New York, 1993.

Milliman, 1.D., PJ. Troy, W.M. Balch, A.K. Adams, Y.-H. Li, and F.T. Mackenzie, Biologically mediated

dissolution of calcium carbonate above the chemicallysocline?, Deep Sea Res., Part 1, 46, 1653-1669,

1999.

Mollenhauer, G., R.R. Schneider, P.1. Mliller, V. SpieD, and G. Wefer, Glacial/interglacial variability in the

Benguela Upwelling System: Spatial distribution and budgets of organic carbon accumulation, Global

Biogeochem. Cycles, 16, 1134, doi: 10102912001GB001488, 2002.

92 The benthic carbon mineralization on a global scale

Mollenhauer, G., R.R. Schneider, T. lenneljahn, P. 1. Muller, and G. Wefer, Organic carbon accumulation in the

South Atlantic Ocean: Its modern, mid-Holocene and Last Glacial basin-wide distribution, Global

Planet. Change, in press.

Morse, 1.W., and K.e. Emeis, Controls on C/S ratios in hemipelagic upwelling sediments, Amer. 1. Sci., 290,

1117-1135,1990.

Muller, PJ., and E. Suess, Productivity, sedimentation rate, and sedimentary organic matter in the oceans, 1.

Organic carbon preservation, Deep Sea Res., 26,1347-1362,1979.

Murray, R.W., and M. Leinen, Scavenged excess aluminum and its relationship to bulk titanium in biogenic

sediment from the central equatorial Pacific Ocean, Geochim. Cosmochim. Acta, 60,3869-3878, 1996.

Murray, R.W., M. Leinen, and A.R. Isern, Biogenic flux of Al to sediment in the central equatorial Pacific

Ocean. Evidence for increased productivity during glacial periods, Paleoceanography, 8, 651-670,

1993.

Neuer, S., V. Ratmeyer, R. Davenport, G. Fischer, and G. Wefer, Deep water particle flux in the Canary Island

region: seasonal trends in relation to long-term satellite derived data and lateral sources, Deep Sea Res.,

Part J, 44, 1451-1466, 1997.

OMEX-I, Ocean Margin Exchange, OMEX-I project data set (1993-1994) 2 CD ROM, Natural Environmental

Research Council, published by British Oceanographic Data Center, Bidston Observatory, Birkenhead,

Merseyside, 1997

Pace, M.L., G.A. Knauer, D.M. Karl, and 1.H. Martin, Primary production, new production and vertical flux in

the eastern Pacific, Nature, 325, 803-804, 1987.

Peinert, R., E. Bauerfeind, R. Gradinger, O. Haupt, M. Krumbholz, 1. Peeken, 1. Werner, and B. Zeitzschel,

Biogenic particle sources and vertical flux patterns in the seasonally ice-covered Greenland Sea, in The

northern North Atlantic, edited by P. Schafer et al., 69-79, Springer, Berlin-Heidelberg, 2001.

Pena, M.A., K.L. Denman, 1.R. Forbes, S.E. Calvert and R.E. Thomson, Particle fluxes from the euphotic zone

over the continental slope in an eastern boundary current region, 1. Mar. Res., 54, 1097-1122, 1996.

Pena, M.A., K.L. Denman, S.E. Calvert, R.E. Thomsen, and 1.R. Forbes, The seasonal cycle in sinking particle

fluxes off Vancouver Island, Brithish Columbia, Deep Sea Res., Part II 46,2969-2992,1999.

Pena, M.A., K.L. Denman, and 1.R. Forbes, Shelf-Ocean Exchange Water Column and Sediment Trap Data from

the Northeast Pacific, JGOFS Canada Data sets 1989-1998, CD-Rom Version 1.0 Dec. 2000, Marine

Environmental Data Service, Department of Fisheries and Oceans, Canada, 2000.

Peterson, R.G., e.S. lohnson, W. Krauss, and R.E. Davis, Langrangian measurements in the Malvonas Current,

in The South Atlantic: Present and past circulation, edited by G. Wefer et al., 239-247, Springer, 1996.

Pfannkuche, 0., and K. Lochte, The biogeochemistry of the deep Arabian Sea: overview, Deep Sea Res., Part II,

47(14),2615-2628,2000.

Ragueneau, 0., P. Treguer, A. Leynaert, R.F. Anderson, M.A. Brzezinski, DJ. DeMaster, R.e. Dugdale, 1.

Dymond, G. Fischer, R. Francois, e. Heinze, E. Maier-Reimer, V. Martin-lezequel, D.M. Nelson, and

B. Queguiner, A review of the Si cycle in the modern ocean: recent progress and missing gaps in the

application of biogenic opal as a paleoproductivity proxy, Global Planet. Change, 26, 317-365, 2000.

Redfield, A.e., The biological control of chemical factors in the environment. American Scientist 46, 205-221,

1958.

Reimers, e.E., Control of benthic fluxes by particulate supply, in Productivity of the Ocean: Present and past,

edited by W.H. Berger et al., 217-233, 1. Wiley & Sons Lim., 1989.

The benthic carbon mineralization on a global scale 93

Reimers, e.E., R.H. lalmke, and D.e. McCorkle, Carbon fluxes and burial rates over the continental slope and

rise off central California with implications for the global carbon cycle, Global Biogeochel71. Cycles, 6,

199-224, 1992.

Ritzrau, W., G. Graf, A. Scheltz, and W. Queisser, Benthic-pelagic coupling and carbon dynamics in the

northern North Atlantic, in The northern North Atlantic: A changing environment, edited by P. Schafer

et aI., 207-224, Springer, Berlin-Heidelberg, 200l.

Romero, 0., and C. Hensen, Oceanographic control of biogenic opal and diatoms in surface sediments of the

Southwestem Atlantic, Mar. Geol., 186, 263-280, 2002.

Rowe, G.T., G.S. Boland, W.e. Phoel, R.F. Anderson, and P.E. Biscaye, Deep-sea floor respiration as an

indication of lateral input of biogenic detritus from continental margins, Deep Sea Res., Part 11, 41, 657-

668,1994.

Sauter, E.J., M. Schliiter, and E. Suess, Organic carbon flux and remineralization in surface sediments from the

northern North Atlantic derived from pore-water oxygen microprofiles, Deep Sea Res., Part 1, 48, 529-

553,200l.

Schlesinger, W.H., and l.M. Melack, Transport of organic carbon in the worlds rivers, Tellus, 172-187, 1981.

Schlitzer, R., Electronic Atlas ofWOCE Hydrographic and Tracer Data Now available, EOS Trans. AGU, 81, p.

45,2000.

Schlitzer, R., Carbon export fluxes in the Southem Ocean: results from inverse modeling and comparison with

satellite-based estimates, Deep Sea Res., Part 11,49, 1623-1644,2002.

Schliiter, M., E.J. Sauter, A. Schafer, and W. Ritzau, Spatial budget of organic carbon flux to the seafloor of the

northern North Atlantic (600N-800N), Global Biogeochem. Cycles, 14, 329-340, doi: 199GBOOl188,

2000.

Schnetger, B., H.-J. Brumsack, H. Schale, J. Hinrichs, and N. Dittert, Geochemical characteristics of deep-sea

sediments from the Arabian Sea: A high-resolution study, Deep Sea Res., Part 11, 47, 2735-2768, 2000.

Seiter, K., e. Hensen, l. Schrbter, and M. Zabel, Regionalization of the organic carbon content in surface

sediment: Defining a new approach of benthic TOC-based regional provinces, Deep Sea Res.,

submitted.

Sirocko, F., D. Garbe-Schbnberg, and e. Devey, Processes controlling trace element geochemistry of Arabian

Sea sediments during the last 25,000 years, Global Planet. Change, 26,217-303,2000.

Smith, K.L.J., R.S. Kaufmann, and R.J. Baldwin, Coupling of near-bottom pelagic and benthic processes at

abyssal depths in the eastem North Pacific Ocean, LilJ1nol.Oceanogr., 39, 1101-1118, 1994.

Stein, R., and R.W. Macdonald, Organic Carbon Budget: Arctic Ocean vs. global ocean, in The organic carbon

cycle in the Arctic Ocean, edited by R. Stein and R.W. MacDonald, 315-322, Springer, Berlin, 2003.

Suess, E., Particulate organic carbon flux in the oceans - surface productivity and oxygen utilization, Natllre,

288,260-263,1980.

Thiede, l., G.W. Diesen, B.E. Knudsen, and T. Snare, Patterns of Cenozoic sedimentation in the Norwegian­

Greenland Sea, Mar. Geol., 69, 323-252, 1986.

Thomsen, e., Verfolgung pelagischer Prozesse mit Hilfe von biochemischen Komponenten am Beispiel der

Alkenone (C37:2, C37:3), Berichte aus dem Sonderforschungsbereich 3l3, Ph.D. Dissertation, Kiel,

Christian-Albrechts-Universitat, 1993.

van Bodungen, B., A.N. Antia, E. Bauerfeind, O. Haupt, W. Koeve, E. Machado, 1. Peeken, T. Peinert, S.

Reitmeier, e. Thomsen, M. Voss, M. Wunsch, U. Zeller, and B. Zeitschel, Pelagic processes and

vertical flux of particles: an overview of a long term comparative study in the Norwegian Sea and

Greenland Sea, Geal. RlIndsch., 84,11-27, 1995.

94 The benthic carbon mineralization on a global scale

van d. Weijden, CH., G.l. Reichart, and H.J. Visser, Enhanced preservation of organic matter in sediments

deposited within the oxygen minimum zone in the northeastern Arabian Sea, Deep Sea Res., Part 1, 46,

807-830, 1999.

Wagner, T., Organisches Material in pelagischen Sedimenten: Glazialelinterglaziale Variationen im

Europaeischen Nordmeer, Berichte alls de111 Sonderforschungsbereich 313, p.138, Kiel, Christian­

Albrechts-Universitat, 1993.

Wagner, T., M. Zabel, L. Dupont, 1. Holtvoeth, and Cl. Schubert, Terrigenous signals in sediments of the low

latitude Atlantic - Implications for environmental variations during the late Quaternary, Part I: Organic

carbon, in The South Atlantic in the late Quaterna7J'." Reconstrllction of material budget and current

systems, edited by G. Wefer et aI., Springer, in press.

Walsh, 1.1., Importance of continental margins in the marine biogeochemical cycling of carbon and nitrogen,

Nature, 350,53-55, 1991.

Walsh, J.J., D.A. Dieterle, and 1.R. Pribble, Organic debris on the continental margins: a simulation analysis of

source and fate, Deep Sea Res., 38, 805-828, 1991.

Wefer, G., Particle Flux in the Ocean: Effects of episodic production, in Productivity of the oceans: Present and

past, edited by W.H. Berger et aI., 139-153, 10hn Wiley and Sons Ltd., 1989.

Wefer, G., and G. Fischer, Seasonal patterns of vertical particle flux in equatorial and coastal up welling areas of

the Eastern Atlantic, Deep Sea Res., 40, 1613-1645, 1993.

WenzhOfer, F., and R. Glud, Benthic carbon mineralization in the Atlantic: a synthesis based on in situ data from

the last decade, Deep Sea Res., Part 1,49,1255-1279,2002.

Wollast, R., Evaluation and comparison of the gobal carbon cycle in the coastal zone and in the open ocean, in

The Sea, edited by K.H. Brink, and A.R. Robinson, 213-252, University of Brussels, 1998.

W ollenburg, 1., and W. Kuhnt, The response of benthic foraminifers to carbon flux and primary production in

the Arctic Ocean, in Foraminiferal proxies of paleoprodllctivity, Marine Micropaleontology (40), 3,

edited by F. 10rissen, and E. Rohling, v. Spec. Issue., pp. 19-231,2000.

Wong, CS., F.A. Whitney, D.W. Crawford, K. Iseki, R.J. Matear, W.K. 10hnson, 1.S. Page, and D. Timothy,

Seasonal and interannual variability in particle fluxes of carbon, nitrogen and silicon from time series

traps at Ocean Station P, 1982-1993: relationship to changes in subarctic primary productivity, Deep

Sea Res., Part 11,46,2735-2760,1999.

Benthic silica release - estimation of non-lithogenic particleflllxes 95

2.3 The benthic silica release and its implication for the estimation of the non-lithogenic particle fluxes to the sea floor

Katherina Seiter l, J an M. Holstein2

, Christian Hensen3, Matthias Zabel l

I Department of Geosciences, University of Bremen, Klagenfurter StraJ3e, D-28359 Bremen, Germany

2Institut fUr Chemie und Biologie des Meeres (ICBM), Carl von Ossietzky Universitat Oldenburg, Carl von

Ossietzky-Str. 9-11

3GEOMAR - Forschungszentrum fUr Marine Geowissenschaften, SFB 574, Wischhofstr. 1-3,24148 Kiel,

Germany

Abstract

In this study, we investigated the potential coupling of the organic carbon mineralization and the dissolution rate

of biogenic opal in surface sediments of the southern Atlantic Ocean. In this kind, we have compared pore water

profiles of oxygen and silica to develop a simple empirical connection between the oxygen uptake and silica

release in marine sediments (Rsilo), Based on the assumption, that this empirical relation is valid for the entire

Ocean, we applied this coupling rate on a basin wide regular grid of the diffusive oxygen uptake (Seiter et al.,

subm. b) to process a regional distribution pattern of diffusive benthic silica release. In comparison with previous

studies, we could show that the estimation of benthic silica release rates by using the oxygen uptake provides a

reliable budget of 3.7 TmolSi yr- I for the investigated area, with the main fraction built up by the deep-sea basin

areas. Related to the total primary production of biogenic silica in the entire global Ocean, we found a benthic

recycling rate of ~ 13 %. Since we assume that the silica reflux at the sediment-water interface is balanced by the

input of biogenic opal rain to the sea floor, a minimum of biogenic opal accumulation could be detected.

Furthermore we enlarge this approach by applying a published correlation between aerob mineralization and

dissolution rates of calcite above the hydrographicallysocline (Pfeifer et al., 2002) and the organic carbon flux to

sum up the minimum of biogenic particle flux to the sea floor.

96 Benthic silica release - estimation of l1ol1-lithogenic particlef!1fxes

1. Introduction

In general the marine biogenic silica cycle is closely coupled with the cycle of organic carbon

(e.g. Ragueneau et aI., 2000). In particular this gets obvious under conditions of an enhanced

nutrient supply to the surface waters, for example in coastal upwelling areas, equatorial

divergence zones, or along the circumpolar CUlTent of the Southern Ocean. Here, biogenic

opal secreting organisms, especially diatoms, can account for 35-75 % of the marine primary

production (Nelson et aI., 1995). The total annual production of marine siliceous shells has

been estimated to 200-280 TmolSi y(1 (Nelson et aI., 1995; Treguer et aI., 1995). However,

after dying of the organisms, degradation and remineralization processes start immediately.

For biogenic opal it is assumed that 90-94 % of the dissolution already occurs during particle

settling in the water column (Treguer et aI., 1995, Hensen et al. 1998). Correspondingly, the

remaining 6-10 % are attributed to surface sediments. Nevertheless, after present estimates, at

the end about 3 % of the biogenic silica production is preserved below the sea floor (Nelson et

aI., 1995; Treguer et aI., 1995). Thus, the preservation efficiency of biogenic opal would be

about one magnitude higher than of organic carbon, at least in the open ocean. From the just

2 % of the primary produced organic matter that on average reaches the sea floor (e.g.

Christens en, 2000; Antia et aI., 2001) the benthic oxygen uptake indicates that 80-99 % are

remineralized on or rather within the surface sediments (Seiter et aI., subm. b). Therefore,

only 0.01-0.4 % of organic carbon produced in the open ocean is preserved for longer periods

of time. These values may illustrate the great importance of the benthic boundary layer for the

calculation of ocean budgets.

However, due to the benthic release deep and bottom waters are continuously enriched with

H4Si04, as the main dissolved silicon species in the sea (Bruland, 1983), during their way

along the ocean sea floor (Bainbridge, 1976a, 1976b, 1978, cited in Broecker and Peng 1983).

But, usually they remain undersaturated with respect to biogenic silica. So, this

thennodynamical state of imbalance is generally seen as the main controlling factor for

biogenic silica dissolution (e.g. Treguer et aI., 1995). Additional rate limiting influence on this

general process has the sinking rate of aggregates, the temperature, the elemental composition

of biogenic opal, and the composition of the sediments (e.g. van Bennekom et aI., 1989;

Treguer et aI., 1995; van Cappellen and Qui, 1997; Smetaczek, 1999; Ragueneau et aI., 2000).

In this context some recent studies have provided evidence that bacterial activity can also play

an important role for the dissolution rate of diatom frustules (Bidle and Azam, 1999; Bidle et

aI., 2003). After these authors, bacterial ectoprotease action on marine diatom detritus

strongly accelerates silica dissolution rates by removing the organic coatings that normally

protect frustules from direct exposure to the undersaturated seawater. This connection or

rather consequences resulting from it - also for the microbial activity within the benthic

regime - may indicate a kind of coupling between the organic carbon and biogenic silica cycle

that is hardly inspected so far.

The main objective of this work was to investigate the potential influence of the organic

carbon mineralization on the dissolution rate of biogenic opal in surface sediments. For this

purpose we compared pore water profiles of oxygen and silica from the South Atlantic in

Benthie silica release - estimation ofnon-lithogenie particle jlllxes 97

detail. By using benthic oxygen respiration rates as a kind of proxy parameter we examined silica release rates into the bottom water. This indirect approach, which differs considerably from former studies (e.g. Treguer et aI., 1995, Hensen et aI., 1998, Zabel et aI., 1998) also enabled us to calculate the distribution pattern of biogenic opal flux rates within the entire ocean basin. Fmihermore, interpreting the sum of opal and calcite dissolution rates (Pfeifer et aI., 2002) and the mineralization rates of organic carbon (Seiter et aI., subm. b) as equivalent to the minimum flux of the non-lithogenic particles to the sea floor, budgets for the specific rain rates were derived. At the end, these minimal input rates were compared with the total particle flux rates by considering burial rates respectively.

2. Study area and locations: Sedimentary settings

This study focuses on the tropical and subtropical southern Atlantic Ocean between lioN and ~45°S (Fig. la). In general, primary production distribution patterns as the major source of biogenic particles in sediments, are mainly induced by the surface and subsurface water circulation, dominated by a subtropical gyre.

In total, the southern Atlantic is an oligotrophic Ocean, with moderate primary production in the central gyre region, but enhanced phytoplankton growth in the coastal and equatorial upwelling areas (e.g. Antoine et aI., 1996; Wefer and Fischer, 1993; Shannon and Nelson, 1996). Driven by the trade wind systems the surface currents induce coastal and open ocean upwelling of nutrient-rich water masses. In the following, we will use the deviation into benthic sedimentary provinces after Seiter et aI. (subm. a; Fig. 1 b) to characterize the different regimes. While the equatorial upwelling area is located at the benthic province ETROPAT (PPmcan : 182 gC m-2 yr-1 after Antoine et aI., 1996; Tab. 1; Fig. 1 b), coastal upwelling mainly occurs at the south-eastern Atlantic provinces NAMBCO and SW ACO between 300 N and 15°S (PPmcan : 225-256 gC m-2 yr- 1

; Shannon and Nelson, 1996). Further high productivity areas are fostered by riverine nutrient discharge and terrigenous input from the Ougoue River (0.5°S) and Congo River (6 0 S). Here, the bulk terrigenous organic matter has been estimated to constitute 58-76 % of the marine benthic TOC content and significant amounts of low reactive mature organic matter (plant debris) has been found as far as the distal base of the Congo deep sea fan (Wagner et aI., in press). However, enhanced primary production is also established along the continental margins of the SW-Atlantic Ocean (ARGCO; 140 gC m-2 yr- 1

) and is accociated with the confluence zone of the Malvinas and Brazil currents (178 gC m-2 y(l).

With exception of the opal belt around Antarctica (e.g. DeMaster, 1981) and some regions of enhanced terrigenous input, above the CCD the biogenic sedimentation in the southern Atlantic Ocean is dominantly calcareous. Nevertheless, a local area of high accumulation rates of amorphous silica has also been documented off the Congo River mouth, where up to 40 % of the sediment consists of biogenic opal (e.g. van der Gaast and Jansen, 1984).

In table 1 we summarize the mean pp after Antoine et al. (1996) and the mean values of the sedimentary composition (TOC, opal, calcite) after recalculations of published data for specific benthic TOC-based provinces defined in a previous study (Seiter et aI., subm. a).

98 Benthic silica release - estimation of non-litho genic particle /1uxes

Longitude (OE) -70 -60 -50 -40 -30 -20 -10 0 10

10

0 0

-10 -10

~ -20 -20

Q) "0 :::l

~ -30 -30

-40 -40

-50 -50

-60 -40 -30 -20 -10 0 10

o JSi sites included in modeling

• J Si sites recalculated by exponential fit (Sayles et al., 1996)

Figure 1 (a): Investigation area and data sites. The white and gray dots mark the GeoB-sites where geochemical investigations were made (cf. Tab. 2). White dots: Silica flux (Js;) data, used for modeling RSiiO .

Additional Js; data (gray dots) along continental margin of SW-Africa are used in this study as control data to examine the processed distribution pattern ofbenthic diffusive silica fluxes.

-70 -60 -5H -40 -30 ·20 .)0 o 20

()

-to

-2U

SOATL

-40

·50

Figure 1 (b): Benthic TOe-based provinces adopted from Seiter et al. (subl11. a).

Benthic silica release - estimation o!non-lithogenic particleflllxes 99

Table 1: Average sedimentary contents of opal, calcite and TOC in surface sediments derived from recalculated grids after Seiter et al. (subm. a) and mean and max primary productivity (recalculated by Seiter et aI., subm. b; after Antoine et aI., 1996). Geographical position of provinces, see figure 1 b.

Province l Opa1rnax OpalmcHn CaJcitcmax Cal ci temean TOCma, TOC lllenn PPm", PPme""

Iwt%] Iwt'1;'] IW(%] Iwt%] [wt%] Iwt%J [g m,2 yr'l] [g m,2 yr'l]

SOATL 89 13.0 106 34 1.7 0.4 360 101

NAMBCO 41 3.7 85 67 8.2 2.7 562 256

SWACOc 41 6.8 93 39 8.2 1.5 1159 225

ETROPAT 28 13.0 96 48 3.5 0.7 1159 182

GUI 23 6.8 46 II 2.1 1.1 819 189

ARGCO 24 2.6 69 30 0.8 0.3 557 140

RIOPLATA 16 4.8 13 3 3.4 0.8 1210 178

GUBRACO 21 1.8 75 49 1.0 0.4 1280 193

BRAZCO 20 3.6 84 46 1.1 0.5 381 128 1: SOA TL: Southern Atlantic Ocean, NAMBCO: Bcngucia upwdling province ofl'Namibia, SWACO: SW-Afbean continental margin, ARGCO: continental margin oll'Arg'::lltina, RIOPL;\TA: olfRio d.:: la Plata mouth, BRAZCO: continental margin ofl'Brazil, GUBRl\CO: contincntal margin ot]'Guyana and Brazil 2: SWACO: including NAl\.JBCO.

3. Methods

Sampling and geochemical analyses

Surface sediments were retrieved by multi corer (MU C) sampling during four crUIses of RV METEOR (M-2911, M-34/2, MA1I1 and M-46/2). The locations are shown in figure la. Geographical positions, water depth, and benthic diffusive flux rates (JSi, J02) are listed in table 2. Methods of measurements and sampling devices are described in the con-esponding references. For detailed description on analysing procedure, the reader is referred to the appropriate literature (e.g. Gundersen and J0rgensen, 1990; Glud et aI., 1994; Zabel et aI., 1998; WenzhOfer et aI., 2001).

Calculation of dijfilsive oxygen and silica flux rates

To allow flux calculations directly across the sediment-water interface, at first, all oxygen and silica pore water profiles were fitted with the transpOli and reaction model CoTReM (e.g. Adler et aI., 2001; WenzhOfer et aI., 2001). Based on these fits, diffusive exchange rates for both constituents (JSi and J02) were calculated by applying Fick's first law. Molecular diffusion coefficients in sea water were taken from Rutgers van der Loeff and van Bennekom (1989) and were recalculated for 4°C. Diffusion coefficients in sediment were calculated after Boudreau (1997) by considering the tortuosity. The mean porosity was set to 0.85 for the Cape Basin and 0.80 for the Argentine Basin.

100 Benthie silica release - estimation of l1ol1-lithogenie particleflllxes

Table 2: Sampling sites, water depth, diffusive benthic oxygen fluxes (101), silica fluxes (Js;) and RSi/O.

Station Longitude' Latitude' wd .ISl .Is! Jrl1 .Tw. Rs!!o(, cruise reference

exp.\ mnd5 CXp.l,j l11od.5 J"i

[1111 [mor'v('1 [Illor'}c('1 Imor'y('1 [mor'v('1

GeoB 6202 -29.087 -47.1700 l497

C)coB 62 I 4 -34.5250 -5 I .4430 1566

GcoB 6219 -35.1860 -50.S650 3551

GcoB 6223 -35.7400 -49.6810 -1280

GcoB 6226 -37.1580 -47.8990 5055

GeoB 6229 -37.2070 -52.6500 3442

GcoB 6230 -37.8950 -51.6900 4381

GeoB 270" -53.9283 -38.9250 3247

GeoB 270S -53.3633 -39.2433 -1474

GcoB 2706 -55.5350 -42.3683 4700

GeoB 2707 -56.3233 -41.9450 3 I 67

GooB 2709 -570633 -41.5383 1080

GeoB3702 13.4550 -26.7917 1319

GooB 3703 13.2317 -25.5167 1376

GeoB3704 130833 -25.4667 1780

GeoB 3705 12.9967 -24.3033 1305

Geo133706 12.6017 -22.7167 1313

GcoB3707 12.1933 -21.6250 1350

CleoB3713 ll.5800 -15.6283 1330

GeoB 3715 11.0567 -18.9550 1204

GeoD 3717 13.3633 -24.8333 855

GeoB 3719 12.8717 -24.9950 1995

GcoB 3721 12.4000 -25.1517 3014

GeoB 3724 &.9267 -26. I 367 4763

(JeoB 4906 8.3783 -0.6900 1272

neoB 4909 8.6250 -2.0683 1305

GeoB 49 I 3 ll.0717 -5.5033 1296

GooB 49 I 7 13.0733 -11.9033 1300

GeoB 490 I 6.7200 2.6820 2177

(JeoB 4417 5.1383 -46.5750 3511

1: negative long.itude indicates west. 2: negative latitude indicates south. 3: after Sayles et al. (1996). 4: in situ data in bold. 5: modclcd ll11XCS.

6: .1,/102 ratios based 011 modeled flllxes.

(J.OS

0.17

0.17

0.11

0.1 l

0.24

0.34

0.31

(J.20

0.22

0.21

0.68

0.82

0.83

045

0.0.43

0.32

0.49

0.16

0.66

1.38

0.33

0.54

0.24

0.28

0.18

0.53

0.21

0.15

0.08

0.07 -0.22 -0.23 -().30 11446:2

0.17 -0.40 -0.39 -0.44 fvl46/2

0.15 -0.29 -0.30 -0.50 1v!4612

0.18 -0.29 -0.41 -0..+4 M46!2

0.10 -0.22 -0.21 -0.48 1v146/2

0.23 -0.50 -0.47 -0.49 M46!2

0..+3 -0.58 -0.82 -0.52 M46/2

0046 -1.15 -1.06 -0.43 Iv! 29/1

0.47 -0.96 -0.8S -0.53 M 29/l

0.33 -0.85 -0.74 -0.45 M 2911

0.23 -0.78 -0.78 -0.30 M 29/1

n,v -0.96 n,\! Ill'. M 29!1

0.95 -1.36 -1.35 -0.70 Iv! 34/2

0.90 -1.86 - 1.]8 -0.65 fvl 3412

0.50 -0.71 -0.68 -0.74 M 34/2

0.66 -1.5.t -124 -0.53 M 34/2

0.49 -1.19 -1. I 9 -0.41 M 34/2

0.54 -0.88 -0.83 -0.65 Iv! 34/2

0.29 -0.43 -0.-18 -0.60 1143412

1.01 -1.15 -1.13 -0.89 1v134/2

n.v. -0.25 ILl'. I1.V. Iv! 34/2

0.32 -0.45 -0040 -0. SO M 34!2

0.25 -0.38 -0.33 -0.75 Iv134/2

0.16 -0.59 -0.45 -0.35 11434/2

0.46 -0.82 -0.8 I -0.57 M 4lil

0.25 -0.51 -Cl.50 -0.50 M 4111

0.53 -0.96 -0.95 -0.56 1144111

0.30 -0.99 -0.98 -0.3 I M 4lil

0.21 -0.26 -0.33 -0.64 M 41/1

0.05 -0.14 -G.OS -0.63

References: [1]:Holstcin (2002). [2]:Wcl1zhiilcr and Glue! (2002), [3J:Frcnz cl al. (in press). [4J:Hcnscn.;;t aI. (1998), [5]: recalculated this study.

11 v.: not validated

[11

11]

111

[1]

[I]

[I]

[I]

141. [5]

[4], [5]

[4], [5]

[4]. [5]

[4J, [5]

14].[5]

HI, [5]

[4], [5]

141·[5]

14], [5J

141· [5]

[4], [5]

[4], [5J

[4].[5]

H],[5J

[4], [5]

[4]. [5]

[I]

[11

[1)

[I]

[I]

[I[

reference .102

[2]: [5]

[2]; [5]

[2; [5]

[2]: [5]

[2]; [5]

12]: 15]

[2]; [5]

[1]

[1]

[l]

[I]

[1[

[I]

[1]

[1]

[2[: [5J

[21: [51

[2]: [5]

[2]: [5]

[2]; [5]

[2]: [5]

[2J; [5]

[2); [5]

[2]: [51

[21; [51

In[51

[2]: [5]

[2]; [5]

[l[:[l[

121

Benthic silica release - estimation of non-lithogenic particle jluxes 101

Regular basin wide grids

A gobal estimate of the particular organic carbon flux to the sea floor has been given by Seiter et al (subm. b). With the assumption of an exclusively oxic organic carbon mineralization, the underlying global grid at least results from a regionalization of in situ measurements on the benthic diffusive oxygen uptake (DOU). Therefore, we used this map for the calculation of the basin wide silica release rates across the sediment-water interface (cf. sect. 4). The budgeting for specific regions is based on the discretization in TOe-based benthic provinces after Seiter et aI. (subm. a) (Fig. 1 b). Distribution patterns of the calcite and biogenic opal contents were also adopted from Seiter et aI. (subm. a). Regional estimates of the sedimentation rates and dry bulk densities were taken from Mollenhauer et al. (in press).

For all mapping and budgeting procedures the geographical information system Arc View 3.2a® was used. For calculating the basin wide budget of benthic silica release rates, we projected the distribution pattern of the processed grid in a Lambert azimuthal equal-area projection. The general map and budgeting is presented in a 1 °x1 ° grid resolution. The necessary resampling of the cells of the metric system (m2) was done by bilinear interpolation of the cell values (Arc View 3.2 a). Distortion is zero in the center of the investigation area and increases radial but moderate within 90°.

4. Results and Discussion

Basin-wide distribution map of benthic silica release rates

As mentioned above, recent studies have given evidence that biogenic opal dissolution is influenced by marine bacterial assemblages (Bidle and Azam, 1999; Bidle et aI., 2003). It seems obvious, that this connection may also be one controlling factor for the quite similar but mirror-imaged shapes of oxygen and silica pore water concentration profiles in marine sediments. The good correlation between both flux rates gives additional support to this assumption (Fig. 2).

1.4

R:0.88

1.2 SD:O.12 n=28

- p< 0.0001 ,.. 1.0 • .-. slope: -0.54 <"8 • •

I 0.8

"0

" d:i 0.6 "8 8 • • ,;: • .... OA

• • • • • •• • 0.2 • ... •

0.0 0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -lA -1.6

\"l11odeled [mol 11,-' yr"]

Figure 2: Regression plot of mode led J02 vs. modeled hi in [mol m-2 yr'l].

102 Benthic silica release - estimation of non-lithogenic particle jluxes

Consequently, it seems that the benthic organic carbon mineralization, as reflected by the oxygen uptake (e.g. Smith and Hinga, 1983; WenzMfer and Glud, 2002), and the silica release can be transferred to each other by using a simple factor (RSilO), at least over a certain depth interval. Accordingly, benthic oxygen respiration rates (J02) and silica flux rates across the sediment-water interface (JsD would be directly coupled by RSilo after equation 1.

(1)

RSilo is simply the ratio between both flux rates. Following this approach we calculated this ratio for all investigated stations (Tab. 2). The range of RSilo-values is -0.30 to -0.90. Unfortunately, the poor data density does not allow a statistically reliable description of special characteristics in this ratio for specific regions. Therefore, as a rough estimation accepting inaccuracies, we transform the DOU grid after Seiter et al. (subm. b) into the distribution pattern of JSi by applying the mean value of Rsi/O (-0.54: SD: 0.15). Therefore, the pattern is comparable to the DOU distribution map (Fig. 3). Highest JSi are estimated in the upwelling area along the continental margin off Namibia (0.83 mol m-2 yr- 1

) and offshore the Congo River mouth (0.53 mol m-2 yr-1

). For the verification of our JSi distribution pattern we recalculated silica flux rates after Hensen et al. (1998) and Zabel et al. (1998) by using the exponential approach after Sayles et al. (1996). In general, a good agreement with results from both former studies can be established. Nevertheless, due to different methods of calculating flux rates, different approaches in regionalization, and the gross generalization by using a constant RsilO, differences against these previous studies arise regionally.

10-

0-

-10-

Z -20-~

-40-

-50-

-60-

-70 I

-70

-50 I

0.01 - 0.1

-60 -50

Longitude eE)

-40 -30 -20 I

0.15 - 0.2 0.25 - 0.3 0.35- 0.4

I

-40 -30 -20

-10 o 10 20 I

-10

-0

--10

--20

--30

--40

--50

0.5 - 0.6 0.7 - 0.8 >0.8 ,

-10 I

o ,

10 20 --60

Figure 3: Regional distribution pattern of diffusive benthic silicate release rates (1 ° x 1°). Locations of modeled fluxes are marked as black dots, only data used for modeling Rs1/0 are displayed.

Benthic silica release - estimation of non-litho genic partic1eflllxes 103

The two former possible reasons described above are hard to discuss, because flux

calculations, based on conventional pore water profiles strongly depend on the method in use.

figure 4 shows a comparison between results by fitting an exponential function on the data (after Sayles et aI., 1996) (Tab. 2) and modeled ]Si (this study). Like to be expected, calculations with a geochemical model tend to higher flux rates in general (mode led data vs.

exponential fit: slope= l.IS).

1.2 R:0.89

SD:O.12 l.O 11:28 • ./

p< 0.0001 ./

./ • ./ ,....., slope: US /

/ ,..., 0.8 ".. c, a ./

/

"0 • ./

~ 0.6 /

./

..:.' /

v ./ /. V • ./ . V • •• ./

V . ./ ./ 0 OA a ./

.;: • ./ / -, • • /

• / • 0.2 . / . / L •

0.0 0.0 0.2 0.4 0.6 O.S 1.0 1.2

JSl

exponential fit [molm-: y/]

Figure 4: Comparison of modeled data and exponential fitted silica flux rates after Sayles et al. (1996). The mean ratio (JSi-moclelecl/JSi-exponential) is given by the slope of l.IS (R=0.89, dashed line). Solid line=­bisecting line.

Because the used RSi/O is based on modeled fluxes, our approach leads to overestimations

compared with common calculations of ]Si by using exponential fits or linear gradients. We do not want to go into this discussion further, which may be relative and would question the

reliability of flux calculations at least. However, as a result, silica fluxes queried from the grid

(JSi-grid) correlate with modeled JSi with R=0.70 and a slope of 0.72 in general, but with R=0.7S and a slope of 0.94, if data offshore Namibia and Argentina are excluded which will

be discussed below (Fig. Sa). But ]Si-grid correlate with the recalculated exponential fitted data (cf. Fig. Sa) with R=0.81 and slope of 1.22, which indicates a potential relative overestimation

of grid values, based on modeled data vs. exponential data already shown in figure 2. Figure Sb shows the estimation error of grid-derived estimations vs. modeled JSi, expressed by the

relative ratio of] Si-gridl] Si-modcled.

104

to

'f- O.X

"0 ,8 :::1 0,6 Sf) r-<

~ <.;:1

] 04 C':l a -.= (/)

o

~-J5 0,2

Benthic silica release - estimation of non-litho genic partic1ef1l1xes

C eXplJllent. recalculated t1uxes (Smles et al.. 19<)6) R (1.81 SD Oil

N 26 l' < ()(lOOI

slope 1.22

, '"

".I

,C ,/

"B

A mocickd data cxcl. Cn:oB 37:\.\: (NamibIa) and 27xx u\rgentinu)

RO.75 SD(U9 N 13 P <() 002 slope: (J.l)4

B all model ed (bra R 0 70 SI) on N2X p: < (W02

slope 072

QO 0,2 Q4 0,6 0,8 1,0 • J, modeled [mol n{ yr-1

] 01. .' -1. a J s, exponentIal fitted [mol m - yr J

Figure 5 (a): Comparison of estimated benthic silica fluxes derived from the silica flux grid vs, modeled silica fluxes, Black dots, line-A: The mean ratio of JSi-gri,/JSi-modeled is 0,94 (R=0,75), if data sites off Argentina and Namibia are excluded, Line-B: If all modeled data are included the ratio is 0,72, Circles, line-C: If ]Si queried from the grid are compared with recalculated exponential fitted J Si (data sites located along the cont. margin of SW-Africa) the ratio is 1.22 (R=0,81), The solid line marks the bisecting line, GeoB37xx or GeoB27xx mean all GeoB sites beginning with number 37 or 27,

Longitude (OE)

-60 -40 o 20

20-

-20-~

z e...- x Ql "0 :J :t: ro

ratio of Jsl-gnd \'::; . .lSi-moddcd

....J 11-0 . .5

-40 x 0.5-1

X I-I 5

X 1.5-2

X .'

-60-

Figure 5 (b): Black crosses: estimation error ratio of J Si queried from the grid vs, modeled ]Si (JSi-gridnSi-mod),

Benthic silica release - estimation of l1ol1-litlzogenic particle jlzc'(es 105

The effect of using a constant Rsilo is obvious as shown in figure 5b. In regions where the

calculated ratios are mostly higher than the mean values the silica fluxes were underestimated

and lower Rsilo will result in overestimations. The first may be the case of the Bengue1a upwelling region and the area north of the Walvis Ridge. The latter may be true for parts of the Angola Basin. Whereas the intense primary production in the area off Namibia enhances

preservation of organic matter, which results in a stagnant J02 with increasing TOC concentrations (WenzhOfer and Glud, 2002; Seiter et aI., subm. b), the north-eastern Angola Basin is mainly influenced by the riverine discharge from the Congo River. So, the lower

Rsilo ratios in this region may be caused by the high input of aluminum, which reduces the

solubility of biogenic silica (e.g. van del' Gaast and Jansen, 1984; van Bennekom et aI., 1989; van Cappellen and Qui, 1997; Dixit et aI., 2001; Dixit and van Cappellen., 2002; Gallinari et aI.,2002).

Minor differences can be observed for the area offshore Argentina, especially off the Rio de la Plata, where maximum values of 0.3 mol m-2 yr- 1 correspond fairly well with flux rates up to 0.4 mol m-2 yr-1 after Hensen et aI. (1998). Silica release rates of the southwestern sites

(GeoB 27xx) are underestimated by a factor of 0.5 (Fig. 5b). These discrepancies in comparison with Hensen et aI. (1998) may be attributed to the differences in the input data.

The data density of DOU values as the basis for the used DOU grid (Seiter et aI., subm. b), which is based on in situ data only, is much lower than the number of available JSi calculations. We have to assume that realistic Rsilo might be higher in this area than the mean

modeled RSi/O for the entire Ocean.

Although, just as previous investigations, with our new approach of estimating JSi we can not exclude effects by seasonal variations, but their influence may be negligible for a basin-wide

budget. However, despite the described uncertainties, it is to emphasize that the incorporation of apparent control parameters like DOU and TOC (as control parameter in processing the

DOU grid after Seiter et aI., subm. b) respectively, gives the fundamental advantage of a huge

number of additional information, especially with regard to spatial resolution.

Quantification of benthic silica release rates

For each benthic province a separate budget was calculated (Tab. 3). In total, we estimate a

benthic silica release rate of about 3.4 TmolSi yr- 1 for the whole investigation area of

34.4-106 km2. In comparison with the estimate after Hensen et aI. (1998) and reduced to the

area below the 1000 m isobath, this budget is higher by a factor of l.5.

106 Benthic silica release - estimation of non-litho genic particle/111xes

Table 3: Budgeting of diffusive benthic silica fluxes in specific provinces as total benthic silica release and min, max and mean values, gueried from the Erocessed grid (JSi-grid).

surface urea2 J si~£1id J::;i-grid Jsi-grid

Provinces (min) (max) (mean) [m2

] [mol m-2 )(1] [mol m-2 YI'-I] [mol m-2 y(l]

SWACO I 1.9E+12 0.07 0.63 0.26

GUI 2.()E+II 0.04 0.31 0.22

GUBRACO 8.7E+II 0.06 0.14 0.10

SOATl. 2.7E+ 13 0.01 0.34 0.09

NOATL 4.0E+II 0.02 0.25 0.09

ARCiCO 7.3E+I0 0.05 0.19 0.09

BRAZCO 1.0E+ 12 0.03 0.17 0.10

RIOPLATA 4.1E+l J 0.06 0.29 0.15

NAMBCO 3.7E+l1 0.07 0.63 0.32 ETROPAT*)3 2.6[+12 0.02 0.37 0.14

E'I'ROPAT4 3.5E+12 0.01 0.6J 0.15

2: excluding wc! < lOOO m. 3: Equatorial Atlantic (ETROPAT*) calculated including continental margin off Guinea (GUll and excluding CONGO fresh water plume area. 4: Equatorial ;\{Jantic (ETROPA T) ca1culat.::d including continental margin olfN-S WACO.

SD Total budget

[mol 111-2

y( 11 [lmol y(l]

0.11 0.49

0.07 0.05

0.01 0.08

0.03 2.14

(L02 0.39

0.03 0.05

0.03 0.11

0.04 0.06

0.15 0.12

0.06 0.36

0.09 0.52

3.7

5: Total area: SWACO+GUBRACO+SOATL+NOATL+ARGCO+BRAZCO+R10PLATA+ETROPAT*-t- CONGO li"esh \vater plume area (Congo area=6.9·1 0 11 m') Sf): standard deviation of mean J5.-",.d.

It must be pointed out, that the continental margin areas are of minor importance for the

general budget of the southern Atlantic Ocean as already shown by Hensen et al. (1998).

Despite lowest flux rates in the deep-sea basin the silica reflux is in the order of 75% of the

total budget of the investigation area. This shows that due to the expanded floor space of 80 %

of the total area, these provinces (SOATL and NOATL) are of major impOltance for the

global budget (Fig. 6, Tab. 3). On the other hand, only ~ 13 % and ~ 10 % respectively of the

total silica release can be attributed to the high productive areas along the SW-African

continental margin (SWACO, including NAMBCO) and of the eastern equatorial upwelling

area (ETROPAT* 1: excluding the fresh water plume of the Congo River, cf. Tab. 3),

respectively. Conspicuously high is the amount of JSi in the area influenced by the Congo

River input. Here, 4.6 % of the total reflux occurs on only 2 % of the total area

(0.17 TmolSi yr- I). This value is nearly balanced by the annual Congo River discharge of

dissolved silica of 0.23 TmolSi yr- I after van Bennekom and Berger (1984). However, related

to the amount of biogenic opal production of ~240 TmolSi yr- I on the global scale (Nelson et

aI., 1995; Treguer et aI., 1995), our results indicate a total benthic reflux of about 13.5 %,

which would be slightly higher than calculated by previous studies (e.g. Hensen et aI., 1998).

In a next step we follow the question how estimates of biogeochemical transfer processes can

be used for the calculation of particle rain rates to the sea floor.

1: ETROPAT*=ETROPAT including GUI and excl. the fresh water plume area off the Congo River.

Benthie silica release - estimation of non-litho genic particlejlllxes 107

5.0 Decreasing pnJ\"lllCe area [m')

- - - - -c - - - - -=: =: =: =: =: =: '""T 4' C' 'r, 'of: s· f~ f- 'n

4.0 rr. ~ r-. C'J cr. er. 'n '""T ('·1

':-. - .-. -;' l- a 3.0 .-, :::: 8 r=:

(3 '" 2:: 8 '" 2.0 :...

0.8 If. Ql 0.. -, :--: § :::

Ql Cl 8 '" 0.5

• '0 E-

0.3

0.0

Figure 6: Gray columns: Budgeting of total silica fluxes in specific provinces (Seiter et a!. subm. a) in [Tmol yr'!]. Black dots: mean JSi in each province in [mol m'l yr'!]. ETROPAT*=ETROPAT exluding the freshwater plume area of the Congo River and inc1. GUI.

Non-lithogenic particle flux to the sediment - the lower limit (Jpmil1)

Conventionally, the amount of paliicles that actually reaches the sea floor can be deduced

from sediment trap results or by calculation of accumulation rates (AR). However, the data

base for trap investigations is sparse and estimations derived from AR can only record the

fraction which is preserved within the sediments. Furthermore, AR strongly depends on

unceliainties in age models. In contrast, benthic biogeochemical transfer rates, closely

coupled with the availability of labile organic carbon components as the main controlling

factor, can be interpreted as a proxy for the particle input or rather its mineralized fraction

(Jahnke, 1996; Wenzhofer and Glud, 2002; Seiter et aI., subm. b). Due to complex and mostly

microbial catalysed interrelations this approach is not inevitably restricted to the rain rate of

organic debris. For example, the degradation of organic carbon is also an important process

driving dissolution of calcite in surface sediments above the lysocline (e.g. Pfeifer et al;

2002).

Accordingly and based on our assumption of a connection between the benthic silica release

rates and the oxygen respiration rates, we can estimate the minimum particle flux of biogenic

opal and calcite to the sea floor (Jopal+calcitc) from the diffusive oxygen uptake (Eq. 2).

J opa!+calcite = M opa! x IJ Si I + Mcaco3 x IJ Ca I (2)

Jopal+calcite is calculated in [g m,2 yr'l]. Diffusive fluxes have the unit [mol m'2 yr- I] and the

molar weights are given in [g mor l]. For calcite dissolution rates above the hydrographical

108 Benthic silica release ~ estimation o(non-lithogenic partic1ef1l1xes

calcite-lysocline (DRCaC03) a linear con-elation between DRcaco) and mineralization rates of

organic carbon expressed by the benthic oxygen consumption (102) has been given by Pfeifer

et al. (2002) (Eq. 3).

DR = 1 1 x 106 x IJ 1 ~ 22 = IJ 1 CaCO; . 138 0, 100 Ca

(3)

In this equation both, DRcaco) and 10" have the unit [mol m-2 yr- l]. A (02:C) ratio of 138:106

is assumed (Froelich et aI., 1979, Redfield, 1958).

By substitution of IJ Si 1 and 11 Ca 1 in equation (2) with the terms on the right hand side of

equation (1) and (3), lopal+calcitc can be calculated as

lopal+caJcite = M opal x 0.54x 110,1 + M caco, x( 1.lx ~~~ x 11 021- ~~~) (4)

After simplification equation (4) can be written as

J opal+calcite = A x IJ 0, 1- 9.3 (5)

Jopal+calcitehas the unit [g m-2 yr- I]

Where A is a constant which amounts to 116.9. To calculate the non-lithogenic particle flux to

the sea floor (1Pmin) we have to add the organic carbon flux. For this we used the estimation of

Seiter et al. (subm. b). The result is presented in figure 7a. Accordingly, as expected the

highest biogenic particle rain rates can be determined at the continental margins, especially

offshore Namibia. However, it has to underline that l Pmin is only the estimated portion of the

particle rain that is influenced by organic carbon mineralization in surface sediments. The

difference against the total flux (lptoD is tally with the pOliion buried and preserved below the

oxic-suboxic interface (see below). There is no general correlation between the l Pmin and l ptot.

Patiicularly when mineralization rates fall below 0.1 mol m-2 yr- l no calcite dissolution has

been observed (Pfeifer et aI., 2002). Consequently, the usage of equation (3) may lead to

extreme underestimates of l ptot at the Mid-Atlantic ridge system where the input of organic

matter is very low (hachure in Fig. 7a).

Total non-lithogenic particle flux to the sediment - the upper limit (Jp /o/)

In high productive areas and at the Middle Atlantic Ridges burial efficiencies l11crease.

Accordingly, the burial rate of biogenic material becomes more important (~). This additional

portion of l ptot can roughly be calculated from the product of dry bulk density (DBD),

sedimentation rate (SR), and sedimentary content of the non-lithogenic fractions (PC).

~ = 0.1 xPC xSR x DBD (6)

PC = (CaC0 3 + opalbio + TOC) (7)

With PC in [g g-l], DBD in [g cm-3], and SR in [cm kyr- l], ~ has the unit [g m-2 yr- l]. So, JPtot

results from the sum of ~ and l Pmin. For SR and DBD the 1 °x1 ° grids from Mollenhauer et al.

Benthie siliea release - estimation of non-lithogenie particle fluxes 109

(in press) were used. Estimations from certain locations are listed in table 4. As a result, the distribution pattern of JPtot is depicted in figure 7b.

---:z ° ~ ~

'"0 ::l .... . ~ .....l

Longitude (0E) -70 -60 -50 -40 -30 -20 -10

I I I o

.... ... •

20

-60- <55-10 20-30 40-50 60-70 80-90100-120130-140 >150

Longitude (OE) -70 -60 -50 -40 -~o -40 -10 0 10 20

I I I

lO-

O-

-10-

-20-

-30-

-40- -'- + .. : ~

....

-50- + • """"" + .... . -.,. -~-..---..

Figure 7: (a) Distribution pattern of the minimum of non-lithogenic particle fluxes to the sea floor (JPmin) in [g m-2 yr-1

] as estimated from diffusive oxygen uptake. Fluxes were validated above the hydrographical lysocline after Pfeifer et al. (2002). Hachured: Area of less calcite dissolution rates by organic carbon mineralization. (b) Distribution pattern of the total non-litho genic particle fluxes to the sea floor (JPlot)

as estimated by the sum of Jpmin and calculated burial rates. Sedimentation rates (SR) and dry bulk density (DBD) grids were adapted from Mollenhauer et al. (in press). Calcite and opal grids were adopted from Seiter et al. (subm. a). Legend in figure 7b see figure 7a.

>-'

Table 4: Sediment composition at specific sites and calculated non-lithogenic particle fluxes to the sediment surface (JP min=minimum particle flux; JPtot= total 0

particle flux).

TOC Opal Opal residual" Calcit SR' DBD' (\'

Event Jpntill-J02...grid ~ " Jpmin-J02-mod"

, Jplot-102-grid JPtOI_J02_mod3" Reference Reference

[\\1%] [wt%] [\\-1%L_ [Wl%] [cm kyr"'] [gE£m"] [g m·2 yr"] [g m" yr"'] Ig nl' yr"] Ig m" yr"] [g m·2 yr"'] TOC Opal

GooB4906 1.51 4.40 2.51 4.90 3.6 0.35 1.14 45.0 94.2 46.1 95.4 [3] [l]

GooB4909 1.32 2.70 1.35 12.3 4.3 0.41 2.66 36.1 55.1 38.8 57.8 [3] [l]

GeoB4913 1.65 6.00 4.31 0.10 12.1 0.36 2.64 119 112 122 115 [8] [I]

GooB4917 2.16 1.10 1.54 0.30 8.3 0.41 1.36 95.2 116 96.6 117 [8] [l]

GooB 4901 0.80 4.70 4.46 4.50 4.1 0.38 1.52 54.5 23.5 56 25 [3] [I]

GeoB 3702 3.10 l.90 1.01 52.4 10.0 0.23 13.00 89.4 162 102 175 [3] [5]

GeoB 3703 6.10 2.00 1.80 45.8 14.5 0.22 17.13 107 225 124 243 [3] [5]

GeoB 3704 4.00 2.80 2.08 51.0 9.9 0.19 10.78 107 80 118 91.l [3] [5]

GooB 3705 4.53 1.20 0.85 58.9 7.9 0.20 10.21 107 185 117 195 [3] [5]

GeoB3706 3.74 1.50 0.48 55.6 8.4 0.12 6.01 60.1 141 66.1 147 [3] [I]

GooB3707 4.92 2.50 1.90 45.3 7.6 0.19 7.57 102 102 109 109 [3] [I]

GeoB 3719 2.74 n.v n.v 62.0 9.9 0.25 16.1 91 48.8 107 65 [3] [5]

GooB 3721 1.03 n.v n.v 70.3 5.4 0.33 12.8 67 38.7 79.3 51.5 [3]

GooB 3713 1.20 n.v n.v 5.10 9.0 0.40 2.27 74.0 51 76.3 53.6 [3]

GeoB 3715 3.10 n.v n.v 25.6 9.0 0.24 6.22 114 133 120 140 [3]

GeoB 3724 0.60 6.50 4.29 27.0 1.7 0.24 1.26 2 1.6 47.5 22.9 48.9 [3] [5]

GooB 6202 0.96 1.00 0.57 1.00 3.5 0.53' 0.47 20.4 18.5 20.9 18.9 [3] [I]

GooB6214 2.48 2.10 2.34 n.v 35.0 0.53' 6.39 40.6 41.2 47.0 47.6 13] [I]

GooB 6219 1.19 5.20 6.26 n.v 2.7 0.53' l.07 27.9 27.3 29.0 28.4 [3] [4]

GeoB 6223 0.50 2.60 n.v 0.00 2.8 0.53 2 0.07 21.5 27.3 21.6 27.4 [3] [4]

GooB6226 1.07 6.70 n.v n.v 3.0 0.53' 0.17 22.0 18.5 22.0 18.7 [3] [4]

GooB 6229 1.49 4.70 4.00 0.70 2.7 0.53' 3.28 42.8 53.8 46.1 57.1 [3] [4]

GeoB 6230 1.02 8.30 n.v n.v 3.0 0.53' 0.16 28.8 63.9 29.0 64. I [3] [4]

GeoB 2704 0.77 2.70 n.v 11.7 3.0 0.53' 1.98 26.5 136 28.4 140 [3] [4]

GooB 2705 1.33 11.90 A.V 1.80 2.9 0.53' 0.50 29.4 112 29.9 112 [3] [4]

GooB 2706 0.50 7.20 n.v 0.30 3.6 0.53' 0.17 17.4 98.0 17.6 98.2 [3] [4]

GeoB 2707 0.40 2.50 n.v 13.3 3.6 0.53' 2.90 14.4 89.2 17.3 92.1 [3] [4]

GcoB 4417 0.64 n.v. n.v 0.60 2.6 0.53' 0.39 n.v n.v 20.9 19.0 [2]

I: SR: Sedimentation rate queried from SR-grid after Mollenhauer et al. (in press). 2: DBD (Dry bulk density) queried from DBD-grid after Mollenhauer et al. (in press). 3: (\: estimated burial rate after equation (6). 4: Jpmm.JO,.gridandJp,o,.JO,.grid: Based on o":ygen flux grid values after Seiter et al. (subm. b). 5: Jpmi".J02.modandJp,o,.J02.,nod: Based on modclcd oxygcn f1uxcs this study. 6: Background value in 4-5 cm sediment depth. References: Holstcin (2002)= l. WenzMfer (1999)=[2]. Mollenhauer et al. (in press)=[3]. Romero and Hensen (2002)=[4]. Schulz (2000)=[5]. Hensen el al. (1998)=[6]: Pfeifer cl al. (2002)=[7]: Wagner et al. (in press)=[8]. Riedinger el al. (subm.)=[9],

Reference

Calcite

rn rn ~

~ rn ~ ~ ~

~ ~ ~

~

~ ~

~

~ ~

[9]

[9]

[6]

[6]

[6]

[6]

[2]

b;J (\:) :;:s ~ (;.

Co ;:::;

~r ~ ~ Sl (\:)

I (\:) Co

§. s:::, ~. :;:s

~ :;:s § ..!..... §::

~ :;:s (;.

'2 ;::;, (;' ~ ';:;:, ~ ~

Benthic silica release - estimation of non-litho genic partic1efluxes 111

In general, differences between Jpmin and J Ptot are conspicuously low. Major deviations can be observed offshore Namibia, where high accumulation rates cause a rapid burial of patiicles. The calculated burial efficiency is ~ 10 % ±2 of Jpto(, of which the main part consists of calcite. Thus, 16 % on average and up to 35 % of calcite that settles on the sea floor is buried, but less than 1 % of opal and TOe. Increasing opal burial efficiency, but less calcite preservation can be observed offshore the Congo River mouth. Due to the position of the hydrographic lysocline the southemmost GeoB sites off Argentina (GeoB 27xx, GeoB 6223, GeoB 6226, and GeoB 6230) are subject to intense calcite dissolution and do not affect the fraction built up by burial considerably.

In comparison with conventional accumulation rates (based on SR and DBD calculations), the calculated total non-lithogenic particle fluxes exceeded these rates by far, since the main portion of biogenic particles settled on the sea floor were immediately remineralized at the sediment-water interface. Direct methods as trap data collected particle fluxes are affected by large unceliainties, due to e.g. intense bottom currents and thus advective transport of suspended particles in the water column or at the sea floor. Calculations of non-lithogenic particle fluxes from diffusive oxygen fluxes (Jpmin) across the sediment-water interface reflect these additional effects and thus are a usefull and practical method for estimating biogenic fluxes to the sea floor.

5. Conclusions

By this completely new approach of developing a simple empirical cOlmection between the oxygen uptake and silica release rates in marine surface sediments, we generated a distribution pattem of the benthic silica flux rates and subsequently of the biogenic opal rain rate to the sea floor for the S-Atlantic Ocean. Despite the rough simplification, a generally good coincidence of the silica flux map with results from previous studies (Zabel et aI., 1998; Hensen et aI., 1998) could be observed. Thus, we could show that the estimation of benthic silica release rates by using the oxygen consumption rates provides a reliable budget of 3.7 TmolSi yr-1 for the investigation area. Related to the total primary production of biogenic silica in the entire global ocean, we found a benthic reflux of ~ 13.5 %. Like to be expected, the pOliion of the deep basin areas (> 4000 m water depth) built up the main fraction of the total budget.

The minimum flux of biogenic opal to the sea floor is assumed as balanced by the reflux, expressed by the diffusive benthic silica release rates. Under consideration of an additional fraction, which is actually buried in the sediments, we could calculate the total biogenic opal rain rate to the sea floor. This approach was enlarged to the calculation of the total non­lithogenic particle flux as the sum of total opal, calcite and organic carbon fluxes to the sea floor for ocean areas above the hydro graphical calcite lysocline. However, the lysocline limits the validity of the regression between organic carbon mineralization rates and calcite dissolution exclusively. In comparison with common accumulation rates, the calculated fluxes exceeded these rates by a multiple, since, the main pOliion of biogenic particles actually settled on the sea floor, were immediately remineralized at the sediment-water interface. Accordingly, the calculated non-lithogenic particle flux is an improved method and good

112 Benthic silica release - estimation oll1on-lithogel1ic particle flllxes

estimation for biogenic pmiic1es that actually is exported to the sea floor for oceanographic areas above the hydrographicallysoc1ine.

Acknowledgements

We would like to thank H. Hecht, K. Pfeifer, N. Riedinger for their valuable and critical

comments during the development of this manuscript. Many thanks also to all the unnamed colleagues who contributed with helpful comments. This research was funded by the Deutsche Forschungsgemeinschaft (DFG, ZA 19911- 1).

Benthic silica release - estimation o!non-lithogenic particle fluxes 113

References

Adler M, Hensen C, Wenzhoefer F, Pfeifer K and Schulz HD (2001) Modelling of calcite dissolution by oxic

respiration in supralysoclinal deep-sea sediments. Marine Geology 177 : 167-189

Antia AN, Koeve W, Fischer G, Blanz T, Schulz-Bull D, Scholten J, Neuer S, Kremling K, Kuss J, Peinert R,

Hebbeln D, Bathmann U, Conte M, Fehner U and Zeitschel B (2001) Basin-wide particulate carbon flux

in the Atlantic Ocean: Regional export patterns and potential for atmospheric CO2 sequestration.

Global Biochemical Cycles 15 : 845-862

Antoine D, Andre J-M and Morel A (1996) Oceanic primary production - 2. Estimation at global scale from

satellite (coastal zone color scanner) chlorophyll. Global Biogechemical Cycles 10: 57-69

Archer DE (1996) An atlas of the distribution of calcium carbonate in sediments of the deep sea. Global

Biogechemical Cycles 10 : 159-174

Bidle KD and Azam F (1999) Accelerated dissolution of diatom silica by marine bacterial assemblages. Nature

397: 508-512

Bidle KD, Brzezinski MA, Long RA, Jones JL and Azam F (2003) Diminished efficiency in the oceanic slica

pump caused by bacteria-mediated silica dissolution. Limnology and Oceanography 48: 1855-1868

Boudreau BP (1997) Diagenetic models and their impletation - Modelling transport and reactions in aquatic

sediments. Spriger, Berlin - Heidelberg

Broecker WS and Peng T-H (1983) Tracers in the sea. Lamont-Dohery Geological Observatory, Columbia

University, Palisades, New York

Bruland KW (1983) Trace Elements in Sea-water. Chemical Oceanography 8: 157-220

Christensen JP (2000) A relationship between deep-sea benthic oxygen demand and oceanic primary

productivity. Oceanologica Acta 23 : 65-82

DeMaster DJ (1981) The supply and accumulation of silica 111 the marine environment. Geochimica et

Cosmochimica Acta 45 : 1715-1732

Dixit S and van Cappellen P (2002) Surface chemistry and reactivity of biogenic silica. Geochimica et

Cosmochimica Acta 66 : 2559-2568

Dixit S, van Cappellen P and van Bennekom AJ (2001) Processes controlling solubilit of biogenic silica and pore

water build-up of silicic acid in marine sediments. Marine Chemistry 73 : 333-352

Frenz M, Heinrich R, Hoppner R, Stuut JB and Wagner T (in press) Surface sediment bulk geochemistry and

grain size composition related to oceanic circulation along the South American continental margin in

the Southwest Atlantic. In: Wefer G, Mulitza Sand Rathmeyer V (Eds.): The South Atlantic in the late

Quaternary - Reconstruction of material budget in current systems. Springer

Gallinari M, Ragueneau 0, Corrin L, DeMaster DJ and Treguer P (2002) The importance of water column

processes on the dissolution properiies of biogenic silica in deep-sea sediments - 1. Solubility.

Geochimica et Cosmochimica Acta 66 : 2701-2717

Glud RN, Gundersen JK, Revsbech NP and J orgensen BB (1994) Effects on the benthic diffusive boundary layer

imposed by microelectrodes. Lirnnology and Oceanography 39 : 462-467

Gundersen JK and J orgensen BB (1990) Microstructure of diffusive boundry layer and the oxygen uptake of the

sea floor. Nature 345 : 604-607

Hensen C, Landenberger H, Zabel M and Schulz HD (1998) Quantification of diffusive benthic fluxes of nitrate,

phosphate and silicate in the Southern Atlantic Ocean. Global Biogeochemical Cycles 12 : 193-210

Holstein JM (2002) Effekt benthischer Mineralisationsprozesse auf die Losung von biogenem Opal. diploma

theses, unpublished, University of Bremen, Depariment of Geosciences, Bremen, Gennany

114 Benthic silica release - estimation o{nol1-lithogenic particle j1l1xes

Jahnke RA (1996) The global ocean flux of particulate organic carbon: Areal distribution and magnitude. Global

Biogeochemical Cycles 10: 71 - 88

Mollenhauer G, Schneider RR, MUller PJ, SpieJ3 V and Wefer G (2002) Glacial/Interglacial variability in the

Benguela Upwelling System: Spatial distribution and budgets of organic carbon accumulation. Global

Biogeochemical Cycles 16 : 1134

Mollenhauer G, Schneider RR, Jenneljahn T, c., MUller PJ and Wefer G (in press) Organic carbon accumulation

in the South Atlantic Ocean: Its modern, mid-Holocene and Last Glacial basin-wide distribution. Global

and Planetary Change

Nelson DM, Treguer P, Brzezinski MA, Leynaert A and Queguiner B (1995) Production and dissolution of

biogenic silica in the ocean: Revised global estimates, comparison with regional data and relationship to

biogenic sedimentation. Global Biogeochemical Cycle 9 : 359 - 372

Pfeifer K, Hensen C, Adler M, Wenzhoefer F, Weber Band Schulz HD (2002) Modeling of subsurface calcite

dissolution including the respiration and reoxidation processes of marine sediments in the region of

equatorial upwelling off Gabon. Geochimica et Cosmochimica Acta 66(24) : 4247-4259

Ragueneau 0, Treguer P, Leynaert A, Anderson RF, Brzezinski MA, DeMaster DJ, Dugdale RC, Dymond J,

Fischer G, Francois R, Heinze C, Maier-Reimer E, Martin-Jezequel V, Nelson DM and Queguiner B

(2000) A review of the Si cycle in the modern ocean: recent progress and missing gaps in the

application of biogenic opal as a paleoproductivity proxy. Global and planetary Change 26 : 317-365

Redfield, A.C. (1958) The biological control of chemical factors in the environment. American Scientist 46 :

205-221.

Riedinger N, Pfeifer K, Kasten S, Ganning JFL, Vogt C and Hensen C (subm) Diagenetic processes within and

below the zone of anaerobic oxidation of methan in sediments of the western Argentine Basin -

Focusing on iron diagenesis. Geochimica et Cocmochimica Acta

Romero 0 and Hensen C (2002) Oceanographic control of biogenic opal and diatoms in surface sediments of the

Southwestern Atlantic. Marine Geology 186 : 263-280

Rutgers van der Loeff MMR and van Bennekom AJ (1989) Weddell Sea contributes little to silicate enrichment

in Antarctic Bottom Water. Deep Sea Research 36: 1341-1357

Sayles FL, Deuser WG, Goudreau JE, Dickinson WH, Jickells TD and King P (1996) The benthic cycle of

biogenic opal at the Bermuda Atlantic Time Series site. Deep Sea Research 43 : 383-409

Schulz S (2000) FrUhdiagenese von Opal in Oberflachensedimenten des S-Atlantiks-Versuch einer Bilanzierung.

UnverOffentlichte Diplomarbeit, Universitat Bremen : 74

Seiter K, Hensen C, Schroter J and Zabel M (subm. a) Regionalization of the organic carbon content in surface

sediment- defining a new approach ofbenthic TOC-based regional provinces. Deep Sea Research.

Seiter K, Hensen C and Zabel M (subm. b) Benthic carbon mineralization on a globale scale. Global

Biogeochemical Cycles.

Shannon LV and Nelson G (1996) The Benguela: Large scale features and processes and system variability. In:

Wefer G, Berger WH and Siedler G (Eds.): The South Atlantic: Present and past circulation. Springer­

Verlag Berlin Heidelberg : 163-210

Smetacek V (1999) Bacteria and silica cycling. Nature 397: 475-476

Smith Jr. KL and Hinga KR (1983) Sediment community respiration in the deep sea. In: Rowe, G.T. (Eds.):

Deep-Sea Biology. Wiley, New York, Chichester, Brisbane, Toronto, Singapore: 331-37

Treguer P, Nelson DM, van Bennekom AJ, DeMaster DJ, Leynaert A and Queguiner B (1995) The silicate

balance in the world ocean: A reestimate. Science 268 : 375-379

Benthic silica release - estimation of non-litho genic partic!ef/1txes 115

van Bennekom AJ and Berger GW (1984) Hydrography and silica budget of the Angola basin. Netherlands

Journal of Sea Research 17 : 149-200

van Bennekom AJ, Jansen JHF, van d Gaast SJ, van Iperen JM and Pieters J (1989) Aluminium-rich opal: an

intermediate in the preservation of biogenic silica in the Zaire (Congo) deep-sea fan. Deep Sea Research

36: 173-190

van Cappellen P and Qiu L (1997) Biogenic silica dissolution in sediments of the Southern Ocean - 1. Solubility.

Deep Sea Research II 44 : 1109-1128

van d Gaast SJ and Jansen JHF (1984) Mineralogy, opal, and manganese of middle and late quaternary

sediments of the Zaire (Congo) Deep-Sea Fan: Origin and climatic variation. Netherlands Journal of Sea

Research l7 : 313-341

Wagner T, Zabel M, Dupont L, Holtvoeth J and Schubert CJ (in press) Terrigenous signals in sediments of the

low latitude Atlantic - Implications for environmental variations during the late Quaternary - Part!:

Organic Carbon. In: Wefer G, Mulitza Sand Rathmeyer V (Eds.): The South Atlantic in the late

Quaternary: Reconstruction of material budget and current systems. Springer

Wefer G and Fischer G (1993) Seasonal Pattems of vertical Particle Flux in equatorial and coastal Up welling

Areas of the Eastern Atlantic. Deep Sea Research 40 : 1613-1645

WenzhOfer F (1999) Biogeochemical Processes at the Sediment Water Interface and Quantification of

Metabolically Driven Calcite dissolution in Deep Sea Sediments. Berichte aus dem Fachbereich

Geowissenschaften 138. Universitat Bremen.

WenzMfer F and Glud R (2002) Benthic carbon mineralization in the Atlantic: A synthesis based on in situ data

from the last decade. Deep Sea Research I 49: 1255-1279

WenzhOfer F, Adler M, Kohls 0, Hensen C, Strotmann B, Boehme Sand Schulz HD (2001) Calcite dissolution

driven by benthic mineralization in the deep sea - In situ measurements of Ca2+, pH, pCO l , O2 ,

Geochimica et Cocmochimica Acta 65: 2677-2690

Zabel M, Dahmke A and Schulz HD (1998) Regional distribution of phosphate and silicon fluxes across the

sediment water interface in the eastem South Atlantic. Deep Sea Research 45 : 277-300

116

2.4

FIlIxes at the benthic bOllndmy layer

Fluxes at the benthic boundary layer - A global view

from the South Atlantic

C. Hensenl,a, K. Pfeifer1, F. WenzhOfer2

.b

, A. Volbers 1•c

, S. Schulz 1, 1. Holstein1

, O. Romero1,

K. Seiter1

1 Fachbereich Geowissenschaften, Universitat Bremen, KlagenfUlier Str., 28334 Bremen, Germany

2 Max Planck Institut fUr Marine Mikrobiologie, Celsiusstr. 1, 28359 Bremen, Germany

a Present address: GEOMAR - Forschungszentrum fUr Marine Geowissenschaften, Wischhofstr. 1-3,

24148 Kiel, Germany

b Present address: Marine Biological Laboratory, University of Copenhagen, Strandpromenaden 5,

3000 Helsing0r, Denmark

C Bundesanstalt fUr Geowissenschaften und Rohstoffe, Stilleweg 2, 30655 Hannover, Germany

The South Atlantic in the late Quaternary

(G. Wefer, S. Mulitza, C. RUhlemann, Springer in press)

Abstract

Fluxes between the ocean waters and the sediments are key regulation processes for the marine biogeochemical

cycles and, thus, their quantification is of crucial importance. At this transition it is ultimately detennined how

much of a primary particulate signal is preserved or mineralized and hence recycled. Our review summarizes two

major approaches how to use spatial information obtained from surface sediments: (1) In the first part we

summarize the state-of-the-art regarding the use of biogenic barium as a proxy for primary productivity. We

discuss the possibilities and limitations of this approach mainly based on the results of a recent study in the

South Atlantic. The general outcome of this study was that the spatial pattern of primary productivity can well be

traced back by calculating (sub-)recent accumulation rates of biogenic barium and applying available and newly

formulated empirical equations. Most of those equations, however, fail to give the really observed magnitude of

today's productivity values. The main reasons for this are mostly the uncertainty of the Corg/Babio depth relation,

which differs between distinct ocean regions, dynamic sedimentary processes at ocean margins combined with

badly constrained values of terrigenous barium input, and the effect of barite dissolution due to subsequent

anoxic diagenesis. To improve the quality of prognoses for past productivity multi proxy approaches are

recommended to bypass the uncertainty in predictions from a single proxy. (2) The more extensive second part is

based on the large amount of studies that aimed at the quantification of benthic fluxes of nutrients and oxygen,

which are good measures for the amount of reactive particulate material being mineralized at the seafloor and

thus returned into the marine cycle. Those results enabled us to give profound calculations of the benthic oxygen

consumption and the release of nitrate, phosphate, and silicate at the seafloor of the South Atlantic and give

upscaled estimates for the global area of the sea floor. Additionally, we discuss more detailed studies focusing on

control parameters for benthic fluxes like primary production and lateral advection along the ocean margins off

Flllxes at the benthie bOl/l1dm)' layer 117

Southwest Africa and Argentina. A very conspicuous result was obtained by calculating mass balances for

biogenic opal in those regions indicating a dramatic underestimation of accumulation fluxes of opal by

"conservative" methods, which is believed to be of global significance. The last section mainly focuses on the

effect of benthic mineralization on the dissolution of calcium carbonate even above the chemicallysocline. This

process is in discussion since more than two decades. A number of studies have been performed, mainly using in

situ devices, to determine CaC03 dissolution. We summarize and discuss the results obtained from the South

Atlantic and use a recently developed empirical algorithm to show the worldwide distribution of supralysoclinal

CaC03 dissolution fluxes in marine surface sediments and give an estimate of their total amount. Finally, a table

for benthic fluxes of major constituents is provided on ocean wide and global scales.

118 ZlIsammenjasslIng

3. Zusammenfassung

Im Rahmen dieser Arbeit wurden drei aufeinander aufbauende Themenkomplexe bearbeitet. Zunachst wurde ein Ansatz zur Regionalisierung und Quantifizierung des Gehaltes organischen Kohlenstoffs (TOC) im Oberflachensediment entwickelt. Die Diskretisierung des Weltozeans in 33 benthische TOC-Provinzen stlltzt sich auf international verfUgbare Daten zur TOC-Konzentration, Partike1fallen der unteren Wassersaule, Sedimentparameter, Porenwasseranalysen und Informationen zu sogenannten Randparametern Wle Stromungssystemen und lateralen patiikularen Eintragspfaden. Es wurden 5553 TOC-Daten kompiliert und eine Diskretisierung unter Einbeziehung der genannten Parameter vorgenommen. Nach der so erfolgten Vor-Charakterisierung in benthische "Roh"-Provinzen konnten die benthischen TOC-Provinzen durch die sich anschliel3ende regional spezifische geostatistische Daten-Analyse (Semi-Variogramm Analyse) nach der Kriging-Methode definiert werden. Die Erfassung globaler Verteilungsmuster des TOC-Gehaltes im Oberflachensediment in einer 1 ° x l°-Auflosung erfolgte durch die Anwendung der Kriging­Methode auf der Grundlage der regional-spezifischen Semi-Variogramm Analysen. So konnten im Rahmen dieser globalen Betrachtung die einflussnehmenden Randparameter berucksichtigt und ortsabhangige Strukturen herausgearbeitet werden. Die TOC­Verteilungskatie bietet daher ein sehr hohes Mal3 an Genauigkeit und bildet fUr weiterfuhrende Untersuchungen zu regional abhangigen Mineralisationsprozessen und Partikelflussen zum Sediment eine entscheidende Grundlage.

Erschwerend fUr die Darstellung moglicher kleinraumiger Strukturen, die regionale und globale Velieilungsmuster pragen, erweist sich die oft nur geringe Anzahl zur VerfUgung stehender Daten geochemischer Parameter. Da das geochemische Milieu in den obersten Sedimentabschnitten entscheidend durch die Mineralisation organischer Substanz mit Sauerstoff gepragt und hierdurch eine Vielzahl frlihdiagenetischer Prozesse gesteueli wird, konnte der in hoher Datendichte verfUgbare Parameter TOC im Oberflachensediment als Kontrollparameter fUr weiterfUhrende Untersuchungen zu Stofffluss-Verteilungsmustern verwendet werden.

So wurde der Frage nachgegangen, ob sich der Fluss partikularen organischen Materials (Jpoc) zum Sediment uber die Konzentration organischen Kohlenstoffs nTI

Oberflachensediment und die dort stattfindende Sauerstoffzehrung (DOU) abschatzen lasst. Um dies zu beantwOlien, wurde der enge konelative Zusammenhang zwischen TOC-Gehalt im Oberflachensediment und Sauerstoffzehrung genutzt. Ausgehend von den entwickelten Provinzen wurden fUr Wassertiefen unterhalb 1000 m in 10 charakteristischen Regionen gute Konelationen zwischen dem TOC-Gehalt im Oberflachensediment und den diffusiven Sauerstoffflussen liber die Sediment-Wasser Grenzschicht nachgewiesen. Ein signifikanter ratenlimitierender Einfluss des Sauerstoffgehaltes im Bodenwasser auf die diffusiven Sauerstofffllisse, war lediglich fUr den Bereich des NW-amerikanischen Kontinentalhanges (NE-Pazifik) zu berucksichtigen. Durch die Anwendung und Ubertragung del' insgesamt 11 empirischen Gleichungen auf Gebiete mit unzureichender Datendichte, in bezug auf die direkte Abschatzung benthischer Stofffllisse, konnte libel' die Berechnung del' benthischen Sauerstoffzehrung (DOU) aus TOC-Verteilungsmustern, global ein minimaler partikularer

Zusammenfassung 119

Fluss organischen Materials (JPOCa) von 0.5 GtC yr-1 zum Sediment abgeschatzt werden. Darliberhinaus ergab sich eine globale Einbettungsrate von 0.002 bis 0.12 GtC yr-1 in Abhangigkeit der zugrundeliegenden Parameter (SR, TOC). Die provinz-spezifische Anwendung empirischer Gleichungen flir die Abschatzung globaler Verteilungsmuster ermoglichte so die Berlicksichtigung regionaler Prozesse, die das organische Material auf seinem Weg aus der euphotischen Zone bis an die Sediment-Wasser Grenzschicht beeinflussen. Es konnte vor all em gezeigt werden, dass regional gesteuerte Prozesse, wie z.B. lateraler Sedimenttransport, Verrugbarkeit des organischen Materials oder Unterschiede im Transportverhalten der organischen Partikel durch die Wassersaule, entscheidend zu einer "Entkopplung" zwischen den Verteilungsmustern der Primarproduktion der euphotischen Zone und benthischen Verteilungsmustern des partikularen organischen Materials beitragen. Die im Rahmen dieser Arbeit entstandenen Verteilungskarten berlicksichtigen diese Prozesse.

Die Ergebnisse der abgeschatzten globalen Fliisse partikularen organischen Materials zum Sediment (JpOCa) sind als vergleichbar mit bisherigen Studien anzusehen. Die regional en Abweichungen zwischen den Primarproduktionsmustern bzw. Jpoc-Velieilungsmustern, die auf der Annahme eines rein vertikalen Flusses durch die Wassersaule basieren, und den hier ermi ttelten V erteil ungsmustern verdeutlichen die N otwendigkeit neben ozeanographischen Parametern auch SteuergroJ3en zu berlicksichtigen, die benthische Mineralisations- und Losungsprozesse beeinflussen (z.B. Lateraltransport, Qualitat des organischen Materials).

Der Ansatz zur Jpoca-Abschatzung anhand des diffusiven Sauerstoffflusses iiber die Sediment-Wasser Grenzschicht wurde fiir den Bereich des S-Atlantik anhand von Untersuchungen zur Kopplung zwischen mikrobieller Aktivitat und benthischen diffusiven Siliziumfliissen weiter entwickelt, urn eine einfache empirische Beziehung zwischen SiliziumrUckfluss und diffusivem Sauerstofffluss iiber die Sediment-Wasser Grenze herzuleiten. Zusatzlich zur aeroben Mineralisation des organischen Materials wurden der so bestimmte Riickfluss von Silizium sowie KalzitlOsungsraten als regional aquivalent zu den jeweiligen minimalen Partikelfliissen (Opal, Kalzit) zum Sediment angenommen und beschreiben in der Summe den minimalen biogenen Partikelfluss zum Sediment.

Unter Einbeziehung der jeweiligen Einbettungsraten konnte gezeigt werden, dass flir groBe

Bereiche entlang der Kontinentalrander die hier abgeschatzten minimalen Partikelfliisse herkommlich ermittelte Akkumulationsraten (Abschatzung iiber die Sedimentationsrate) iibersteigen. Die Abbweichungen lassen sich dadurch erklaren, dass der Hauptanteil des akkumulierten biogenen Materials im Oberflachensediment remineralisiert wird und bei Betrachtungen herkommlich bestimmten Akkumulationsraten nicht berlicksichtigt wird.

120

o Lage der Stationer

Toe [wt"Io)

< -0.25 0.25 -0.5 0.5 - 0.75 0.75 -1 1 - 1.25 1.25 - 1.5 1.5-2 2-2.5 25-3 3 - 3.5 3.5 - 4 4-4.5 4_5 - 5 5 - 5.5 5.5 - 6 6-6.5 6.5 -7 >7

Jpol' [g lli'Jalu:'l ~'0.25 0.25 -0.5 0.5 - 0.75 0.75 -1 1 -1.25 1.25 - 1.5 1.5 -1.75 US -2 2 - 2.25 225-2.5 2.5 - 2.75 2.75 - 3 3 -3.5 3.5-4 4-4.5 4.5 -5 5-6 6-8 8 -12 12-20 >20

Kartenanhang

4. Kartenanhang

Longitude (OE)

Longitude (OE)

A I: Verteilungskarte der TOC­Konzentrationen im Oberflachensediment fUr den SW­Afrikanischen Kontinentalrand in 0.1 ° x 0.1 ° Auflosung. Die Karten entstanden im Rahmen des 1. Manuskriptes (Abschnitt 2.1) durch Anwendung der dort beschriebenen Kriging-Methode. Aufgrund der hohen Datenlage entlang des Kontinentalhanges (mehrere Stationen innerhalb 1°) werden regionale Strukturen dort besser sichtbar.

All: Darstellung des Flusses organischen Kohlenstoffs (JPOCa) zum Sediment (0.1 ° x 0.1°). Die J pOCu Karten 11, IV, VI entstanden auf der Grundlage der jeweiligen TOC­Verteilungskarten im Rahmen des 2. Manuskriptes (Abschnitt 2.2) und des dort beschriebenen Verfahrens zur Abschatzung der C-Fltisse zum Sediment.

Kartenanhang

o Lage der Stationen

TOe [wt"Io) _< -0.25 _ 0.25-0.5 _ 0.5-0.75 III 0.75-1 _1-1.25 _ 1.25-1.5 _ 1.5-2 _2-2.5 l~~~~ 2.5 - 3 lili~~ 3 -3.5

3.5 - 4 .. 4 -4.5

.4.5-5 _5-5.5 1115.5-6 111 6 - 6.5 _6.5-7 _>7

Jr(lC [gni'Jah,:'J

.'::0.25 • 0.25-0.5 .0.5-0.75 .0.75-1 .1-1.25 1IIIIIIII1.25 -1.5 1IIIIIIII 1.5 -1.75 Im 1.75-2 1IIIIIIII 2 - 2.25 1m2.25-2.5 1IIIIIIII 2.5-2.75 1IIIIIIII2.75-3 1m3 -3.5 .. ~ 3.5-4

...... ; 4 -4.5 ···"':'·j4S - 5 .5'-6 1IIIIIIII 6 - 8 1IIIIIIII 8 - 12 1IIIIIIII 12 - 20 .>20

Longitude (OE)

A Ill: TOC-Verteilung im o berflachensediment des nordlichen N-Atlantik (Gronland-N orwegische­Island See) (0.1° x 0.1°, vgl. Manuskript 1, Abschnitt 2.1).

121

A IV: JPoca-Abschatzung im nordlichen N-Atlantik (0.1 ° x 0.1 0, vgl. Manuskript 2, Abschnitt 2.2) .

122 Kartenanhang

A V: TOC-Verteilung im Oberflachensediment der Arabischen See (0.1 ° x 0.1 0). Es wurden hauptsachlich die Strukturen der ostlichen Arabischen See herausgearbeitet (Verfahren: vgl. Manuskript 1, Abschnitt 2.1).

Longitude (OE)

A VI: JPOCa-Verteilung basierend auf der TOC-Verteilung in 0.1 ° x 0.1 o-Auflosung (Abschnitt 2.1 und Abschnitt 2.2).

Longitude (OE)

o Lage der Stationen

TOC [wt"Io) _< -0.25 _ 0.25-0.5 .0.5-0.75 .0.75-1 111 1 -1.25 1111.25 -1.5 1111.5 - 2 111 2 - 2.5 ¥tl~~~% 2. 5 ~ 3

~~:~ ~.5~·~ : 4 -4.5 1114.5-5 1115 -5.5 .5.5-6 .6-6.5 .6.5-7 _>7

J i'()e [g lli'Jah,:'] _~'0.25 _0.25 -0.5 _0.5-0.75 _0.75-1 _1-1.25 IR1.25 -1.5 IR1.5 -1.75 IR1.75 -2 IR2-2.25 IR2.25-2.5 IR2.5-2.75 IR2.75-3 IR3-3.5 ~·13.5-4

14 -4.5 ··~··~i 4.5 - 5

-t~ .8-12 _12-20 _>20

Kartenanhang

B I: TOC-Gehalt im Oberflachensediment « 5 cm Sedimenttiefe).

o 00-.....

o \0

0-

0 \0-

I

0 C"l_ .....

I

0 00_ .....

I

0 0

\0 I

g I I I

123

0 -00 ..... r--

1\

r--I

lr'l \Ci

0 -C"l ..... \0

I lr'l tri

-0

Cil 0

lr'l ..... -\0

I

lr'l ~ .....

I .....

0 -C"l lr'l .....

I r--c? lr'l 0 ,......,

;::R

:t u lr'l

0 0 C"l 0 _00 ~ ..... V I

co >---' IV

~ .j::o.. ~

30 60 90 120 150 180 [/) ~ I I I I I I C

-180 I

-150 I

-120 I

-90 I

-60 I

-30 I

o I

(1) >-; 'J] .-+ 0 :::i:l

(JQ (1)

0"' ~ ::+ S· td 0 0... (1)

::s ~ ~ 'J] 'J] (1) >-;

~

td 0 n

'--'

+--60

I i i I I I I I I I I i I

-180 -150 -120 -90 -60 -30 0 30 60 90 120 150 180

Bottom water oxygen concentration [mmol ni'] I

~ .." ~ ::: \:l

0-20 40-60 80 - 100 120 - 140 160 -180 200 -220 240 -260 280 -300 >300 ::: :::-\:l

~

Kartenanhang

BIn: Minimaler partikuHirer Fluss organischen Kohlenstoffs zum Sediment (JPOCa).

~ 0

0 \0 I

0 I I I

00- + .......

0 C'-l-.......

I

0 00_ .......

I

0 -00 .......

0 -C'-l .......

-~

-0

o -\0

I

0 -C'-l .......

I

0 _00

....... I

0 C'-l 1\

~

C'-l .......

00

I.Ij

I.Ij

-.::t

-.:::t

111 ('f')

('f')

I.Ij r---~

I.Ij C'-l ~

.......

I.Ij c-q .......

I.Ij r---0

..r::;-' 'I-< >0

("

',-. c: I.Ij

60 C'-l '--' 0 U V

125

ApQendix I > >-' N

distanee/ max. "0 0\

d' I seareh- "0 Zone model hmax

lag Iw Co

Ir a ratio a ~h seale slope

data no. of ('!l

n width [0] dirroJ

Radii points/ Seetors == aniso aniso. (SR) ~

unit seetor ..... i><

Polar Seas ""'" ,-... BARENTKARA 282 exponential 500000 [m] 20000 25 0.04 n.sp. 313500 0.12 2 range 4 4 '" om n.sp. .....

('!l

33 0.008 =-KARA2 90 spherical 330000 [m] 110000 5 om n.sp. n.sp. 110000 0.02 range 4 4 ('!l

> LAPTEV SEA 161 spherieal 500000 [mJ 20000 25 0.025 146 35 n.sp. 150000 0.08 n.sp. n.sp. n.sp. 0-

'" GROE 2 n

145 spherical 1200000 [m] 40000 30 0.009 85 65 n.sp. 750000 0.05 range 4 4 =-== ANT 266 nugget 1110000 [m] 0.085 222000 n. sp. n.sp. ..... ..... .....

South Atlantic areas ~ I-"

SOATL 258 exponential IS n 0.50 30 0.010 om n.sp. n.sp. 8.55 0.055 1.5 range 4 4 '-' Ul

SOATL + Coast 911 spherical 15 n 0.50 30 0.050 om 8.00 0.15 1.5 range 6 4 C/J . n.sp. n.sp. c 0 S

NAMBCO 139 spherical 5 [0] 0.2 25 0.005 135 50 3 135 3.66 0.125 range 2 4 S ~ l'>

...,.. SWACO (incl. NAMBCO. ::l. ('e

285 spherical 15 [0] 0.50 30 0.02 119 45 n.sp. 5.00 0.10 range 4 4 N = inel. intersect. ETROPAT) (!) p.. ~

ARGCO (incl. RIOPLATA) 73 spherical 15 [0] 0.75 20 0.02 45 60 n.sp. 10.00 0.23 range 4 4 -< = P' =-RIOPLATA 50 spherical 3 [oJ 0.20 15 0.001 153 60 2.00 0.15 2 4 ::l. ~ n.sp. range 0

(Jq = 4 4 ..,

CJC/ BRAZCO 144 nugget 10 n 0.09 10 l'> S

GUBRACO 190 spherical 10 [0] 0.40 25 0.06 118 60 n.sp. 3.00 0.06 range 2 4 '0 .., SRI:2;

0 '0 (!)

GUI 79 10 [0] 0.085 SR2:1O

4 2 :4-nugget angle: (D'

m

90° l'> ;:;

ETROPAT 170 linear 10 [0] 0.40 25 0.01 16 45 n.sp. n.sp. n.sp. 0.02 asGUI 4 2 p.. m (!) ;:::I: ::r

(Jq m C m

b (!) p..

~

8' ~ .., ::s ~.

~ ::s ;:;..

(!) ~ ~ ~ 0

[ o· ;:; '0 .., 0 0 (!) p.. C .., (!)

'"

Appendix I continued

distance! Zone n model hrnax lag width Iw

unit

;\"ortlz At/unlic areas

NOATL+ Coast 1762 exponential 30 [oJ 1.20 25

NEAl'vlCO 105 spherical 15 [O) 0.50 30

NOATL 385 exponential 30 raj 1.0 30

WAFCO 64 nugget 15 1°1

CANAR 85 spherical 4 1°] 0.16 25

fURl 114 Gauss 4 f01 0.13 30

FUR2 238 exponential 2 fOI 0.04 50

Pacific areas

NWAMCO 68 exponential 5 n 0.33 15

Cl [fCO and PERCO without 125 spherical 12 ["I 0.60 20

anisotropy- a

CHlCO with anisolropy- h 113 spherical 12 1°] 0.4 30

NW-iNEPAC' 243 exponential 40 10J 1.60 25

TROPAC (cast) 183 cxponential 25 1°1 1.00 25

TROPAC2 (\ycsO'

SE-!S\VP;\C 79 11. sp. raJ

d' I ratio C

Ir (J. (t i'\h Q fO] dir fOJ ,lniso al1l~O

0.065 om I1.Sp. n.sp. 7.13

0.18 om n.sp. ~ 55 10.3 .' 0.015 om n.sp. 1I.8p. 14.25

0.196

0 74 30 2 74 2.2

0.043 om n.sp. n.sp. 3.98

0.01 om n.sp. n.sp. 1.14

0.005 113 45 2 113 2.14

0.01 om n.Sp. n.sp. 10.00

0.006 75 30 " 75 6

0.03 om n.sp. n.sp. 17.10

0.()2 om n.sp. 0 0 1 1.4 .)

Kriging linear with adjaccnt provinces

search-scale slope Radii

(SR)

0.10 range

o.n range

0'()6 range

10

0'()4 range

SRI:6:

0.35 SR2:3: angle: 1100

0.20 2 range

0.12 2 range

0.12 range

SR1:6:

0.08 SR2:2: angle:

75°

0.08 range

0.098 range

15

max. data no. of

points/ Sectors

6 4

3 4

6 4

all data n.sp. ine!.

4 2

8

4 4

2 4

3 2

4 2

6 4

4 4

n.sp. n.sp.

U ~ (j) ::l ~ ::l ::l" ~ ::l

(JCj

>-'

N -...}

Appendix I continued

Zone n model

Ifldi c areas

INn) 223 spherical

EARAB 206 spherical

WARAB 121 exponential

SEAFCO 94 nugget

G!o/Jal ari!a

East of zero meridian 4 [334 n.sp

West of zero meridian 4 2670 n.sp

j: direction with zero on the x- axis

2: (;ROE: > 550 In \Vd

3: processed with Longitude corr. =Longitude-360c

4: distortiol1l1ot corrected (within 65°$ and 65°N) 5: including SOMALlCo, TANZACO, EICO

11

hmi/X

wd: water depth, n.sp. not specified, SR: search radii, om: omnidirectional

1w

Co

dirrl

(1 dw /,,1 r(Jti()al!I,~I)

0: <1J/1.1'O

20

3

10

70

70

distance! lag width [w

unit

[oJ 1.00

raJ 0.10

ID] 0.50

20

30

20

dir l ex Co fOI dir [Cl

0.04 om n.sp.

0.009 om Il.sp.

0.02 om n.sp.

n 1.4 50 0.06 om n.sp.

ratio u Llh

aniso aniso

n.sp. lO.OO

2 125

n.sp. 4.56

n.sp. n.sp.

scarch-max. data no. or

scale slope Radii points! Sectors

(SR) sector

0.(l4 2 range 6.00 4

SRI:I

0.18 SR2:3

4 2 angle: 215 0

(),09 2 range LOO 4.00

SRl:8; SR2:4;

4 2 angle:

75°

incl. all lOO 4

data n.sp.

incl. all I" 1 1.4 50 0.13 0111 n.sp. n.Sp. n.sp. n.sp.

data 100 4

number ofdala points

maxil1lul1llag distance

lagwidth

number of lags (number of computed variogram points)

nugget el1ect

direction of the variogram view

size (lUhe angular window

unisotropyellipse

direction of ellipse orientation

Llh

scale

slope

search radW and angle

110. o.lsectors

max. data pOillls/secror

mode!

dislal1ce!lag lridth unit

range

vertical scale orthe variogram (sill+nugget etTect)

in the case ofa linear variogram=scale

search ellipse: radius horizontal and vertical extend in d,lla units l{)r data search/rotation of the search ellipse

defines the number of sectors for data search

specifies the number of data used from each sector

theoretical i1ttcd semi-variogram model

unit

N 00

\:J I::)

~ ~ :::;

~ ~

Datenanizang 129

Appendix II-IV (siehe Abschnitt 2.2)

II: SUlmnarized budgeting and reference data: Primary production (PP) after Behrenfeld and Falkowski (1997-a,b), PP after Antoine et al. (1996), particulate organic carbon flux to the sea floor (hoe) recalculated after Antia et al. (2001) based on PP after Antoine et al. (1996), Jpoc recalculated after Jahnke (1996) and calculated JPOCa of this study. III: Calculated mean values of each province: PP after Behrenfeld and Falkowski (1997-a,b) and after Antoine et al. (1996), particulate organic carbon flux to the sea floor (Jpoe) recalculated after Antia et al. (2001) based on PP after Antoine et al. (1996), hoc recalculated after Jahnke (1996), calculated JPOCa of this study and calculated mean oxygen bottom water concentration, queried from the 10 x 10 grid. IV: Collection of mooring sites, hoc and duration of trap sampling, JPOCa estimated (this study), Jpoc recalculated after Antia et al. (2001) at the trap sites and data sources (Eds.)

Appendix Il ProvinL:; Province calculated

description (> 1000111 wd)

CiROF' CTrcenland Sea 1963860 3.94E-OI

LAPTEV SEA Laptcv Sea 185066 no data

KAR/\2 Kara Sea 15357 no data

BARI:NTKARA Barent Sea and Kura

98439 no data Sea

NWAMCO con!. marg. olfNW-

1088530 U5E-()] America

NEPAC N [,-Pacific Ocean 25750300 3.17E+OO

NEAMCO con!. marg, ofl N 1:-

551546 1.05E-OI America

SWACO" con!. marg, off S W-

1929801 6,78F-Ol Africa

IND Indian Ocean 55354800 5,79F+OO

NWPAC NW-Paeilic Oc~an 20552300 2.45F+OO

GUl Guinea Basin 200633 5,17E-02

UllCO conI. marg, olH'hile 356027 l,nE-OI

conI, marg, off G1.mRACO Guyana, Venezuela. 1084600 1.191:-01

Surina1ll

WAFCO cont marg, ofTW-

965796 4,8IE-Ol Africa

SOATL southern Atlantic

37269900 4.501'+00 Ocean

ARGCO coni. marg. oir

556267 6.46L-02 Argentina

BRAZCO conI, marg, ofT Brazil 1044870 9,89E-02

NOATL northern Atlantic

27971100 4571'+00 Ocean

Jpoc

2.21:-01 2.7F-03 llO data no data

no data no data no data no data

no data no data 110 data no data

no data no data no data no data

1.82E-OI 5.28L-03 4.UE-03 5.29E-03

2.231:+00 23 I E-02 2.10E-02 269E-02

7.081':-02 1.681::-03 I.2IE-03 1.55F-03

4.05F-OI 1.32F-02 5.491'-03 7,05E-03

5.39E+OO 7.38[-02 6.86E-02 8.811'-02

1 98E+OO 2.26E-02 205F-02 2.63E-02

3,97[-02 1.07[-03 9,03[:-04 1 16[-03

6.691'-02 1,911'-03 3.171'-04 4,071'-04

1.31E-01 2.8IF-03 2,21 E-03 2,841'-03

2.341'-01 9.46[-()3 2,61F-03 334E-03

3.77E+OO 4.85E-02 3,881'-02 498E-02

4,21E-02 6.85E-04 5.15E-04 6.621-:-04

121£'-01 226E-03 USE-03 I 77E-03

;>, 95F+00 409F-02 338E-()2 434F-02

JP(lCu

(this stud,,)

3.66E-03 0.93

2.611:-04 no data

no data no data

no data no data

4.IOE-()3 UN

HOE-02 1.20

1. lOF-03 I,OS

1.001:-02 1.48

7.85E-02 1.36

370E-02 1.51

7.50[-04 1.45

1.571:-03 1.29

1,80F-03 U2

2.30E-03 0,.18

5 55E-02 1.23

g,99E-04 139

190E-03 1.92

395[-02 0.86

2.18

HO data

no data

IlO data

2.25

1.70

1.55

2.47

1.46

I 87

1,89

2.35

1.38

0.98

1.47

2.13

1.57

I 34

,....... UJ 0

to ~ '" 2 :::l

is ~

Appendix II continued b !::)

Province Province calculated area' PP' pp3 Jpex• Jpoe 5 Jpoe

(> JpOCo. JpeX,,!PP' J"oculP!" (t

description (> 1000 III wd) (this study) :::: §

[CitC y(', [CitC yf' 1 IGtC y(' I [Citeyr", [CitC y(', [CitCyr"1 :::;-

Ikm"] 1%1 [%>1 !::l

CANAR Canary Islands area 422511 7.611:-02 5.11[-02 1.121'-03 7.561'-04 9.70[-04 1.0 11:-03 U3 1.98 ~

I'URl southern European

209682 5.111'-02 2.731'-02 conL marg.

6.291'-04 3.211'-04 4.111'-04 4.75E-04 0.93 174

EARAH eastern Arabian Sea 344618 6.071'-02 6.951:-02 2.641:-03 1.54F-O] 1.971:-03 4.471-03 7.36 6.43

EUR2 northern European 31\5925 1.18E-O I 5.11 E-(U

cont. luarg. 1.27E-03 6.58E-04 8.441:-04 6.261'-04 0.53 1.22

WARA13 western Arabian Sea 1889100 5.48E-Ol 4.8IE-OI U61'-02 5.52E-03 7.08E-03 2.23E-02 4.07 4.64

RIOPLATA Rio de la Plata 409923 7.20E-02 4.831'-02 9.881'-04 9.72E-04 1251'-03 1.071'-03 1.49 2.22

PERCO cont. marg. 011' Peru 355240 1.36E-0 1 8.25E-02 2.66E-03 1.22E-03 1.56E-03 3.23E-03 2.37 3.91

TROPAC castcrn tropicnl

30110000 3.191'+00 3.84E+00 5.95E-02 5.01 E-02 6.43E-02 7.IOE-02 2.22 1.85 Pacific

TROPAC2 western tropical

11809400 7.631'-01 1.25E+00 1.94E-02 1.041'-02 1341'-02 1.87E-02 2.45 1.50 Pacific

SEPAC SE-Pacific Ocean 53018800 5.31 E+OO 4.151:+00 4.58E-02 3.20F-02 4.IOF.-02 5.90F-02 1.11 1.42

SWP;\C SW-Pacific Ocean 8297580 9.261'-01 7.90E-01 1.33E-02 7.991'-03 1.03E-02 140E-02 1.51 1.77

SEAFCO conI. marg. oITSE-

1969750 2.451'-0 [ 2.281":-01 Africa

4.46F-03 3.95E-03 5.07E-03 3.67E-03 1.50 1.61

NAM13CO conI. marg. off

367829 9.72E-02 4.88£-02 1.68E-03 6.81 E-04 8.74E-04 2.04E-03 2.10 4.18 Namibia

EICO con!. margin ofT E-

[86078 30 I F-02 3.73E-02 India

1.54E-03 6.46E-04 8.29E-04 1.7 [E-03 5.68 4.59

ETROPAr E-Equatorial Mlantic 3460340 7.90£-01 6.11E-Ol 1.40E-02 7.25E-03 9.30E-03 9.00E-03 1.14 1.47

TAN7.ACO con!. marg. oil'

2107830 1.93F-01 2.34E-Ol 3.70E-03 3.70E-03 4.741'-03 6.281:-03 3.26 2(i8 Tanzania

SOMALICO conI. marg:. olT

183718 4.151:-02 3.88E-02 Somalia

1.21 LA)] 4.92L-04 6.311.>04 5.271:-04 1.27 1.36

ANT (area < 65°S) all areas> 65° N 4017892 1.431:-01 3.971'-0 I 5.381'-03 no data no data 5.2%-03 3.69 133

Sum 2961 [7579 3.7IE+Ol 3.12E+01 4.44E-O 1 3JOE-OI 4.23E-O 1 5.01'-01 135 1.61

I: average calculated cell area (> 1000111 wd): slight difTerences ofl11ean values are caused by diflerences in average integrated eell area. 2: primary production after Bchrcllfcld and Falkowski (1997-a,b). 3: primary production all er Antoine et al. ( 1996). 4:.Il'oc after Antia et al. (200 I) are based on primary productiou after ;\lltoillC et al. (1996) and \\ater depth (wd) fi'om [TOI'05. 5: hoc recalculated atter Jahnkc (1996) from a global oxygcu consumption grid with a CO, ratio of 0.6 as described by Jahnke (1996). 6: .lroc recalculated Idler Jahnke (1996) with a C:O, ratio of 0.77 based on this study. 7: .I eoe recalculated aflcr Schlliter et al. (2000) instead ofAnlia et al. (200 I) Jor GROE; wd> 550 111.

S: all budgets including N/\MBCO corrected in global budget and area. projection: Mollweide, eenter: zero meridian, all calculations> lOOO m wd. GROE; :> 550 m wd. all areas and all grids were resampled to a 2048 x ]()24 grid resolution.

w

Appendix III Province pp' pp' Jpoc

j

(mean) (mean) (mean)

IgC m·' y('] IgC Ill"' )"('] IgC nf' yr·']

GROE" 201 112 1.37

J.;\PTEV SEA no data no data no data

KARJ\2 no data no data no data

BARENTKAJZA 110 data 110 data no data

NWAMCO 345 167 4.85

NEPAC 123 87 0.90

NEAMCO 191 128 H)4

SWACO') 351 210 6.86

IN]) 105 97 133

NWP;\C 119 96 UO

GUI 258 198 5.34

ClHeO 342 188 5.36

GUBRACO 109 121 2.59

WAFCO 498 243 9.79

SOATL. 121 101 1.30

ARGCO 116 76 1.23

BRA7CO 95 116 2.16

NOATL 164 106 1,46

CANAR 180 III 2.65

EURI 244 130 3.00

EARAB 176 202 7.65

EUR2 305 132 3.29

JI'OC , Jroe;

(mean) (mean) C/O,:0.6 ClO,:(I.77

IgC nI"' )"('] IgC n;·2 y('1

no data no data

no data no data

no data no data

no data no data

3.79 4.86

0.81 un 2.20 2.89

2.85 3.67

1.24 1.66

l.OO 130

4.50 4.43

0.89 1.0S

204 2.58

2.70 3.34

1.04 1.41

0.93 1.27

1.32 1.67

1.21 1.60

1.79 2.22

1.53 1.91

4.47 5.31

1.70 2.28

Jpoc(), TClC(' (mean) (mean)

this studv IgC m·' Y;·' 1 [wt%1

1.86 0.72

141 0.97

no data no data

110 data no data

3.77 1.56

1.48 044

1.99 0.65

5.18 1.22

1.42 0.35

1.80 0.58

3.74 l.OO

4.41 1.45

1.66 0.43

2JS 0.64

1.49 0.35

1.62 0.38

1.82 0.54

1.41 0.36

2.39 0.62

2.27 0.80

12.97 1,48

1.62 0.46

BOC' (mean)

this study 111111101111·']

310

316

202

275

103

154

228

206

203

163

214

192

207

212

231

249

219

257

213

223

93

249

........ V-J N

tJ t:) (i)

~ ::: ::0-t:)

~

Appendix III continued Pmvince ppl pp' Jpoe " Jpoc 4

Jpoc' JpoC't( TOe(' (mean) (mean) (mean) (mean) (mean) (mean) (mean)

C;02:0J, ClO,:(L77 this studv igCm,2 yr'l] [gem" y(IJ [gCm" y(l] IgC !If' y(l] [gem" y(IJ IgC nf';,;,IJ !wt""(,J

W/\RAB 290 254 9.31 2.92 3.79 11.80 1.36

RIOPL;\TA 176 118 2.41 2.37 3.0S 2,61 0.94

PFRCO 383 232 7.49 3.42 4.06 9.09 3.82

TROPAC 106 127 1.98 1.66 2.1,5 2.36 1.22

TROPAC2 65 106 1.64 088 1.[9 1.58 0.76

SEPAC lOO 78 0.86 0.60 08} l.ll 0.49

SWPAC 112 95 1.60 0.96 1.35 1.69 0.82

SEAFCO 125 116 2.27 2.01 2.52 186 050

NAMRCO 264 133 455 U5 3.52 5.55 1.87

F[eO 162 200 830 3.47 4,73 9.19 1.12

ETROPAT 228 177 4.06 209 2.72 2.60 0.64

T;\N7:ACO 91 III 175 175 2.24 2.98 1.05

SOMALlCO 226 211 6.58 2.68 3J9 2.87 0.71

ANT « 65°S) 36 99 134 no data no data 1.32 0.32

I: primal)' production after Behrenteld and Falkowski (1997-a.b). 2: primary production after Antoine et al. (1996). 3: mean Jpoc recalculated af\cr AlItia et al. (2001) is based Oll primary production aner Antoine et a!. (1996) and water depth (\Vd) /i'Olll

ETOP05. 4: mean .Ipot: recalculated after .Iahnke (1996) from a global oxygen cOl1smnption grid \vith a C:02 ratio of 0.6 as described by Jahnke (1996). 5: llleall .11'0(' recalculated atter .lahnkc (1996) with a C:O, ratio 01'0.77 hased on this study 6: mean 'roe values recalculated trom global TOe grid after Sciter et al. (submitted). 7: mean HOC values queried from the grid, this study. 8: mean Jpoc values for GROE recalculated afkr Schlliter et al. (2000). 9: SWACO: all mean values calculated including NAMBCO. all calculations corresponds to wd > 1000 m: GROE: wd> 550 m.

130C' (mean)

this study inullolm"]

lOO

202

[30

160

158

190

172

194

191

107

233

191

186

234

b ~ (';) :::: I:l :::: ;:;-8

(iQ

u..> u..>

Appendix IV Longitude' L-atitude' Trap Trap wd Trap depth Jp"c.,-",»' Jpoc </ JpO(:u- .I,·'oc' Editor start date end date COlTIlTIent

-151.0000 15.0000

-145.0000 50.0000

-140.0000 11.0333

-139.7500 -1.9500

-138.9317 1.0555

-13X.0000 49.0000

-127.7350 49.1 IS]

-127.7350 49.1183

-127.7000 39.5000

-127.7000 39.5000

-127.5800 42.1900

-125.7700 420900

-123.0000 34.8330

-104.0000 8.8000

-92.8000 6.5500

-86.0000 0.0000

-85.5833 5.3667

-7i.OOOO 39.0000

-71.0000 39.4167

no.

2

-'

4

5

(j

7

8

9

Im]

PI 5792

PAPA 4240

MANOP_ 4620 S

EqPac 92 3593

MANOP_ 4450 <.'3 PC 4008

10S_lrap_ 1978 05

J()S_.lrap.... 1.978 06

1V1I7.-4 4356

10 MF7-1 I 4230

II

12

J3

14

15

16

17

18

19

MW·I

NSI

IVI-S

MANOI' M

2830

21129

4100

31S0

MANOI'_ 3565 H

epac 2670

PS 3860

WD-] 2750

WD-2 2300

ahove bottom

Im]

210

386

10

o 530

50S

78

78

571

445

500

500

50

100

20

lOO

300

50

50

/JpnCT!ap trap trap

IgC nf' y('] IgC nf' yf'] 1%] rgC m'! yf'l

0.24 1.41 578 0.74 Ik'J1ioeta!.(1982):Francoiseta1.(1995) 01.09.78 01.11.78

2.19 2.19 lOO 1.07 Wong et al. (1999); Francois cl al. (1995) 23.09.82 ]6.05.94

[x]

intcranl1lwl variations

0.20

1.30

1.69

2.17

3.21

4.44

1.44

0.95

1.78

4.98

3.00

1.50

1.40

1.63

4.31

2.26

2.81

3.29

1.58

1.10

1.60

5.03

5.03

1.93

1.93

3.75

4.40

6.00

3.85

2.56

3.86

4.07

3.67

3.55

1613

122

65

74

157

113

134

203

211

88

200

257

183

237

94

162

126

1.09

2.55

2.20

1.10

7.72

7.72

1.(14

1.04

1.83

3.88

8.89

2.84

1.77

3.64

3.47

4.02

3.56

Emerson et al. (1985)

HOl1jo et al. ( ! (95)

Dymond et al. ( 1992)

Francois et al. (1995)

Pefia et al. (1996); Pefia et al. (1999); Pefia et 31. (20()O)

Pcna ct al. ( 1(96): Pena ct al. (1999): PClia ct al. (2000)

Dymond and Lyle (1994): Ragueneau et al. (2(}OO)

DYll10nd et al. ( 1(92)

Dymond and Collier (1988); Dymond et al. (1992)

])ymond and Collier (1988); DYlllond et al. (1992)

Smith ct al. (1994)

DYl110nd and Lyle (1985): Dymond et al. (1992): Emcrson ct 31. (1985)

Dymond and Lylc (1985); Dymond et al. (1992): Emcrsoll ct 31. (1985)

Collier and Dymond (1980); Emcrsoll and Bender (1981 )

Wcfcr(1989)

Walsh et a!. (1991)

Walsh et al. (1991)

10.06.80

annual

01.05.85

annual

03.10.95

14.G4.96

annual

annual

anllual

annual

26.02.90

12.09.80

20.09.80

annual

03.12.79

annual

annual

13.10.81 Ix]

water depth 29.(13.96 1Tol11 ETOPO

\Yakr depth 01.09.96 j[OI11 FTOI'O

24.10.91

23.10.81

17.10.81

02.12.80

5

Ix]

[x]

Ixl

Ixl

[xJ

Ixl

'"-' V.l ..".

b !::l (\) ::l !::l ::l ;:,.. !::l

~

Appendix lV continued Longitude' ~ Latitude'

136.2766 41250

154~8250

1 56.0()()O

174.9450

175.0000

175.1500

177.7366

6.8000

-7.7117

0.4000

0.4633

1.0667

13667

10.0000

114660

12.4867

15.0000

-15.4500

-17.7617

-13.0000

37.4033

7.9266

0.0000

34.4217

78.9000

723817

70.0000

69.6867

65.5167

78.8650

69.5000

75.8500

75.1967

74.0000

29.1167

I : Longitude: wcst,·o'negativc 2: I ,atitllde: sOllth~negative.

Trap no.

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

Trap

K2

K12

Kll

K7

K4

K3

K5

SI'

OG72

NB-a

NB-b

NA-I

FS-l

LB (I)

BI-I

BI-2

BS

\\d

!!:!1 4888

2821

1832

5105

5260

4880

3365

1676

2631

3269

3292

3058

2823

3161

2123

2000

1608

ESTOC 3600

Trap depth above bottom

~ 314

517

517

517

517

517

517

551

431

269

292

428

381

400

423

50

20

523

3: particulale organic carbon tluxe, collected by near bol1ol11 traps. 4: particulate organic carbon nuxe, this study.

JpoC' '(bp Jpoc'U JPOC5

iJp<X"!t:1P

3.98 2.01 50 1.61

0.37

0.99

2'()]

051

0.99

2.81

6.60

0.55

3.56

2.76

0.59

OAI

1.37

2.85

5.22

18.81

0.84

0.96

1.61

2.7R

0.97

0.93

2.30

5.67

1.32

1.86

1.99

2.49

3.37

1.91

3.02

2.10

2.00

2.63

264

163

139

190

94

82

86

240

52

72

422

822

139

106

40

11

313

1.36

1.61

0.80

0.71

2.48

128

S.IO

0.73

0.23

0.19

0.62

1.87

0.84

0.46

1.14

0.92

1.35

Editor

Kawahata (2002)

Kawahata (2002)

Kawahata (2002)

Kawahata (2002)

Kawahata (2002)

Kawahata (2002)

Kawahata (2002)

Antia et 31. (2001); I lebbdn (2000)

Antia et al. (2001): von Bodungen et a!. (1995)

'Ion Bodullgcn et al. ( 1995)

Thol1lscn ( 1993 )

Wefer (1989)

WeJcr (1989)

Wder (1989); Franeois et al. (1995)

WeJer (1989); Francoi, et al. (1995)

Thofllscn ( 1993)

Fnmeois et aI. (1995)

Ncuer et al. ( 19(7)

5: partieulatc organic carbon nuxes recalculated aner Antia et al. (200 It t()r the GrecnlanJ-Norwegian-lce]and Sea recalculated after Sehllilcr ct at. (2000). [xJ: not included in regression analyses and not ,hown in Fig. 10a,b.

stmi date trap

15'()4.92

16.05.95

01.04.96

09.04.94

13.04.93

160493

01.06.94

9.12.88

25.2.91

1986

6.8.91

19.8.85

10.8.84

15.8.83

12.8.84

16.3.91

18.5.89

cnd date trap

25.1 l.OO

16.03.96

1()11.00

08.11.00

17Jl7.00

14.Il.OO

15.12.00

9.12.89

25.2.92

1989

3.7.92

18.7.86

30.7.85

1.8.84

10.8.85

23,7.91

10.8.89

25.1 1.1991 906.1994

Comment

Ix]

Ix]

1.'1

Ix]

Ix]

1.'.1

1.'.1 Ixl

".I [x]

[x]

Ix]

b s:::,

~ ~ ::> ;:,... s:::,

~

>-" W Vl

In dieser Reihe bereits erschienen:

Nr.l

Nr.2

Nr.3

Nr.4

Nr.5

Nr.6

Nr.7

Nr. S

Nr.9

Nr.lO

Nr.11

Nr.12

Nr.13

Nr.14

Nr.15

Nr.16

Nr.17

Nr.lS

Nr.19

Wefer, G., E. Suess und Fahrtteilnehmer Bericht Uber die POLARSTERN-Fahrt ANT IV/2, Rio de Janeiro - Punta Arenas, 6.11. - l.l2.1985. 60 Seiten, Bremen, 1986. Hoffmann, G. Holozanstratigraphie und KUstenlinienverlagerung an der andalusischen MittelmeerkUste. 173 Seiten, Bremen, 1988. (vergriffen) Wefer, G. und Fahrtteilnehmer Bericht Uber die METEOR-Fahr! M 6/6, Libreville - Las Palmas, 18.2. - 23.3.1988. 97 Seiten, Bremen, 1988. Wefer, G., G.F. Lutze, T.J. Miiller, O. Pfannkuche, W. Schenke, G. Siedler, W. Zenk Kurzbericht Uber die METEOR-Expedition Nr. 6, Hamburg - Hamburg, 28.10.1987 - 19.5.1988. 29 Seiten, Bremen, 1988. (vergriffen) Fischer, G. Stabile Koh1enstoff-Isotope in partiku1arer organischer Substanz aus dem SUdpo1armeer (Atlantischer Sektor). 161 Seiten, Bremen, 1989. Berger, W.H. und G. Wefer Partike1fluB und Koh1enstoffkreis1auf im Ozean. Bericht und Kurzfassungen iiber den Workshop vom 3.-4. Juli 1989 in Bremen. 57 Seiten, Bremen, 1989. Wefer, G. und Fahrtteilnehmer Bericht Uber die METEOR - Fahrt M 9/4, Dakar - Santa Cruz, 19.2. - 16.3.1989. 103 Seiten, Bremen, 1989. Kolling, M. Modellierung geochemischer Prozesse im Sickerwasser und Grundwasser. 135 Seiten, Bremen, 1990. Heinze, P.-M. Das Auftriebsgeschehen vor Peru im Spatquartar. 204 Seiten, Bremen, 1990. (vergriffen) WiIlems, H., G. Wefer, M. Rinski, B. Donner, H.-J. Bellmann, L. Eifimann, A. Miiller, B.W. Flemming, H.-C. Hone, J. Merkt, H. Streif, G. Hertweck, H. Kuntze, J. Schwaar, W. Schiifer, M.-G. Schulz, F. Grube, B. Menke Beitrage zur Geologie und Pa1aonto1ogie Norddeutsch1ands: ExkursionsfUhrer. 202 Seiten, Bremen, 1990. Wefer, G. und Fahrtteilnehmer Bericht Uber die METEOR-Fahrt M 1211, Kapstadt - Funcha1, 13.3.1990 - 14.4.1990. 66 Seiten, Bremen, 1990. Dahmke, A., H.D. Schulz, A. Kolling, F. Kracht, A. Liicke Schwennetallspuren und geochemische G1eichgewichte zwischen Poren16sung und Sediment im WesermUndungsgebiet. BMFT-Projekt MFU 0562, Absch1uBbericht. 121 Seiten, Bremen, 1991. Rostek, F. Physikalische Strukturen von Tiefseesedimenten des SUdatlantiks und ihre Erfassung in Echolotregistrierungen. 209 Seiten, Bremen, 1991. Baumann, M. Die Ablagerung von Tschernoby1-Radiocasium in der Norwegischen See und in der Nordsee. 133 Seiten, Bremen, 1991. (vergriffen) Kolling, A. FrUhdiagenetische Prozesse und Stoff-FlUsse in marinen und astuarinen Sedimenten. 140 Seiten, Bremen, 1991. SFB 261 (Hrsg.) 1. Kolloquium des Sonderforschungsbereichs 261 der Universitat Bremen (l4.Juni 1991): Der SUdatlantik im Spatquartar: Rekonstruktion von StoffhaushaIt und Stromsystemen. Kurzfassungen der Vortrage und Poster. 66 Seiten, Bremen, 1991. Piitzold, J., T. Bickert, L. Briick, C. Gaedicke, K. Heidland, G. Meinecke, S. Mulitza Bericht und erste Ergebnisse liber die METEOR-Fahrt M 1512, Rio de Janeiro - Vitoria, 18.1. -7.2.1991. 46 Seiten, Bremen, 1993. Wefer, G. und Fahrtteilnehmer Bericht und erste Ergebnisse liber die METEOR-Fahrt M 1611, Pointe Noire - Recife, 27.3. - 25.4.1991. 120 Seiten, Bremen, 1991. Schulz, H.D. und Fahrtteilnehmer Bericht und erste Ergebnisse lib er die METEOR-Fahrt M 16/2, Recife - Belem, 28.4. - 20.5.1991. 149 Seiten, Bremen, 1991.

Nr.20

Nr.21

Nr.22

Nr.23

Nr.24

Nr.25

Nr.26

Nr.27

Nr.28

Nr.29

Nr.30

Nr.31

Nr. 32

Nr.33

Nr.34

Nr.35

Nr.36

Nr.37

Nr.38

Nr.39

Nr.40

Berner, H. Mechanismen der Sedimentbildung in der Fram-Stral3e, im Arktischen Ozean und in der Norwegischen See. 167 Seiten, Bremen, 1991. Schneider, R. Spatquartare Produktivitatsanderungen im ost1ichen Ango1a-Becken: Reaktion aufVariationen im Passat-Monsun-Windsystem und in der Advektion des Bengue1a-Kiistenstroms. 198 Seiten, Bremen, 1991. (vergriffen) Hebbeln, D. Spatquartare Stratigraphie und Pa1aozeanographie in der Fram-Stral3e. 174 Seiten, Bremen, 1991. Liicke, A. Umsetzungsprozesse organischer Substanz wahrend der Friihdiagenese in astuarinen Sedimenten. 137 Seiten, Bremen, 1991. Wefer, G. und Fahrtteilnehmer Bericht und erste Ergebnisse der METEOR-Fahrt M 2011, Bremen - Abidjan, 18.11.- 22.12.1991. 74 Seiten, Bremen, 1992. Schulz, H.D. und Fahrtteilnehmer Bericht und erste Ergebnisse der METEOR-Fahrt M 20/2, Abidjan - Dakar, 27.12.1991 - 3.2.1992. 173 Seiten, Bremen, 1992. Gingele, F. Zur k1imaabhangigen Bi1dung biogener und terrigener Sedimente und ihrer Veranderung durch die Friihdiagenese im zentra1en und ostlichen Siidat1antik. 202 Seiten, Bremen, 1992. Bickert, T. Rekonstruktion der spatquartaren Bodenwasserzirkulation im ostlichen Siidatlantik iiber stabi1e Isotope benthischer Foraminiferen. 205 Seiten, Bremen, 1992. (vergriffen) Schmidt, H. Der Bengue1a-Strom im Bereich des Wa1fisch-Riickens im Spatquartar. 172 Seiten, Bremen, 1992. Meinecke, G. Spatquartare Oberflachenwasseliemperaturen im ostlichen aquatoria1en Atlantik. 181 Seiten, Bremen, 1992. Bathmann, U., U. Bleil, A. Dahmke, P. Miiller, A. Nehrkorn, E.-M. Nothig, M. Olesch, J. Plitzold, H.D. Schulz, V. Smetacek, V. Spiefi, G. Wefer, H. WiIlems Bericht des Graduierten Kollegs. Stoff-F1iisse in marinen Geosystemen. Berichtszeitraum Oktober 1990 - Dezember 1992. 396 Seiten, Bremen, 1992. Damm, E. Friihdiagenetische Vertei1ung von Schwermetallen in Schlicksedimenten der westlichen Ostsee. 115 Seiten, Bremen, 1992. Antia, E.E. Sedimento1ogy, Morphodynamics and Facies Association of a mesotidal Barrier Island Shore face (Spiekeroog, Southern North Sea). 370 Seiten, Bremen, 1993. Duinker, J. und G. Wefer (Hrsg.) Bericht iiber den 1. lGOFS-Workshop. 1.12. Dezember 1992 in Bremen. 83 Seiten, Bremen, 1993. Kasten, S. Die Vertei1ung von Schwermetallen in den Sedimenten eines stadtbremischen Hafenbeckens. 103 Seiten, Bremen, 1993. Spiefi, V. Digitale Sedimentographie. Neue Wege zu einer hochauflosenden Akustostratigraphie. 199 Seiten, Bremen, 1993. Schinzel, U. Laborversuche zu friihdiagenetischen Reaktionen von Eisen (In) - Oxidhydraten in marinen Sedimenten. 189 Seiten, Bremen, 1993. Sieger, R. CoTAM - ein Modell zur Modellierung des Schwennetalltransports in Grundwasserleitern. 56 Seiten, Bremen, 1993. (vergriffen) WiIlems, H. (Ed.) Geoscientific Investigations in the Tethyan Himalayas. 183 Seiten, Bremen, 1993. Hamer, K. Entwicklung von Laborversuchen als Grund1age fiir die Modellierung des TranspOliverhaltens von Arsenat, Blei, Cadmium und Kupfer in wassergesattigten Sau1en. 147 Seiten, Bremen, 1993. Sieger, R. Modellierung des Stofftransports in porosen Medien unter Ankopp1ung kinetisch gesteuerter Sorptions- und Redoxprozesse sowie thermischer Gleichgewichte. 158 Seiten, Bremen, 1993.

NI'. 41

NI'. 42

NI'. 43

NI'. 44

NI'. 45

NI'. 46

Nr. 47

NI'. 48

NI'. 49

NI'. 50

NI'. 51

NI'. 52

NI'. 53

NI'. 54

NI'. 55

NI'. 56

Nr. 57

NI'. 58

NI'. 59

NI'. 60

ThieBen, W. Magnetische Eigenschaften von Sedimenten des ostlichen Slidatlantiks und ihre palaozeanographische Relevanz. 170 Seiten, Bremen, 1993. SpieB, V. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 23/1, Kapstadt - Rio de Janeiro, 4.-25.2.1993. 139 Seiten, Bremen, 1994. B1eil, U. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 2312, Rio de Janeiro - Recife, 27.2.-19.3.1993 133 Seiten, Bremen, 1994. Wefer, G. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 23/3, Recife - Las Palmas, 21.3. - 12.4.1993 71 Seiten, Bremen, 1994. Giese, M. und G. Wefer (Hrsg.) Bericht lib er den 2. JGOFS-Workshop. 18 .. /19. November 1993 in Bremen. 93 Seiten, Bremen, 1994. BaizeI', W. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 2211, Hamburg - Recife, 22.9. - 21.10.1992. 24 Seiten, Bremen, 1994. Stax, R. Zyklische Sedimentation von organischem Kohlenstoffin der Japan See: Anzeiger fUr Anderungen von Palaoozeanographie und Palaoklima im Spatkanozoikum. 150 Seiten, Bremen, 1994. Skowronek, F. Frlihdiagenetische Stoff-FlUsse gelOster Schwermetalle an der Oberflache von Sedimenten des Weser Astuares. 107 Seiten, Bremen, 1994. Dersch-Hansmann, M. Zur Klimaentwicklung in Ostasien wahrend der letzten 5 Millionen Jahre: Terrigener Sedimenteintrag in die Japan See (ODP Ausfahrt 128). 149 Seiten, Bremen, 1994. Zabel, M. Frlihdiagenetische Stoff-Fliisse in Oberflachen-Sedimenten des aquatorialen und ostlichen Slidatlantik. 129 Seiten, Bremen, 1994. Bleil, U. und Fahrtteilnehmer Report and preliminary results of SONNE-Cruise SO 86, Buenos Aires - Capetown, 22.4. - 31.5.93 116 Seiten, Bremen, 1994. Symposium: The South Atlantic: Present and Past Circulation. Bremen, Germany, 15 - 19 August 1994. Abstracts. 167 Seiten, Bremen, 1994. Kretzmann, U.B. 57Fe-Mossbauer-Spektroskopie an Sedimenten - Moglichkeiten und Grenzen. 183 Seiten, Bremen, 1994. Bachmann, M. Die Karbonatrampe von Organya im oberen Oberapt und unteren Unteralb (NE-Spanien, Prov. Lerida): Fazies, Zyklo- und Sequenzstratigraphie. 147 Seiten, Bremen, 1994. (vergriffen) Kemle-von Miicke, S. Oberflachenwasserstruktur und -zirkulation des Slidostatlantiks im Spatquartar. 151 Seiten, Bremen, 1994. Petermann, H. Magnetotaktische Bakterien und ihre Magnetosome in Oberflachensedimenten des Slidatlantiks. 134 Seiten, Bremen, 1994. Mulitza, S. Spatquartiire Variationen der oberflachennahen Hydrographie im westlichen aquatorialen Atlantik. 97 Seiten, Bremen, 1994. Segl, M. und Fahrteilnehmer Report and preliminary results of METEOR -Cruise M 2911, Buenos-Aires - Montevideo, 17.6. - 13.7.1994.94 Seiten, Bremen, 1994. B1eil, U. und Fahrtteilnehmer Report and preliminary results of METEOR -Cruise M 29/2, Montevideo - Rio de J aniero 15.7. - 8.8.1994.153 Seiten, Bremen, 1994. Henrich, R. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 29/3, Rio de Janeiro - Las Palmas 11.8. - 5.9.1994. Bremen, 1994. (vergriffen)

Nr.61

Nr.62

Nr.63

Nr.64

Nr.65

Nr.66

Nr.67

Nr.68

Nr.69

Nr.70

Nr.71

Nr.72

Nr.73

Nr.74

Nr.75

Nr.76

Nr.77

Nr.78

Nr.79

Sagemann, J. Saisonale Variationen von Porenwasserprofilen, Nahrstoff-Fllissen und Reaktionen in intertidalen Sedimenten des Weser-Astuars. 110 Seiten, Bremen, 1994. (vergriffen) Giese, M. und G. Wefer Bericht liber den 3. JGOFS-Workshop. 5./6. Dezember 1994 in Bremen. 84 Seiten, Bremen, 1995. Mann, U. Genese kretazischer Schwarzschiefer in Kolumbien: Globale vs. regionale/lokale Prozesse. 153 Seiten, Bremen, 1995. (vergriffen) WilIems, H., Wan X., Yin J., Dongdui L., Liu G., S. Diirr, K.-U. Grafe The Mesozoic development of the N-Indian passive margin and of the Xigaze Forearc Basin in southern Tibet, China. - Excursion Guide to IGCP 362 Working-Group Meeting "Integrated Stratigraphy". 113 Seiten, Bremen, 1995. (vergriffen) Hiinken, U. Liefergebiets - Charakterisierung proterozoischer Goldseifen in Ghana anhand von Fluideinschlu!3 - Untersuchungen. 270 Seiten, Bremen, 1995. Nyandwi, N. The Nature of the Sediment Distribution Patterns in ther Spiekeroog Backbarrier Area, the East Frisian Islands. 162 Seiten, Bremen, 1995. Isenbeck-Schroter, M. Transportverhalten von Schwermetallkationen und Oxoanionen in wassergesattigten Sanden. - Laborversuche in Saulen und ihre Modellierung -182 Seiten, Bremen, 1995. Hebbeln, D. und Fahrtteilnehmer Report and preliminary results of SONNE-Cruise SO 102, Valparaiso - Valparaiso, 95 l34 Seiten, Bremen, 1995. Willems, H. (Sprecher), U. Bathmann, U. Bleil, T. v. Dobeneck, K. Herterich, B.B. Jorgensen, E.-M. Nothig, M. Olesch, J. Patzold, H.D. Schulz, V. Smetacek, V. Spiefl, G. Wefer Bericht des Graduierten-Kollegs Stoff-Fllisse in marine Geosystemen. Berichtszeitraum Januar 1993 - Dezember 1995. 45 & 468 Seiten, Bremen, 1995. Giese, M. und G. Wefer Bericht liber den 4. JGOFS-Workshop. 20,/21. November 1995 in Bremen. 60 Seiten, Bremen, 1996. (vergriffen) Meggers, H. Pliozan-quartare Karbonatsedimentation und Palaozeanographie des Nordatlantiks und des Europaischen Nordmeeres - Hinweise aus planktischen Foraminiferengemeinschaften. 143 Seiten, Bremen, 1996. (vergriffen) Teske, A. Phylogenetische und okologische Untersuchungen an Bakterien des oxidativen und reduktiven marinen Schwefelkreislaufs mittels ribosomaler RNA.. 220 Seiten, Bremen, 1996. (vergriffen) Andersen, N. Biogeochemische Charakterisierung von Sinkstoffen und Sedimenten aus ostatlantischen Produktions-Systemen mit Hilfe von Biomarkern. 215 Seiten, Bremen, 1996. Treppke, U. Saisonalitat im Diatomeen- und Silikoflagellatenflul3 im ostlichen tropischen und subtropischen Atlantik. 200 Seiten, Bremen, 1996. Schiiring, J. Die Verwendung von Steinkohlebergematerialien im Deponiebau im Hinblick auf die Pyritvelwitterung und die Eignung als geochemische Barriere. 110 Seiten, Bremen, 1996. Patzold, J. und Fahrtteilnehmer Report and preliminary results of VICTOR HENSEN cruise JOPS II, Leg 6, Fortaleza - Recife, 10.3. - 26.3. 1995 and Leg 8, Vitoria - Vitoria, lOA. - 23.4.1995. 87 Seiten, Bremen, 1996. Bleil, U. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 3411, Cape Town - Walvis Bay, 3.-26.1.1996. 129 Seiten, Bremen, 1996. Schulz, H.D. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 3412, Walvis Bay - Walvis Bay, 29.1.-18.2.96 133 Seiten, Bremen, 1996. Wefer, G. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 34/3, Walvis Bay - Recife, 21.2.-l7.3.1996. 168 Seiten, Bremen, 1996.

Nr.80

Nr.81

Nr.82

Nr.83

Nr.84

Nr.85

Nr.86

Nr.87

Nr.88

Nr.89

Nr.90

Nr.91

Nr.92

Nr.93

Nr.94

Nr.95

Nr.96

Nr.97

Nr.98

Nr.99

Fischer, G. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 34/4, Recife - Bridgetown, 19.3.-15.4.1996. lO5 Seiten, Bremen, 1996. Kulbrok, F. Biostratigraphie, Fazies und Sequenzstratigraphie einer Karbonatrampe in den Schichten der Oberkreide und des Alttertiars Nordost-Agyptens (Eastern Desert, N'Golfvon Suez, Sinai). 153 Seiten, Brel11en, 1996. Kasten, S. Early Diagenetic Metal Enrichl11ents in Marine Sediments as Documents of Nonsteady-State Depositional Conditions. Bremen, 1996. Holmes, M.E. Reconstruction of Surface Ocean Nitrate Utilization in the Southeast Atlantic Ocean Based on Stable Nitrogen Isotopes. 113 Seiten, Brel11en, 1996. Riihlemann, C. Akkumulation von Carbonat und organischem Kohlenstoff im tropischen Atlantik: Spatquart1ire Produktivitats-Variationen und ihre Steuerungsmechanismen. 139 Seiten, Bremen, 1996. Ratmeyer, V. Untersuchungen zum Eintrag und Transport lithogener und organischer partikularer Substanz im ostlichen subtropischen Nordatlantik. 154 Seiten, Bremen, 1996. Cepek, M. Zeitliche und raumliche Variationen von Coccolithophoriden-Gemeinschaften im subtropischen Ost-Atlantik: Untersuchungen an Plankton, Sinkstoffen und Sedimenten. 156 Seiten, Brel11en, 1996. Otto, S. Die Bedeutung von gelOstem organischen Kohlenstoff (DOC) fUr den Kohlenstoffflu13 im Ozean. 150 Seiten, Bremen, 1996. Hensen, C. FrUhdiagenetische Prozesse und Quantifizierung benthischer Stoff-Fllisse in Oberflachensedimenten des Sudatlantiks. 132 Seiten, Brel11en, 1996. Giese, M. und G. Wefer Bericht uber den 5. JGOFS-Workshop. 27.128. November 1996 in Bremen. 73 Seiten, Bremen, 1997. Wefer, G. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 3711, Lisbon - Las Palmas, 4.-23.12.1996. 79 Seiten, Bremen, 1997. Isenbeck-Schroter, M., E. Bedbur, M. Kofod, B. Konig, T. Schramm & G. Matthefi Occurrence of Pesticide Residues in Water - Assessment of the Current Situation in Selected EU Countries. 65 Seiten, Bremen 1997. Kiihn, M. Geochemische Folgereaktionen bei der hydrogeothermalen Energiegewinnung. 129 Seiten, Bremen 1997. Determann, S. & K. Herterich JGOFS-A6 "Daten und Modelle": Smmnlung JGOFS-relevanter Modelle in Deutschland. 26 Seiten, Bremen, 1997. Fischer, G. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 3811, Las Palmas - Recife, 25.1.-l.3.l997, with Appendix: Core Descriptions from METEOR Cruise M 37/1. Bremen, 1997. Bleil, U. und Fahrtteilnehmer Report and preliminmy results of METEOR-Cruise M 3812, Recife - Las Palmas, 4.3.-14.4.1997. 126 Seiten, Brel11en, 1998. Neuer, S. und Fahrtteilnehmer Report and preliminary results of VICTOR HENSEN-Cruise 9611, Las Palmas - Las Palmas, 10.1. - 4-3.1996.76 Seiten, Bremen, 1997. ViIIinger, H. und Fahrtteilnehmer Fahrtbericht SO Ill, 20.8. - 16.9.1996. 115 Seiten, Bremen, 1997. Liining, S. Late Cretaceous - Early Tertiary sequence stratigraphy, paleoecology and geodynamics of Eastern Sinai, Egypt. 218 Seiten, Bremen, 1997. Haese, R.R. Beschreibung und Quantifizierung friihdiagenetischer Reaktionen des Eisens in Sedimenten des Sudatlantiks. 118 Seiten, Bremen, 1997.

Nr.l00

Nr.l0l

Nr.l02

Nr.l03

Nr.l04

Nr.l05

Nr.l06

Nr.l07

Nr.l08

Nr.l09

Nr.110

Nr.l11

Nr.l12

Nr.l13

Nr.114

Nr.115

Nr.116

Nr. 117

Nr.118

Nr.119

Liihrte, R. von Verwertung von Bremer Baggergut als Material zur Oberflachenabdichtung von Deponien -Geochemisches Langzeitverhalten und Schwermetall-Mobilitat (Cd, Cu, Ni, Pb, Zn). Bremen, 1997. Ebert, M. Der EinfluB des Redoxmilieus auf die Mobilitat von Chrom im durchstromten Aquifer. 135 Seiten, Bremen, 1997. Krogel, F. EinfluB von Viskositat und Dichte des Seewassers aufTransport und Ablagerung von Wattsedimenten (Langeooger RLickseitenwatt, sLidliche Nordsee). 168 Seiten, Bremen, 1997. Kerntopf, B. Dinoflagellate Distribution Pattems and Preservation in the Equatorial Atlantic and Offshore North-West Africa. 137 Seiten, Bremen, 1997. Breitzke, M. Elastische Wellenausbreitung in marinen Sedimenten - Neue Entwicklungen der Ultraschall Sedimentphysik und Sedimentechographie. 298 Seiten, Bremen, 1997. Marchant, M. Rezente und spatquartare Sedimentation planktischer Foraminiferen im Peru-Chile Strom. 115 Seiten, Bremen, 1997. Habicht, K.S. Sulfur isotope fractionation in marine sediments and bacterial cultures. 125 Seiten, Bremen, 1997. Hamer, K., R.v. Liihrte, G. Becker, T. Felis, S. Keffel, B. Strotmann, C. Waschkowitz, M. Kolling, M. lsenbeck-Schroter, H.D. Schulz Endbericht zum Forschungsvorhaben 060 des Landes Bremen: Baggergut del' Hafengruppe Bremen-Stadt: Modelluntersuchungen zur Schwermetallmobilitat und Moglichkeiten der Verwertung von Hafenschlick aus Bremischen Hafen. 98 Seiten, Bremen, 1997. Greeff,O.W. Entwicklung und Erprobung eines benthischen Landersystemes zur in situ-Bestimmung von Sulfatreduktionsraten mariner Sedimente. 121 Seiten, Bremen, 1997. Patzold, M. und G. Wefer Bericht libel' den 6. lGOFS-Workshop am 4./5.12.1997 in Bremen. Im Anhang: Publikationen zum deutschen Beitrag zur Joint Global Ocean Flux Study (JGOFS), Stand 1/1998. 122 Seiten, Bremen, 1998. Landenberger, H. CoTReM, ein Multi-Komponenten TranspOli- und Reaktions-Mode11. 142 Seiten, Bremen, 1998. VilIinger, H. und Fahrtteilnehmer Fahrtbericht SO 124, 4.10. - 16.10.1997. 90 Seiten, Bremen, 1997. Gietl, R. Biostratigraphie und Sedimentationsmuster einer nordostagyptischen Karbonatrampe unter Berlicksichtigung der A1veolinen-Faunen. 142 Seiten, Bremen, 1998. Ziebis, W. The Impact of the Thalassinidean Shrimp Callianassa trllncata on the Geochemistry of penneab1e, coastal Sediments. 158 Seiten, Bremen 1998. Schulz, H.D. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 4111, Malaga - Libreville, 13.2.-15.3.1998. Bremen, 1998. Volker, D.J. Untersuchungen an stromungsbeeinfluBten Sedimentationsmustem im SLidozean. Interpretation sedimentechographischer Daten und numerische Modellierung. 152 Seiten, Bremen, 1998. Schliinz, B. Riverine Organic Carbon Input into the Ocean in Relation to Late Quatemary Climate Change. 136 Seiten, Bremen, 1998. Kuhnert, H. Aufzeichnung des Klimas vor Westaustralien in stabilen Isotopen in Korallenskeletten. 109 Seiten, Bremen, 1998. Kirst, G. Rekonstruktion von Oberflachenwassertemperaturen im ostlichen Slidatlantik anhand von Alkenonen. 130 Seiten, Bremen, 1998. Diirkoop, A. Del' Brasil-Strom im Spatquartar: Rekonstruktion del' oberflachennahen Hydrographie wahrend del' Ietzten 400000 lahre. 121 Seiten, Bremen, 1998.

NI'. 120

NI'. 121

NI'. 122

NI'. 123

NI'. 124

NI'. 125

NI'. 126

NI'. 127

NI'. 128

NI'. 129

NI'. 130

NI'. 131

NI'. 132

NI'. 133

NI'. 134

NI'. 135

NI'. 136

NI'. 137

NI'. 138

NI'. 139

Lamy, F. Spatquartare Variationen des terrigenen Sedimenteintrags entlang des chilenischen Konti­nentalhangs als Abbild von Klimavariabilitat im Milankovil- und Sub-Milankovil-Zeitbereich. 141 Seiten, Bremen, 1998. Neuer, S. und Fahrtteilnehmer Report and preliminary results of POSEIDON -Cruise Pos 237/2, Vigo - Las Palmas, 18.3.-31.3.1998.39 Seiten, Bremen, 1998. Romero, O.E. Marine planktonic diatoms from the tropical and equatorial Atlantic: temporal flux patterns and the sediment record. 205 Seiten, Bremen, 1998. Spiess, V. und Fahrtteilnehmer Report and preliminary results of RV SONNE Cruise 125, Cochin - Chittagong, 17.1O.-17.1l.l997. 128 Seiten, Bremen, 1998. Arz, H.W. Dokllmentation von kurzfristigen Klimaschwankungen des Spatqllartars in Sedimenten des westlichen aquatorialen Atlantiks. 96 Seiten, Bremen, 1998. Wolff, T. Mixed layer characteristics in the equatorial Atlantic during the late QlIatemary as deduced from planktonic foraminifera. 132 Seiten, Bremen, 1998. Dittert, N. Late Quaternary Planktic Foraminifera Assemblages in the South Atlantic Ocean: Quantitative Determination and Preservational Aspects. 165 Seiten, Bremen, 1998. Him, c. Kalkige und organisch-wandige Dinoflagellaten-Zysten in Spatquartaren Sedimenten des tropischen Atlantiks und ihre palOkologische Auswertbarkeit. 121 Seiten, Bremen, 1998. Hencke, J. Redoxreaktionen im Grundwasser: Etablierung und Verlagerung von Reaktionsfronten und ihre Bedeutung fUr die Spurenelement-Mobilitat. 122 Seiten, Bremen 1998. Piitzold, J. und FahrtteiInehmer Report and preliminary results of METEOR-Cruise M 4113, Vitoria, Brasilien - Salvador de Bahia, Brasilien, 18.4. - 15.5.1998. Bremen, 1999. Fischer, G. und FahrtteiInehmer Report and preliminary results of METEOR-Cruise M 4114, Salvador de Bahia, Brasilien -Las Palmas, Spanien, 18.5. - 13.6.1998. Bremen, 1999. Schliinz, B. und G. Wefer Bericht liber den 7. JGOFS-Workshop am 3. und 4.12.1998 in Bremen. Im Anhang: Publikationen zum deutschen Beitrag zur Joint Global Ocean Flux Study (JGOFS), Stand 11 1999. 100 Seiten, Bremen, 1999. Wefer, G. und Fahrtteilnehmer Report and preliminary results of METEOR-Cruise M 42/4, Las Palmas - Las Palmas - Viena do Castelo; 26.09.1998 - 26.10.1998.104 Seiten, Bremen, 1999. Felis, T. Climate and ocean variability reconstructed from stable isotope records of modem SUbtropical Corals (Northern Red Sea). III Seiten, Bremen, 1999. Draschba , S. North Atlantic climate variability recorded in reef corals from Bermuda. 108 Seiten, Bremen, 1999. Schmieder, F. Magnetic Cyc10stratigraphy of South Atlantic Sediments. 82 Seiten, Bremen, 1999. Riefi, W. In situ measurements of respiration and mineralisation processes - Interaction between fauna and geochemical fluxes at active interfaces. 68 Seiten, Bremen, 1999. Devey, C.W. und Fahrtteilnehmer Report and shipboard results from METEOR-cruise M 4112, Libreville - Vitoria, 18.3. - 15.4.98. 59 Seiten, Bremen, 1999. WenzhOfer, F. Biogeochemical processes at the sediment water interface and quantification of metabolically driven calcite dissolution in deep sea sediments. 103 Seiten, Bremen, 1999. Klump, J. Biogenic barite as a proxy of pal eo productivity variations in the Southern Peru-Chile Current. 107 Seiten, Bremen, 1999.

Nr.140

Nr.141

Nr.142

Nr.143

Nr.144

Nr.145

Nr.146

Nr.147

Nr.148

Nr.149

Nr.150

Nr.151

Nr.152

Nr.153

Nr. 154

Nr.155

Nr.156

Nr.157

Nr.158

Nr.159

Nr.160

Huber, R. Carbonate sedimentation in the northern Northatlantic since the late pliocene. 103 Seiten, Bremen, 1999. Schulz, H. Nitrate-storing sulfur bacteria in sediments of coastal upwelling. 94 Seiten, Bremen, 1999. Mai, S. Die Sedimentverteilung im Wattenmeer: ein Simulationsmodell. 114 Seiten, Bremen, 1999. Neuer, S. und Fahrtteilnehmer Report and preliminary results of Poseidon Cruise 248, Las Palmas - Las Palmas, 15.2.-26.2.1999. 45 Seiten, Bremen, 1999. Weber, A. Schwefelkreislauf in marinen Sedimenten und Messung von in situ Sulfatreduktionsraten. 122 Seiten, Bremen, 1999. Hadeler, A. Sorptionsreaktionen im Grundwasser: Unterschiedliche Aspekte bei der Modellierung des Transportverhaltens von Zink. 122 Seiten, Bremen, 1999. Dierfien, H. Zum Kreislauf ausgewahlter Spurenmetalle im Siidatlantik: Vertikaltransport und Wechselwirkung zwischen Partikeln und L6sung. 167 Seiten, Bremen, 1999. Ziihlsdorff, L. High resolution multi-frequency seismic surveys at the Eastern Juan de Fuca Ridge Flank and the Cascadia Margin - Evidence for thenl1ally and tectonically driven fluid upflow in marine sediments. 118 Seiten, Bremen 1999. Kinkel, H. Living and late Quaternary Coccolithophores in the equatorial Atlantic Ocean: response of Distribution and productivity patterns to changing surface water circulation. 183 Seiten, Bremen 2000. Piitzold, J. und Fahrtteilnehmer Report and preliminary results of METEOR Cruise M 44/3, Aqaba (Jordan) - Safaga (Egypt) - DuM (Saudi Arabia) - Suez (Egypt) - Haifa (Israel), 12.3.-26.3.-2.4.-4.4.1999. 135 Seiten, Bremen, 2000. Schliinz, B. und G. Wefer Bericht iiber den 8. JGOFS-Workshop am 2. und 3.12.1999 in Bremen.Im Anhang: Publikationen zum deutschen Beitrag zur Joint Global Ocean Flux Study (JGOFS), Stand 1/2000. 95 Seiten, Bremen, 2000. Schnack, K. Biostratigraphie und fazielle Entwicklung in der Oberkreide und im Alttertiar im Bereich der Kharga Schwelle, Westliche Wi.iste, SW-Agypten. 142 Seiten, Bremen, 2000. Karwath, B. Ecological studies on living and fossil calcareous dinoflagellates of the equatorial and tropical Atlantic Ocean. 175 Seiten, Bremen, 2000. Moustafa, Y. Paleoclimatic reconstructions of the Northern Red Sea during the Holocene infelTed from stable isotope records of modem and fossil corals and molluscs. 102 Seiten, Bremen, 2000. Villinger, H. and cruise participants Report and prelminary results of SONNE-cruise 145-1 Balboa - Talcahuana, 21.12.1999 - 28.01.2000. 147 Seiten, Bremen, 2000. Rusch, A. Dynamik der Feinfraktion im Oberflachenhorizont permeabler Schelfsedimente. 102 Seiten, Bremen, 2000. Moos, C. Reconstruction of upwelling intensity and paleo-nutrient gradients in the northwest Arabian Sea derived from stable carbon and oxygen isotopes of plank tic foraminifera. 103 Seiten, Bremen, 2000. XU,W. Mass physical sediment properties and trends in a Wadden Sea tidal basin. 127 Seiten, Bremen, 2000. Meinecke, G. and cruise participants Report and preliminary results of METEOR Cruise M 4511, Malaga (Spain) - Lissabon (Portugal), 19.05. - 08.06.1999.39 Seiten, Bremen, 2000. Vink, A. Reconstruction of recent and late Quaternary surface water masses of the western subtropical Atlantic Ocean based on calcareous and organic-walled dinoflagellate cysts. 160 Seiten, Bremen, 2000. Willems, H. (Sprecher), U. Bleil, R. Henrich, K. Herterich, B.B. Jorgensen, H.-J. Kufi, M. Olesch, H.D. Schulz,V. Spiefi, G. Wefer Abschlu13bericht des Graduierten-Kollegs Stoff-Fliisse in marine Geosystemen. Zusammenfassung und Berichtszeitraum Januar 1996 - Dezember 2000. 340 Seiten, Bremen, 2000.

Nr.161

Nr.162

Nr.163

Nr.164

Nr.165

Nr. 166

Nr.167

Nr.168

Nr.169

Nr.170

Nr.l71

Nr.l72

Nr. 173

Nr.174

Nr.175

Nr.176

Nr.l77

Nr.178

Nr.179

Nr.180

Sprengel, C. Untersuchungen zur Sedimentation und Okologie von Coccolithophoriden im Bereich der Kanarischen Inseln: Saisonale Flussmuster und Karbonatexport. 165 Seiten, Bremen, 2000. Donner, B. und G. Wefer Bericht Liber den JGOFS-Workshop am 18.-21.9.2000 in Bremen: Biogeochemical Cycles: German Contributions to the International Joint Global Ocean Flux Study. 87 Seiten, Bremen, 2000. Neuer, S. and cruise participants Report and preliminiary results of Meteor Cruise M 45/5, Bremen - Las Palmas, October 1 - November 3, 1999.93 Seiten, Bremen, 2000. Devey, C. and cruise participants Report and preliminiary results of Sonne Cruise SO 145/2, Talcahuano (Chile) - Arica (Chile), February 4 - February 29, 2000. 63 Seiten, Bremen, 2000. Freudenthal, T. Reconstruction of productivity gradients in the Canary Islands region off Morocco by means of sinking particles and sediments. 147 Seiten, Bremen, 2000. Adler, M. Modeling of one-dimensional transport in porous media with respect to simultaneous geochemical reactions in CoTReM. 147 Seiten, Bremen, 2000. Santamarina Cuneo, P. Fluxes of suspended particulate matter through a tidal inlet of the East Frisian Wadden Sea (southern N0l1h Sea). 91 Seiten, Bremen, 2000. Benthien, A. Effects of CO2 and nutrient concentration on the stable carbon isotope composition of C37:2 alkenones in sediments of the South Atlantic Ocean. 104 Seiten, Bremen, 200 I. Lavik, G. Nitrogen isotopes of sinking matter and sediments in the South Atlantic. 140 Seiten, Bremen, 2001. Budziak, D. Late Quaternary monsoonal climate and related variations in paleoproductivity and alkenone-derived sea-surface temperatures in the western Arabian Sea. 114 Seiten, Bremen, 2001. Gerhardt, S. Late Quaternary water mass variability derived from the pteropod preservation state in sediments of the western South Atlantic Ocean and the Caribbean Sea. 109 Seiten, Bremen, 2001. Bleil, U. und Fah.·tteilnehmer Report and preliminary results of Meteor Cruise M 46/3, Montevideo (Uruguay) - Mar del Plata (Argentina), January 4 - February 7, 2000. Bremen, 2001. Wefer, G. und Fahrtteilnehmer Report and preliminary results of Meteor Cruise M 46/4, Mar del Plata (Argentina) - Salvador da Bahia (Brazil), February 10 - March 13,2000. With partial results of METEOR cruise M 46/2. 136 Seiten, Bremen, 2001. Schulz, H.D. und Fahrtteilnehmer Report and preliminary results of Meteor Cruise M 46/2, Recife (Brazil) - Montevideo (Uruguay), December 2 - December 29,1999. 107 Seiten, Bremen, 2001. Schmidt, A. Magnetic mineral fluxes in the Quatemary South Atlantic: Implications for the paleoenvironment. 97 Seiten, Bremen, 2001. Bruhns, P. Crystal chemical characterization of heavy metal incorporation in brick burning processes. 93 Seiten, Bremen, 2001. Karius, V. Baggergut der Hafengruppe Bremen-Stadt in del' Ziegelherstellung. 131 Seiten, Bremen, 2001. Adegbie, A. T. Reconstruction of paleoenvironmental conditions in Equatorial Atlantic and the Gulf of Guinea Basins for the last 245,000 years. 113 Seiten, Bremen, 2001. Spie6, V. und Fahrtteilnehmer Report and preliminary results ofRlV Sonne Cruise SO 149, Victoria - Victoria, 16.8. - 16.9.2000. 100 Seiten, Bremen, 2001. Kim, J.-H. Reconstruction of past sea-surface temperatures in the eastem South Atlantic and the eastem South Pacific across Termination I based on the Alkenone Method. 114 Seiten, Bremen, 2001.

Nr.181

Nr.182

Nr.183

Nr.184

Nr.185

Nr.186

Nr.187

Nr.188

Nr.189

Nr.190

Nr.191

Nr.192

Nr.193

Nr.194

Nr.195

Nr.196

Nr.197

Nr.198

Nr.199

Nr. 200

von Lom-Keil, H. Sedimentary waves on the Namibian continental margin and in the Argentine Basin - Bottom flow reconstructions based on high resolution echo sounder data. 126 Seiten, Bremen, 200l. Hebbeln, D. und Fahrtteilnehmer PUCK: Report and preliminary results of R/V Sonne Cruise SO 156, Valparaiso (Chile) - Talcahuano (Chile), March 29 - May 14, 200 l. 195 Seiten, Bremen, 200 l. Wendler, J. Reconstruction of astronomically-forced cyclic and abrupt paleoecological changes in the Upper Cretaceous Boreal Realm based on calcareous dinoflagellate cysts. 149 Seiten, Bremen, 2001. Volbers, A. Planktic foraminifera as paleoceanographic indicators: production, preservation, and reconstruction of upwelling intensity. Implications from late Quaternary South Atlantic sediments. 122 Seiten, Bremen, 200l. Bleil, U. and cruise participants Report and preliminary results of R/V METEOR Cruise M 49/3, Montevideo (Uruguay) - Salvador (Brasil), March 9 - April 1, 200l. 99 Seiten, Bremen, 200l. Scheibner, C. Architecture of a carbonate platform-to-basin transition on a structural high (Campanian-early Eocene, Eastern Desert, Egypt) - classical and modelling approaches combined. 173 Seiten, Bremen, 2001. Schneider, S. QuarHire Schwankungen in Stromungsintensitat und Produktivitat als Abbild der Wassermassen­Variabilitat im aquatorialen Atlantik (ODP Sites 959 und 663): Ergebnisse aus Siltkorn-Analysen. 134 Seiten, Bremen, 2001. Uliana, E. Late Quaternary biogenic opal sedimentation in diatom assemblages in Kongo Fan sediments. 96 Seiten, Bremen, 2002. Esper, O. Reconstruction of Recent and Late Quaternary oceanographic conditions in the eastern South Atlantic Ocean based on calcareous- and organic-walled dinoflagellate cysts. 130 Seiten, Bremen, 2001. Wendler, I. Production and preservation of calcareous dinoflagellate cysts in the modern Arabian Sea. 117 Seiten, Bremen, 2002. Bauer, J. Late Cenomanian - Santonian carbonate platfonn evolution of Sinai (Egypt): stratigraphy, facies, and sequence architecture. 178 Seiten, Bremen, 2002. Hildebrand-Habel, T. Die Entwicklung kalkiger Dinoflagellaten im SUdatlantik seit der hOheren Oberkreide. 152 Seiten, Bremen, 2002. Hecht, H. Sauerstoff-Optopoden zur Quantifizierung von Pyritverwitterungsprozessen im Labor- und Langzeit-in-situ­Einsatz. Entwicklung - Anwendung - Modellierung. 130 Seiten, Bremen, 2002. Fischer, G. und Fahrtteilnehmer Report and Preliminary Results of RV METEOR Cruise M49/4, Salvador da Bahia - Halifax, 4.4.-5.5.200l. 84 Seiten, Bremen, 2002. Groger, M. Deep-water circulation in the western equatorial Atlantic: inferences from carbonate preservation studies and silt grain-size analysis. 95 Seiten, Bremen, 2002. Meinecke,G. und Fahrtteilnehmer Report of RV POSEIDON Cruise POS 271, Las Palmas - Las Palmas, 19.3.-29.3.200l. 19 Seiten, Bremen, 2002. Meggers, H und Fahrtteilnehmer Report of RV POSEIDON Cruise POS 272, Las Palmas - Las Palmas, 1.4.-14.4.200l. 19 Seiten, Bremen, 2002. Grafe, K.-U. Stratigraphische Korrelation und Steuerungsfaktoren Sedimentarer Zyklen in ausgewahlten Borealen und Tethyalen Becken des Cenoman/Turon (Oberkreide) Europas und Nordwestafrikas. 197 Seiten, Bremen, 2002. Jahn, B. Mid to Late Pleistocene Variations of Marine Productivity in and Terrigenous Input to the Southeast Atlantic. 97 Seiten, Bremen, 2002. AI-Rous an, S. Ocean and climate history recorded in stable isotopes of coral and foraminifers from the northern Gulf of Aqaba. 116 Seiten, Bremen, 2002.

NI'. 201

NI". 202

NI'. 203

NI'. 204

NI'. 205

NI'. 206

NI'. 207

NI'. 208

NI'. 209

NI'. 210

NI'. 211

NI'. 212

NI'. 213

NI'. 214

NI'. 215

NI'. 216

NI'. 217

NI'. 218

NI'. 219

NI'. 220

NI'. 221

Azouzi, B. Regionalisierung hydraulischer und hydrogeochemischer Daten mit geostatistischen Methoden. 108 Seiten, Bremen, 2002. Spiefi, V. und Fahrtteilnehmer Report and preliminary results of METEOR Cruise M 47/3, Libreville (Gabun) - Walvis Bay (Namibia), 01.06 - 03.07.2000.70 Seiten, Bremen 2002. Spiefi, V. und Fahrtteilnehmer Report and preliminary results of METEOR Cruise M 49/2, Montevideo (Uruguay) - Montevideo, 13.02 - 07.03.2001. 84 Seiten, Bremen 2002. Mollenhauer, G. Organic carbon accumulation in the South Atlantic Ocean: Sedimentary processes and glacial/interglacial Budgets. 139 Seiten, Bremen 2002. Spiefi, V. und Fahrtteilnehmer Report and preliminary results of METEOR Cruise M4911, Cape Town (South Africa) - Montevideo (Uruguay), 04.01.2000 - 10.02.2000.57 Seiten, Bremen, 2003. Meier, KJ.S. Calcareous dinoflagellates from the Mediterranean Sea: taxonomy, ecology and palaeoenvironmental application. 126 Seiten, Bremen, 2003. Rakic, S. Untersuchungen zur Polymorphie und Kristallchemie von Silikaten der Zusammensetzung Me2Si20S (Me:Na, K). 139 Seiten, Bremen, 2003.

Pfeifer, K Auswirkungen friihdiagenetischer Prozesse auf Calcit- und Barytgehalte in marinen Oberflachen­sedimenten. 110 Seiten, Bremen, 2003. Heuer, V. Spurenelemente in Sedimenten des Siidatlantik. Primarer Eintrag und friihdiagenetische Uberpragung. 136 Seiten, Bremen, 2003. Streng, M. Phylogenetic Aspects and Taxonomy of Calcareous Dinoflagellates. 157 Seiten, Bremen 2003. Boeckel, B. Present and past coccolith assemblages in the South Atlantic: implications for species ecology, carbonate contribution and palaeoceanographic applicability. 157 Seiten, Bremen, 2003. Precht, E. Advective interfacial exchange in permeable sediments driven by surface gravity waves and its ecological consequences. 131 Seiten, Bremen, 2003. Frenz, M. Grain-size composition of Quaternary South Atlantic sediments and its paleoceanographic significance. 123 Seiten, Bremen, 2003. Meggers, H. und Fahrtteilnehmer Report and preliminary results of METEOR Cruise M 5311, Limassol - Las Palmas - Mindelo, 30.03.2002 - 03.05.2002.81 Seiten, Bremen, 2003. Schulz, H.D. und Fahrtteilnehmer Report and preliminary results of METEOR Cruise M 5811, Dakar - Las Palmas, 15.04 .. 2003 - 12.05.2003. Bremen,2003. Schneider, R. und Fahrtteilnehmer Report and preliminary results of METEOR Cruise M 57/1, Cape Town - Walvis Bay, 20.01. - 08.02.2003. 123 Seiten, Bremen, 2003. Kallmeyer, J. Sulfate reduction in the deep Biosphere. 157 Seiten, Bremen, 2003. Roy,H. Dynamic Structure and Function of the Diffusive Boundary Layer at the Seafloor. 149 Seiten, Bremen,2003. Plitzold, J., C. Hiibscher und Fahrtteilnehmer Report and preliminary results of METEOR Cruise M 52/2&3, Istanbul- Limassol - Limassol, 04.02. - 27.03.2002, Bremen, 2003.

Zabel, M. and cruise participants Report and preliminary results of METEOR Cruise M 57/2, Walvis Bay - Walvis Bay, 11.02. - 12.03.2003, 136 Seiten, Bremen 2003. Salem, M. Geophysical investigations of submarine prolongations of alluvial fans on the western side of the Gulf of Aqaba-Red Sea. 100 Seiten, Bremen, 2003.

Nr. 222

Nr.223

Nr.224

Nr.225

Nr.226

Tilch, E. Oszillation von Wattflachen und deren fossiles Erhaltungspotential (Spiekerooger Riickseitenwatt, siidliche Nordsee). 137 Seiten, Bremen, 2003. Frisch, U. nnd F. Kockel Der Bremen-Knoten im Strukturnetz Nordwest-Deutschlands. Stratigraphie, Palaogeographie, Strukturgeologie. 379 Seiten, Bremen, 2004. Kolonic, S. Mechanisms and biogeochemical implications of Cenomanian/Turonian black shale formation in North Africa: An integrated geochemical, millennial-scale study from the Tarfaya-LaAyoune Basin in SW Morocco. 174 Seiten, Bremen, 2004. Panteleit, B. Geochemische Prozesse in der Salz- Siil3wasser Obergangszone. 106 Seiten, Bremen, 2004. Seiter, K. Regionalisierung und Quantifizierung benthischer Mineralisationsprozesse. 135 Seiten, Bremen, 2004.