untersuchungen zur speziation von verschmutzungsrelevanten ... · ost-hainan generell höher als in...
Post on 09-Sep-2019
5 Views
Preview:
TRANSCRIPT
Untersuchungen zur Speziation von verschmutzungsrelevanten Spurenmetallen in
tropischen Ästuar- und Küstensystem (Südchina und Nord-Ost-Brasilien)
Dissertation zum Erlangen des akademischen Grades eines
„Doktor der Naturwissenschaften“
(Dr. rer. nat.)
des Fachbereichs 02, Biologie/Chemie
der Universität Bremen
Vorgelegt von:
Jun Fu
Bremen, 2014
I
Untersuchungen zur Speziation von verschmutzungsrelevanten Spuren-metallen in tropischen Ästuar- und Küstensystem (Südchina und Nord-Ost-Brasilien).
Dissertation zum Erlangen des akademischen Grades eines „Doktor der
Naturwissenschaften (Dr. rer. nat.)“ des Fachbereichs 02, Biologie/Chemie der
Universität Bremen.
Vorgelegt von: Jun Fu
Gutachter: 1. Prof. Dr. Wolfgang Balzer
2. Prof. Dr. Wolfram Thiemann
Prüfer: 1. Prof. Dr. Otto Schrems
2. Dr. Uwe Schüßler
Datum des Kolloquiums: 04.02.2014
II
Selbstständigkeitserklärung:
Hiermit erkläre ich, dass ich die Doktorarbeit mit dem Titel:
„Untersuchungen zur Speziation von verschmutzungsrelevanten Spuren-metallen in
tropischen Ästuar- und Küstensystem (Südchina und Nord-Ost-Brasilien).“
Selbstständig verfasst und geschrieben habe und außer den angegebenen Quellen
keine weiteren Hilfsmittel verwendet habe.
Jun Fu
III
Danksagung
Mein besonderer Dank gilt Herrn Prof. Dr. Wolfgang Balzer von der Universität
Bremen für die Ermöglichung dieser Arbeit sowie seine wissenschaftlichen
Anregungen und Diskussionen als Doktorvater während dieser Arbeit. Ich möchte
mich bei meinem Zweitgutachter Herrn Prof. Dr. Wolfram Thiemann von der
Universität bedanken. Herrn Prof. Dr. Otto Schrems vom Alfred-Wegener-Institut für
Polar- und Meeresforschung (AWI) und Herrn Dr. Uwe Schüßler danke ich für ihre
Funktion als Prüfer.
Für die gute Zusammenarbeit und die hilfreichen Erörterungen organisatorischer
sowie analytischer und inhaltlicher Problemstellungen möchte ich mich bei meinen
Kollegen und ehemaligen Kollegen Dr. U. Schüßler, Dr. W. Barkmann, Dr. M.
Lukman, Dr. B. Bach, U. Wolpmann, I. Becker, O. Wilhelm, T. Daberkow und X.-L.
Tang bedanken. Insbesondere danke ich Herrn u.a. X.-L. Tang für die Durchführung
von Analysen eines Teils der verwendeten Proben.
Daneben wurden zu Vergleichs- und Interpretationszwecken einige Ergebnisse
(Balzer, unpublished) von Expeditionen für die beiden Projekte in die vorliegend
Arbeit mit aufgenommen, die vor Beginn dieser Dissertation durchgeführt worden
waren; dies gilt auch für die Verwendung von DOC-Daten, die sämtlich im Rahmen
eines anderen Vorhabens erhoben wurden.
Diese Arbeit wurde im Rahmen des Projekts "LANCET (Land-Sea Interactions in
Coastal Ecosystems of Tropical China: Hainan)" bzw. des Projekts „POLCAMAR
(Impact of Pollutants from Sugar Cane Monoculture on Estuaries and Coastal Waters
of NE-Brazil)“ durchgeführt. Gedankt sei den Kommissionen der jeweiligen Projekte
für die finanzielle Unterstützung.
Ich möchte mich auch bei den vielen Wissenschaftlerinnen und Wissenschaftlern, die
an den beiden wissenschaftlichen Projekten „LANCET“ und „POLCAMAR“ beteiligt
waren, für die Kollegialität bedanken und ganz besonders bei den Koordinatoren Dr.
L. Dsikowitzky und Dr. D.-R. Wang für „LANCET“ sowie O. Wilhelm und Prof. Dr. B.
Knoppers für „POLCAMAR“, die mir in China bzw. in Brasilien stets hilfreich zur Seite
standen.
Schließlich danke ich meiner Familie und meinen Freunden für ihre Ermutigung und
Unterstützung während der Arbeit.
IV
Kurzfassung
Diese Arbeit befasste sich im Wesentlichen mit der Konzentration, der Verteilung und
dem Verhalten von fünf verschmutzungsrelevanten Spurenmetallen (Nickel, Cobalt,
Kupfer, Cadmium und Blei) plus Eisen in der wässrigen und in der partikulären
Phase sowie der Abschätzung des Kontaminationsniveaus der untersuchten Metalle
in tropischen Ästuaren und Lagunen von China und Brasilien.
Die Feldarbeiten fanden im Wanquan Ästuar und im Wenchang-Wenjiao Ästuar auf
der Südchinesischen Insel Hainan sowie in der Mundaú Lagune und in der
Manguaba Lagune in Nordostbrasilien statt, die alle durch relativ niedrige DOC- und
Huminstoff-Konzentrationen gekennzeichnet sind. Die Oberflächenwasserproben
wurden durch Filtration in eine gelöste (< 0.45 μm) und eine partikuläre (> 0.45 μm)
Phase separiert. Danach erfolgte eine weitere Separation der gelösten Phase mittels
Querstrom-Ultrafiltration in zumeist zwei kolloidale Fraktionen (HMW "high molecular
weight“: 10 kDa - 0.45 μm; LMW "low molecular weight“: 5 kDa - 10 kDa) sowie eine
als "echt gelöst" angesehene Fraktion (TD "truely dissolved": < 5 kDa).
Die Konzentrationen der sechs verschmutzungsrelevanten Spurenmetalle liegen in
Ost-Hainan generell höher als in Nordost-Brasilien. Das ästuarine Verhalten der
Metalle in den beiden Studiengebieten ist jedoch kaum voneinander unterscheidbar.
Das gelöste Cadmium scheint mehr oder weniger in der gelösten Phase angereichert
zu werden, wohingegen gelöstes Eisen und Blei mit zunehmendem Salzgehalt rasch
aus der gelösten Phase entfernt werden. Nickel, Cobalt und Kupfer in Lösung sowie
alle Spurenmetalle in partikulärer Form folgen allgemein dem Muster konservativer
Mischung. In Hinblick auf die kolloidalen Fraktionen gewinnen die echt gelösten
Metalle (< 5 kDa; TD) mit steigender Salinität zunehmend an Bedeutung. Gelöstes
Eisen, Nickel und Blei sind hauptsächlich in der hochmolekularen Fraktion (10 kDa -
0.45 μm; HMW) vorhanden, wohingegen ein dominanter Anteil des gelösten
Cadmium in der echt gelösten Fraktion vorliegt. Beim gelösten Cobalt und Kupfer
sind zwar jeweils die echt gelösten Fraktionen ebenfalls vorherrschend; jedoch
können die hochmolekulare und die Fraktion mit niedrigem Molekulargewicht (LMW:
5 kDa - 10 kDa) nicht vernachlässigt werden.
Trotz der intensiven Aquakulturindustrie in Ost-Hainan und des großflächigen
Anbaus von Zuckerrohr in Monokultur in Nordost-Brasilien sowie der anthropogenen
Aktivitäten in den beiden Studiengebieten konnten keine nennenswerten
Umweltbeeinträchtigungen in Form von Spurenmetallverschmutzungen festgestellt
werden. Wenn erhöhte Metallkonzentrationen auftraten (im Wenchang-Wenjiao
V
Ästuar und in der Mundaú Lagune), wurden sie im Ästuar effektiv gedämpft. Die
Metallkonzentrationen bei hohen Salinitäten in Ost-Hainan und Nordost-Brasilien
liegen etwa auf dem gleichen Niveau. Im weltweiten Vergleich liegen sowohl die
gelösten als auch die partikulären Metallenkonzentrationen in den beiden
Studiengebieten auf einem niedrigen bis mittleren Niveau, mit Ausnahme von Eisen,
was vermutlich auf die geologische Gegebenheiten zurückzuführen ist.
VI
Abstract
This work investigated the concentrations, the distributions and the behaviors of five
trace metals (nickel, cobalt, copper, cadmium and lead) plus iron in the aqueous and
in the particulate phase in two tropical estuarine systems of China and Brazil, and
estimated their contamination levels. The field work was carried out in the Wanquan
Estuary and in the Wenchang-Wenjiao Estuary of the Southern Chinese island of
Hainan as well as in the Mundaú Lagoon and in the Manguaba Lagoon of Northeast-
Brazil, which are all characterised by relatively low contents of dissolved humics and
DOC. The surface water samples were seperated into the dissolved (< 0.45 μm) and
the particulate (> 0.45 μm) phase by conventional filtration followed by a further
separation of the dissolved phase into two colloidal fractions (HMW "high molecular
weight“: 10 kDa - 0.45 μm; LMW "low molecular weight“: 5 kDa - 10 kDa) and a
fraction (TD: < 5 kDa) considered as "truly dissolved" fraction.
The levels of these trace metals with high contamination riscs is generally higher in
East-Hainan than in Northeast-Brazil. However, the estuarine behaviors of these
metals in the two study areas are very similar. Dissolved cadmium appears more or
less to be enriched in the dissolved phase, while dissolved iron and lead are rapidly
removed from the dissolved phase when salinity increases. Nickel, cobalt and copper
in solution, as well as all trace metals in particulate forms essentially follow a
conservative mixing line. The relative significance of the truly dissolved fraction (< 5
kDa; TD) of the metals increased with increasing salinity. Dissolved iron, nickel and
lead are mainly associated with the high molecular weight fraction (10 kDa - 0.45 μm;
HMW), whereas dissolved cadmium predominantly occurs in the truly dissolved
fraction. The truly dissolved fraction is also dominant for dissolved cobalt and copper;
however, the high molecular weight fraction and the low molecular weight fraction (5
kDa - 10 kDa; LMW) cannot be neglected for these two metals.
Despite the massive aquaculture industry in East-Hainan and the extensive
cultivation of sugar cane in monocultural form in Northeast-Brazil as well as general
anthropogenic activities, no significant impact of trace metals on the local
environment could be detected in the two study areas. Although elevated metal
concentrations occurred (in the Wenchang-Wenjiao Estuary and in the Mundaú
Lagoon), they were leveled out effectively during estuarine processing. At high
salinities, the metal concentrations in East-Hainan and in Northeast-Brazil are
approximately at the same level. In comparison with averages of estuaries worldwide,
both dissolved and particulate metal concentrations in the two study areas are similar
VII
or lower, except for iron probably originating from specific geographic settings.
VIII
Inhaltsverzeichnis Abbildungsverzeichnis 1 Tabellenverzeichnis 5 1. Einführung 6
1.1. Verschmutzungsrelevante Spurenmetalle in aquatischen Systemen 8
1.1.1. Eisen 8
1.1.2. Nickel 8
1.1.3. Cobalt 9
1.1.4. Kupfer 9
1.1.5. Cadmium 10
1.1.6. Blei 10
1.2. Ästuar-Typen 11
1.3. Mastervariablen für Spurenmetalle 12
1.3.1. Salinität und chemische Zusammensetzung 12
1.3.2. Residenzzeit 13
1.3.3. Suspendiertes partikuläres Material 13
1.3.4. Größenspeziation (Kolloide) 14
1.3.5. Redoxpotential und pH-Wert 15
1.3.6. Biologische Aktivität 16
1.4. Ästuarine Mischungsprozesse für Spurenmetalle 17
2. Material und Methoden 18
2.1. Studiengebiete 18
2.1.1. Ästuare auf der südchinesischen Insel Hainan 18
2.1.2. Lagunen-System in Nordost-Brasilien 19
2.2. Maßnahmen zur Kontaminationsminderung 20
2.3. Probenahme 21
2.4. Filtration und Querstrom-Ultrafiltration 21
2.5. UV-Aufschluss 27
2.6. Flüssig-Flüssig-Extraktion 27
2.7. Mikrowellen-Aufschluss-Verfahren 28
2.8. Massenspektrometrie mit HR-ICP-MS 30
2.9. Referenzmaterialien 30
2.10. Andere Parameter 32
IX
3. Vergleichende Diskussion der Ergebnisse 33 3.1. Eisen 34
3.2. Nickel 37
3.3. Cobalt 39
3.4. Kupfer 42
3.5. Cadmium 44
3.6. Blei 47
3.7. Verteilungskoeffizienten 49
3.8. Vergleich der Studiengebiete und ihrer Kontaminationsniveaus 52
3.9. Offene Fragen und Perspektiven 57
4. Publikation 1 (erschienen) : "Estuarine modification of dissolved
and particulate trace metals in major rivers of East-Hainan" Continental Shelf Research 57 (2013): 59-72 59
Abstract 59
4.1. Introduction 61
4.2. Materials and Methods 62
4.2.1. Study area 62
4.2.2. Sampling 64
4.2.3. Pretreatment and measurement of dissolved metals 64
4.2.4. Pretreatment and analysis of particulate metals 65
4.2.5. Other parameters 65
4.3. Results and discussion 66
4.3.1. Distribution of suspended particulate matter and environmental
Conditions 66
4.3.2. Dissolved trace metals 69
4.3.3. Particulate trace metals 75
4.3.4. Distribution coefficients 79
4.3.5. Comparison of WR and WWR with other Chinese estuaries 82
4.4. Conclusions 84
Acknowledgements 85
References 86
5. Entwurf zur Publikation 2: "Transport of trace metals in colloidal
form through estuarine systems of East-Hainan" 91 Abstract 91
5.1. Introduction 92
X
5.2. Materials and Methods 93
5.2.1. Study area 93
5.2.2. Sampling and filtration 95
5.2.3. Pretreatment and measurement 96
5.2.4. CFF blanks and mass balance 97
5.3. Results and discussion 98
5.3.1. Iron 98
5.3.2. Lead 101
5.3.3. Cadmium 101
5.3.4. Copper 102
5.3.5. Cobalt 103
5.3.6. Nickel 104
5.3.7. Spatial variability of colloidal trace metals in East-Hainan 104
5.4. Conclusions 105
6. Entwurf zur Publikation 3: "Cd, Co, Cu, Fe, Ni and Pb in solution,
colloidal and particulate phases of Northeast Brazilian estuaries" 107 Abstract 107
6.1. Introduction 108
6.2. Materials and Methods 110
6.2.1. Study area 110
6.2.2. Sampling and filtration 111
6.2.3. Pretreatment and measurement of dissolved and colloidal metals 112
6.2.4. Cross-flow filtration Blanks and Mass balance 112
6.2.5. Pretreatment and measurement of particulate metals 114
6.2.6. Other parameters 115
6.3. Results and Discussion 115
6.3.1. Salinity, suspended particulate matter and dissolved organic carbon 115
6.3.2. Dissolved trace metals 117
6.3.3. Particulate trace metals 130
6.3.4. Distribution coefficient 131
6.3.5. Colloidal trace metals 133
6.3.6. Spatial and seasonal differences 137
6.4. Conclusions 138
Literaturverzeichnis 141 Anhang 155
Abbildungsverzeichnis
- 1 -
Abbildungsverzeichnis
Kapitel 1-3
Abb. 1. Korngrößenverteilung von verschiedenen Kolloiden und Partikeln
in der Umwelt. 15
Abb. 2. Die Studiengebiete: (a) Wenchang-Wenjiao Ästuar und (b)
Wanquan Ästuar in Ost-Hainan. 18
Abb. 3. Das Studiengebiet Mundaú-Manguaba Lagunen-System in
Nordost-Brasilien. 19
Abb. 4. Schematische Darstellung der Querstrom-Ultrafiltrations-Kassette
Vivaflow® von Sartorius. 25
Abb. 5. Ablauf der Querstrom-Ultrafiltrations-Prozedur. 25
Abb. 6. Salinitäts-Abhängigkeit des gelösten Eisens. 35
Abb. 7. Salinitäts-Abhängigkeit des partikulären Eisens. 36
Abb. 8. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Eisen. 36
Abb. 9. Salinitäts-Abhängigkeit des gelösten Nickels. 37
Abb. 10. Salinitäts-Abhängigkeit des partikulären Nickels. 38
Abb. 11. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Nickel. 38
Abb. 12. Salinitäts-Abhängigkeit des gelösten Cobalts. 40
Abb. 13. Salinitäts-Abhängigkeit des partikulären Cobalts. 40
Abb. 14. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Cobalt. 41
Abb. 15. Salinitäts-Abhängigkeit des gelösten Kupfers. 42
Abb. 16. Salinitäts-Abhängigkeit des partikulären Kupfers. 43
Abb. 17. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Kupfer. 43
Abb. 18. Salinitäts-Abhängigkeit des gelösten Cadmiums. 45
Abb. 19. Salinitäts-Abhängigkeit des partikulären Cadmiums. 45
Abb. 20. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Cadmium. 46
Abb. 21. Salinitäts-Abhängigkeit des gelösten Bleis. 47
Abb. 22. Salinitäts-Abhängigkeit des partikulären Bleis. 48
Abbildungsverzeichnis
- 2 -
Abb. 23. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Blei. 48
Abb. 24. Salinitäts-Abhängigkeit des log KD. 50
Abb. 25. Gelöstes Cadmium in der Weser an der FGG-Weser-Messstelle
Bremerhaven. 57
Kapitel 4
Fig. 1. Estuaries of East Hainan under investigation: (a)
Wenchang/Wenjiao River estuary and (b) Wanquan River estuary. 62
Fig. 2. Distribution of SPM and DOC vs. salinity. 66
Fig. 3. Dissolved trace metals vs. salinity. 68
Fig. 4. Dissolved Pb vs. dissolved Fe. 72
Fig. 5. Dissolved Co vs. dissolved Fe during the wet seasons. 72
Fig. 6. Particulate trace metals vs. salinity. 74
Fig. 7. Ratio of the percentages of the particulate metal on the sum of the
dissolved and particulate metal: Pb vs. Fe. 78
Fig. 8. Log10KD vs. salinity. 80
Kapitel 5
Fig. 1. Cross-flow filtration apparatus. 93
Fig. 2. Wenchang/Wenjiao River estuary (a) and Wanquan River estuary
(b). 94
Fig. 3. Flow diagram of the sample filtration procedure. 96
Fig. 4. Colloidal and truly dissolved concentration vs. total dissolved
concentration. 97
Fig. 5. Trace metal ratios of various size fractions vs. salinity. 99
Fig. 6. Fe concentration vs. DOC in the HMW fraction. 101
Fig. 7. Pb vs. Fe concentrations in the HMW fraction. 101
Fig. 8. Co vs. Fe concentrations in the HMW fraction and Co vs. DOC in
the UD fraction. 104
Abbildungsverzeichnis
- 3 -
Kapitel 6
Fig. 1. The Mundaú-Manguaba lagoon system in Northeast Brazil. 110
Fig. 2. Calculated total dissolved concentration vs. separately measured
total dissolved trace metal concentration. 113
Fig. 3. The estuarine distribution of suspended particulate matter vs.
salinity (a) in the Mundaú Lagoon and (b) in the Manguaba
Lagoon. 115
Fig. 4. The estuarine distribution of DOC vs. salinity (a) in the Mundaú
Lagoon and (b) in the Manguaba Lagoon. 116
Fig. 5. Estuarine distribution of Cd in the Mundaú Lagoon: (a) total
dissolved Cd, (b) particulate Cd and (c) Cd percentage of colloidal
size fractions. 117
Fig. 6. Estuarine distribution of Cd in the Manguaba Lagoon: (a) total
dissolved Cd, (b) particulate Cd and (c) Cd percentage of colloidal
size fractions. 118
Fig. 7. Estuarine distribution of Cu in the Mundaú Lagoon: (a) total
dissolved Cu, (b) particulate Cu and (c) Cu percentage of colloidal
size fractions. 120
Fig. 8. Estuarine distribution of Cu in the Manguaba Lagoon: (a) total
dissolved Cu, (b) particulate Cu and (c) Cu percentage of colloidal
size fractions. 121
Fig. 9. Estuarine distribution of Fe in the Mundaú Lagoon: (a) total
dissolved Fe, (b) particulate Fe and (c) Fe percentage of colloidal
size fractions. 122
Fig. 10. Estuarine distribution of Fe in the Manguaba Lagoon: (a) total
dissolved Fe, (b) particulate Fe and (c) Fe percentage of colloidal
size fractions. 123
Fig. 11. Estuarine distribution of Pb in the Mundaú Lagoon: (a) total
dissolved Pb, (b) particulate Pb and (c) Pb percentage of colloidal
size fractions. 124
Fig. 12. Estuarine distribution of Pb in the Manguaba Lagoon: (a) total
dissolved Pb, (b) particulate Pb and (c) Pb percentage of colloidal 125
Abbildungsverzeichnis
- 4 -
size fractions.
Fig. 13. Estuarine distribution of Co in the Mundaú Lagoon: (a) total
dissolved Co, (b) particulate Co and (c) Co percentage of colloidal
size fractions. 126
Fig. 14. Estuarine distribution of Co in the Manguaba Lagoon: (a) total
dissolved Co, (b) particulate Co and (c) Co percentage of colloidal
size fractions. 127
Fig. 15. Estuarine distribution of Ni in the Mundaú Lagoon: (a) total
dissolved Ni, (b) particulate Ni and (c) Ni percentage of colloidal
size fractions. 128
Fig. 16. Estuarine distribution of Ni in the Manguaba Lagoon: (a) total
dissolved Ni, (b) particulate Ni and (c) Ni percentage of colloidal
size fractions. 129
Fig. 17. Log10KD (L/kg) vs. salinity. 132
Tabellenverzeichnis
5
Tabellenverzeichnis
Kapitel 1-3
Tab. 1. Mikrowellen-Programme einschließlich der Parameter. 29
Tab. 2. Massenbilanzen der Metalle. 32
Tab. 3. Ästuarines Verhalten von gelösten Spurenmetallen. 34
Tab. 4. Durchschnittliche Konzentrationen der gelösten Spurenmetalle
und des DOC in den vier Studiengebieten. 52
Tab. 5. Durchschnittliche Konzentrationen der partikulären Spurenmetalle
und des SPM in den vier Studiengebieten. 53
Tab. 6. Konzentrationen der gelösten Spurenmetalle in Ästuaren. 55
Tab. 7. Konzentrationen der partikulären Spurenmetalle und des SPMs in
Ästuaren. 56
Kapitel 4
Tab. 1. Dissolved trace metals in freshwater, in the estuarine mixing zone
and at high salinities of the WR and the WWR during the dry
season (2006) and the wet seasons (2007 and 2008). 69
Tab. 2. Particulate trace metals in freshwater, in the estuarine mixing
zone and at high salinities of the WR and the WWR during the dry
season (2006) and the wet seasons (2007 and 2008). 75
Tab. 3. Particulate trace metals and SPM of East-Hainan estuaries in
comparison to selected Chinese rivers and estuaries. 79
Tab. 4. Dissolved trace metals of East-Hainan estuaries in comparison to
selected Chinese rivers and estuaries. 83
Kapitel1: Einführung
- 6 -
1. Einführung
Der Ausdruck „Umwelt“ wurde von Jakob Johann von Uexküll als zentraler Begriff
der Ökologie eingeführt. Mit seinem Buch „Umwelt und Innenwelt der Tiere“ (1909)
war er maßgeblich beteiligt an der Geburt des Begriffs „Umwelt“ als „die Umgebung
eines Lebewesens, die auf dieses einwirkt und seine Lebensumstände
beeinflusst“ (z.B. Mildenberger, 2009). Seit der menschlichen Zivilisation, spätestens
seit der Industrialisierung, ist das Streben der Menschen nach Energie und
Rohstoffen nicht mehr zu stoppen. Dementsprechend werden auch eine Vielzahl von
Umweltchemikalien (Stoffe, die durch menschliches Zutun in die Umwelt gebracht
werden und in Mengen oder Konzentrationen auftreten können, die geeignet sind,
Lebewesen, insbesondere den Menschen, zu gefährden; Bundesregierung, 1972)
absichtlich oder unbeabsichtigt in die Umwelt entlassen, zu den auch Spurenmetalle
gehören. Insbesondere die industrielle Anwendung ermöglicht die Ausbreitung von
Spurenmetallen, die anderenfalls nicht leicht in die Natur gelangen würden (Hem,
1970). Die dadurch entstandenen Umweltveränderungen können lokale und
regionale Ausmaße haben, teilweise jedoch auch globalen Charakter annehmen
(Koch, 1995). Bisher ist nur ein Teil der Umweltchemikalien einschließlich deren
Einflüssen und Wirkungen auf natürliche Ökosysteme sowie auf die
Lebensgrundlagen und die Gesundheit des Menschen ausreichend erforscht (z.B.
Judson et al., 2009). Umso wichtiger ist es daher, auch kleine negative
Veränderungen in der Umwelt durch gezielte Beobachtung und Überwachung
frühzeitig zu erkennen, ihre Ursachen aufzuklären und ggf. ihrem Fortgang
entgegenzuwirken.
Diese Arbeit beschäftigt sich mit der Konzentration, der Verteilung und dem
Verhalten von verschmutzungsrelevanten Spurenmetallen in der wässrigen und
partikulären Phase sowie der Abschätzung der Kontaminationsniveaus der
untersuchten Metalle in den jeweiligen Studiengebieten. Dabei ging es um zwei
ästuarbezogene wissenschaftliche Projekte, die jeweils zu den Verbund-Projekten
„LANCET (Land-Sea Interactions along Coastal Ecosystems of Tropical China:
Hainan) “ und „POLCAMAR (Impact of Pollutants from Sugar Cane Monoculture on
Estuaries and Coastal Waters of North-East-Brazil)“ gehörten. Bei dem Projekt
LANCET sollte aufgeklärt werden, inwieweit anthropogene Einflüsse zum
Niedergang der Korallenriffe an der Ost-Küste der chinesischen Insel Hainan
beitragen, und bei dem Projekt POLCAMAR stand die Untersuchung der Auswirkung
der Zuckerrohrmonokultur auf Ästuare und Küstenwässer Nordostbrasiliens im
Vordergrund. Da Ost-Hainan möglicherweise durch lokale Aquakultur verschmutzt ist
Kapitel1: Einführung
- 7 -
und das Mundaú-Manguaba Lagunen-System von Abwässern der
Zuckerrohrmonokultur und der Zuckerrohrfabriken belastet sein kann, standen die
Beeinträchtigung der jeweilige Studiengebiete durch Aquakultur bzw.
Zuckerrohrmonokultur im Fokus der beiden Projekte. Die Forschungsergebnisse
können dazu beitragen, die Wirkung von anthropogenen Aktivitäten mit dem
Schwerpunkt Spurenmetalle auf lokale Ökosysteme generell besser zu verstehen.
Die Ziele dieser Arbeit waren daher:
1. Bestimmung der gelösten, partikulären und kolloidalen Konzentrationen von
sechs verschmutzungsrelevanten Spurenmetallen im Wanquan Ästuar und im
Wenchang-Wenjiao Ästuar in Ost-Hainan sowie in der Mundaú Lagune und in der
Manguaba Lagune in Nordost-Brasilien.
2. Interpretation der ästuarinen Verteilung und des ästuarinen Verhaltens der
gelösten, partikulären und kolloidalen Spurenmetalle in den oben genannten
Studiengebieten.
3. Vergleich der Konzentrationen und des Verhaltens der sechs untersuchten
Spurenmetalle in den genannten Studiengebieten mit anderen Ästuaren weltweit
sowie Beurteilung der Kontaminationsniveaus in den Studiengebieten.
Kapitel1: Einführung
- 8 -
1.1. Verschmutzungsrelevante Spurenmetalle in aquatischen Systemen
Spurenmetalle sind in der Umwelt allgegenwärtig. Dabei sind die Hydrosphäre und
vor allem die Ozeane von besonderer Bedeutung. Denn Ozeane können als
Depositionsort aller Materialien, die aus anderen Geosphären stammen, betrachtet
werden (Chester, 1990). Spurenmetalle gelangen in die Ozeane hauptsächlich durch
Transport über Flüsse, hydrothermale Prozesse und Einträge aus der Atmosphäre.
Der erste Weg ist besonders von Interesse, da verschmutzungsrelevante
Spurenmetalle als Folge von anthropogenen Aktivitäten in großem Umfang in die
Ozeane gelangen. Während Organismen gegenüber essentiellen Metallen eine
relativ hohe Toleranz haben, wirken unbenötigte Metalle oft toxisch (z.B. Bach, 2012).
Schließlich kann eine kleine Erhöhung des Eintrags verschmutzungsrelevanter
Spurenmetalle große Beeinträchtigungen für Lebewesen im Meer mit sich bringen.
1.1.1. Eisen
Eisen ist als Bestandteil verschiedener Enzyme und Proteine von essentieller
biologischer Bedeutung. Negative Auswirkungen an Menschen infolge einer
Überdosis von Eisen wurden jedoch auch nachgewiesen (z.B. Pigeon et al., 2001).
Eisen ist an der Ästuarchemie entscheidend beteiligt (z.B. Santschi et al., 1997).
Deshalb wurde es in dieser Arbeit untersucht, obwohl es nicht zu den
verschmutzungsrelevanten Spurenmetallen im engeren Sinne gehört.
Im Meer liegt Eisen entweder als reduziertes Eisen(II) oder als oxidiertes Eisen (III)
vor (z.B. Rose and Waite, 2005). Als Spezies kommen vor allem Eisen(hydr)oxide
(z.B. Sulzberger et al., 1989) und organische Eisenkomplexe (z.B. Mill, 1980) in
Betracht sowie evtl. Eisensulfide in den Sedimenten (z.B. Davison et al., 1999; Lapp
and Balzer, 1994). Jedoch ist die Löslichkeit des Eisens durch die Gleichgewichte
zwischen gelöster Phase und festen Phasen begrenzt (z.B. Balzer, 1982). Unter
oxischen Bedingungen sollte die maximale Konzentration für gelöstes Eisen im Meer
nur etwa 0.2 nmol/L betragen (Millero, 1995), wegen organischer Komplexierung liegt
sie jedoch um den Faktor 3-5 höher (Rue and Bruland, 1995). Das fluviale Eisen wird
hauptsächlich in Form von Kolloiden und Partikeln in die Ozeanen transportiert (z.B.
Fox and Wofsy, 1983).
1.1.2. Nickel
Nickel findet vor allem Verwendung bei der Produktion von Stahl und Legierungen
Kapitel1: Einführung
- 9 -
sowie in der Galvanik- und Elektronikindustrie (Galler, 1999). Nickel wird bei
manchen Tieren als essentielles Spurenelement angesehen; auch für Pflanzen und
Plankton kann es bioaktiv wirken (z.B. Bruland et al., 1991; Galler, 1999).
Nickel existiert in natürlichen Wässern als Nickel(II) (z.B. Mouvet and Bourg, 1983).
In Meerwasser liegt Nickel zum Teil in Komplexen mit organischen Liganden vor, die
eine hohe Stabilität besitzen (z.B. Achterberg and Van den Berg, 1997).
Anorganisches Nickel ist entweder hydratisiert oder mit u.a. Chlorid- und Carbonat-
Ionen komplexiert (z.B. Donat et al., 1994).
1.1.3. Cobalt
Cobalt hat eine Vielzahl von Anwendungen z.B. in der Farbenindustrie, Glas- und
Keramikproduktion, Lebensmittelindustrie und als industrieller Katalysator (z.B.
Hamilton, 1994). In der Natur ist Cobalt essentiell für die biotische Stickstofffixierung,
kann jedoch auch schädlich für lebende Organismen sein (z.B. Kim et al., 2006).
Zwar ist Cobalt für die Bildung von Vitamin B12 (Cobalamin) im menschlichen Körper
notwendig, doch eine Überdosis kann die Schilddrüsenaktivität mindern (z.B.
Barceloux, 1999).
In Ozeanen ist Cobalt primär als Cobalt(II) vorhanden (z.B. Nolan et al., 1992). Ältere
Untersuchungen behaupteten, dass hydratisierte Ionen sowie Ionenpaare mit Chlorid
und Sulfat die überwiegenden Cobalt-Spezies im Meerwasser darstellen (z.B.
Ćosović et al., 1982). Neuere Studien weisen jedoch darauf hin, dass mindestens ein
Teil des Cobalts organisch gebunden ist (z.B. Ellwood and Van den Berg, 2001). Als
biologisches Abbauprodukt von Vitamin B12 soll inertes Cobalt(III) ebenfalls
anwesend sein (z.B. Saito and Moffett, 2001).
1.1.4. Kupfer
Neben Gold, Silber und Zink war Kupfer ein Metall, welches als eines der Ersten in
der Menschheit eine Anwendung fand. Es ist vor allem wegen seiner sehr guten
elektrischen und thermischen Leitfähigkeit vielseitig verwendbar (z.B. Wiberg, 2007),
und dementsprechend in der Industrie allgegenwärtig (z.B. Blake et al., 2004). Kupfer
ist ein für die Ernährung aller Lebewesen essentielles Element, jedoch können bei
Überschuss toxische Wirkungen bei Pflanzen und Tieren auftreten (z.B. Galler, 1999).
Obwohl unter anoxischen Bedingungen das reduziertes Kupfer(I) die dominante
Spezies sein kann (Boulègue, 1983), ist im sauerstoffhaltigen Meerwasser Kupfer(II)
Kapitel1: Einführung
- 10 -
vorherrschend (z.B. Gledhill et al., 1997). Es bestehen Gleichgewichte zwischen
freien Kupfer(II)-Ionen und organisch sowie anorganisch komplexiertem Kupfer(II)
(z.B. Gledhill et al., 1997), wobei organisches Kupfer(II) über 90% der gesamten
Kupfer(II)-Spezies darstellt (z.B. Buckley and Van den Berg, 1986).
1.1.5. Cadmium
Cadmium ist ein toxisches Metall, welches auch in großem Umfang eine industrielle
Herkunft hat (Järup and Åkesson, 2009). Es kann bei Korrosionsschutz und
Akkumulatorherstellung sowie als Stabilisator für Kunststoffe eingesetzt werden (z.B.
Wurtz and Maeder, 2002). Cadmium wirkt bereits in geringer Konzentration toxisch
und ist daneben relativ mobil (Galler, 1999). Es greift vor allem die Nieren und den
Knochen des menschlichen Körper an und führt zu z.B. Nierenversagen und der so
genannten „Itai-Itai Krankheit“ (z.B. Satarug et al., 2010). Doch es kann auch die
biokatalytische Aktivität von u.a. marinen Kieselalgen ankurbeln (z.B. Lane and Morel,
2000).
Unter marinen Bedingungen wird Cadmium(II) als Hauptspezies angesehen (z.B.
Turner et al., 1981). Obwohl ein gewisser Anteil organischer Cadmiumspezies
identifiziert wurde (z.B. Ellwood, 2004), scheint die Assoziation von Cadmium mit
organischen Materien in den Ozeanen nicht sehr effektiv zu sein (z.B. Sunda and
Huntsman, 1998). Vielmehr ist Cadmium stark mit Chloriden komplexiert (z.B. Byrne
et al., 1988).
1.1.6. Blei
Blei ist eines der am längsten von Menschen verwendeten Metalle überhaupt.
Bereits in der frühen Bronzezeit wurde Blei verwendet, um Bronzen zu erzeugen.
Gegenwärtig findet es u.a. bei der Akkumulatoren-, Pigment- und
Legierungsherstellung Verwendung und bis vor kurzem als Treibstoffzusatz (z.B.
Wurtz and Maeder, 2002). Blei ist für die Menschen krebserregend und
nervensystemschädigend. Es kann einerseits eine große Zahl von Sulfhydryl-haltigen
Enzymen hemmen, Nukleinsäuren deformieren und die oxidative Phosphorylierung
stören, doch andererseits kann es mit Carboxypeptidase A gebunden z.B. die
Estersubstrate effektiv hydrolysieren (Vallee and Ulmer, 1972).
Die Oxidationsstufe +II ist die stabilste für Blei, und kommt in Ozeanen am häufigsten
vor (z.B. Town and Filella, 2002). Etwa 50% des gelösten Bleis in
Kapitel1: Einführung
- 11 -
Oberflächenmeerwasser ist assoziiert mit organischen Materien (z.B. Capodaglio et
al., 1990). Die andere Hälfte besteht wesentlich aus Carbonat- und Chlorid-
Komplexen (z.B. Millero et al., 2009).
1.2. Ästuar-Typen
Flüsse gehören zu den wichtigsten Trägern für die chemischen Signale, die sich in
Ozeanen zusammenfinden (z.B. Duce and Tindale, 1991). Dabei spielt das Ästuar
eine fundamentale Rolle für die Modifikation der Signale. “An estuary is a semi-
enclosed coastal body of water which has a free connection with the open sea and
within which sea water is measurably diluted with fresh water derived from land
drainage" (Pritchard, 1967). Als komplexe aquatische Systeme sind Ästuare sowohl
fluvialen als auch marinen Einflüssen ausgesetzt (z.B. Bach, 2012). Darüber hinaus
zählen die Ästuargebiete zu den am dichtesten besiedelten Gebieten der Welt. Auf
Grund dessen sind die Ästuare auch stark von anthropogenen Aktivitäten beeinflusst.
Wegen der unterschiedlichen Zusammensetzung von Fluss- und Meerwasser spielen
sich in Ästuaren eine Vielzahl von biologischen, physikalischen und chemischen
Reaktionen bzw. Umverteilungen ab. Angesichts solcher Modifikationen kann das
Ästuar als eine Art „Filter“ angesehen werden, welcher die fluvialen chemischen
Signale in der Mischzone erheblich verändern kann (Chester, 1990). Diese Fähigkeit
des Ästuars basiert größtenteils auf der Wechselwirkung zwischen gelöster und
fester Phase, welche sowohl durch physikochemische als auch durch biologische
Prozesse gesteuert wird.
Ästuare lassen sich in unterschiedliche Typen unterteilen. Nach u.a. Davies (1964)
werden Ästuare je nach ihre Tide zu micro-tidal (< 2 m), meso-tidal (2-4 m) und
macro-tidal (> 6 m) Typ klassifiziert. Jedoch ist eine ausreichende Vermischung von
Frisch- und Salzwasser einer der wichtigen Voraussetzungen für die filternde
Wirkung eines Ästuars. So werden Ästuare nach ihrem Mischungsgrad zwischen
salt-wedge, highly-stratified, partially-stratified und well-mixed Ästuaren
unterschieden, wobei sich Frisch- und Salzwasser in den salt-wedge Ästuaren am
schlechtesten und in den well-mixed Ästuaren am besten miteinander vermischen
(z.B. Berner and Berner, 1996). Außerdem ist von einem Ästuar zum anderen die
Geomorphologie auch unterschiedlich (z.B. Kjerfve and Magill, 1989). Unter den in
dieser Arbeit behandelten Studiengebieten gehört beispielsweise das Wanquan
Ästuar zu den klassischen Ästuaren mit einer direkten Flussmündung, wohingegen in
dem Wenchang-Wenjiao Ästuar und dem Mundaú-Manguaba Lagunen-System der
Kapitel1: Einführung
- 12 -
Fluss durch eine Lagune mit dem Meer verbunden ist. Unterschiede wie solche
können die ästuarinen Mastervariablen entscheidend beeinflussen, und so auch die
ästuarine Verteilung der Spurenmetalle.
Die in dieser Arbeit untersuchten Ästuarsysteme werden durch sog.
Weißwasserflüsse (Ertel et al., 1986; Konhauser et al., 1994; Bach, 2012) gespeist,
die sich durch relativ niedrige Huminstoff- und DOC- Konzentrationen (< ca. 5 mg/L)
sowie eine blasse Färbung auszeichnen; sie sind - im Gegensatz zu den sog.
Schwarzwasserflüssen - häufig durch höhere pH-Werte, niedrigere
Lösungskonzentrationen für Spurenmetalle und eher physikalische Verwitterungspro-
zesse im mineralisch geprägten Einzugsgebiet gekennzeichnet.
1.3. Mastervariablen für Spurenmetalle
Spurenmetalle können in verschiedenen physikochemischen Formen vorkommen,
u.a. als gelöste anorganische und organische Komplexe, Metallspezies in Form von
Kolloiden und Metallspezies, die an Kolloide oder Schwebstoffe adsorbieren (Mota
and Correia Dos Santos, 1995). Dabei streben sie nach der höchst möglichen
Stabilität, welche sie durch die Bindung an eine andere Spezies erreichen können
(Bach, 2012). Das Verhalten, der Transport und die Bioverfügbarkeit von
Spurenmetallen in Ästuaren hängen eng mit ihrer chemischen Speziation zusammen
(Chester, 1990). Einige Metalle wie z.B. Eisen (z.B. Wen et al., 1999) und Cadmium
(z.B. Elbaz-Poulichet et al., 1996) existieren in Ästuaren hauptsächlich in
anorganischer Form, während andere Metalle wie Kupfer eher als organische
Komplexe vorliegen (z.B. Shank et al., 2004). Jedoch sind die meiste Metalle in
Ästuaren über ein weiteres Spektrum von freien Ionen bis hin zu makromolekularen
Formen verteilt. Eine Menge von Mastervariablen, die teilweise voneinander
abhängen, können dabei ausschlaggebend sein, u.a. die Salinität, die chemische
Zusammensetzung des Wassers, die Residenzzeit, das suspendierte partikuläre
Material, das Vorkommen von Kolloiden, das Redoxpotential, der pH-Wert sowie die
biologische Aktivität.
1.3.1. Salinität und chemische Zusammensetzung
Als Salinität wird der Salzgehalt in einer wässrigen Lösung bezeichnet (z.B. Chester,
1990). Häufig wird für die Salinität in der ästuarinen Chemie die dimensionslose
Einheit „psu“ (practical salinity units) verwendet. In den meisten Veröffentlichungen
Kapitel1: Einführung
- 13 -
wird die Spurenmetallverteilung im Ästuar als eine Funktion der Salinität diskutiert.
Dabei handelt es sich darum, dass bei einer Veränderung der Salinität die
Grenzfläche Lösung/Festkörper, die Löslichkeit der Spurenmetalle und der
Aggregatszustand sich ändern können (z.B. Stolpe and Hassellöv, 2007). Solche
Veränderungen treten vor allem bei niedriger bis mittlerer Salinität auf. Das
berühmteste Beispiel dafür ist die Salz-induzierte Aggregation des Eisens bei
niedriger Salinität (z.B. Nowostawska et al., 2008).
Außerdem sind im Flusswasser Calcium und Bicarbonat die wichtigsten
anorganischen Ionen, während das Meerwasser von Natrium und Chlorid dominiert
ist (Chester, 1990). Diese Veränderung spiegelt sich im Ästuar wider, und kann eine
Speziationsänderung der Spurenmetalle hervorrufen. Nickel z.B. konkurriert mit
Calcium und Magnesium um die Bindungsstellen an organischen Materien (z.B.
Mandal et al., 2002), und die Zunahme von Chloridionen zum Meer hin kann die
Bildung von stabilen Cadmiumchlorid-Komplexe begünstigen (z.B. Dabrin et al.,
2009).
1.3.2. Residenzzeit
Die Zeitdauer, in der das Flusswasser im Ästuar verbleibt, wird als Residenzzeit
bezeichnet (z.B. Duinker, 1986). Die Residenzzeit eines Ästuars ist einer der
entscheidenden Faktoren in der ästuarinen Chemie. Je länger diese ist, desto besser
ist das ästuarine Wasser vermischt und umso mehr Zeit besteht für
Gleichgewichtseinstellungen zwischen Lösungsbestandteilen und verschiedenen
Partikeln. Die ästuarine Residenzzeit hängt u.a. stark von den Gezeiten und der
Geomorphologie eines Ästuars ab, und variiert beispielsweise für Ästuare in Europa
und den USA zwischen 1 und 200 Tagen (Uncles et al., 2002). In geomorphologisch
fast geschlossenen Ästuaren (z.B. Lagunen) kann die Residenzzeit extrem lang sein
- im Gegensatz zu klassischen Ästuaren (Kjerfve and Magill, 1989). Darüber hinaus
kann sich die Residenzzeit in einem Ästuar stark verändern. Je nach Tidensituation
variiert die Residenzzeit beispielsweise in der San Francisco Bay von 1 Tag bis zu
60 Tagen (Walters et al., 1985).
1.3.3. Suspendiertes partikuläres Material
Suspendiertes partikuläres Material (suspended particulate matter, SPM) hat große
Bedeutung für die ästuarine Modifikation der Spurenmetalle (z.B. Turner, 1996). Je
Kapitel1: Einführung
- 14 -
nach den Eigenschaften der jeweiligen Oberfläche können die im Wasser
schwebenden Partikeln sowohl adsorbierende als auch abstoßende Wirkung für die
im Wasser gelösten Stoffe haben. Spurenmetalle können sowohl an organische
Oberflächenfilme des SPM oder direkt an organische Partikel (z.B. Lion et al., 1982)
als auch an anorganische Partikel/Kolloide wie Eisen(hydr)oxide adsorbiert werden
(z.B. Turner et al., 2004). Wenn metallreiches SPM auf metallarmes Wasser trifft,
kann es auch zur Freisetzung von Metallen aus dem SPM in das Wasser kommen
(z.B. Edmond et al., 1985). Windaktivität und Flussmorphologie bewirken häufig eine
Aufwirbelung des Sediments, die beim bodennahen Zusammentreffen von Fluss-
und Meerwasser zu einem "Trübungsmaximum" - meist bei niedriger bis mittlerer
Salinität - führen kann (z.B. Turner et al., 1992; Chiffoleau et al., 1994; Kraepiel et al.,
1997; Martino et al., 2002). Dies ermöglicht einen intensiven Kontakt zwischen der
Festphase und der gelösten Phase, und kann einen besseren Austausch zwischen
den beiden Phasen hervorrufen. Zudem ist dort die Fließgeschwindigkeit des
Wassers sehr deutlich verringert. Etwa 90% des fluvialen SPM sedimentieren in den
Ästuaren (z.B. Chester, 1990), so dass die im SPM gebundenen Spurenmetalle
durch die Ästuare zum größten Teil abgefangen werden können, bevor sie das Meer
erreichen.
1.3.4. Größenspeziation (Kolloide)
Kolloide, bei denen häufig zwischen hoch- und niedermolekularen Entitäten
unterschieden wird, spielen eine zentrale Rolle für die Regulierung von Konzentration
und Speziation, den Transport und die Bioverfügbarkeit im natürlichen wässrigen
System (z.B. Guo et al., 2000a). Aufgrund ihrer vergleichsweise großen spezifischen
Oberfläche, die reaktive funktionale Gruppen enthält, können Kolloide eine Reihe von
Spurenmetallen adsorbieren (z.B. Waeles et al., 2008a). „An aquatic colloid is any
constituent that provides a molecular milieu into and onto which chemicals can
escape from the bulk aqueous solution, while its vertical movement is not significantly
affected by gravitational settling“ (Gustafsson and Gschwend, 1997). Bei Kolloiden
wird die Wechselwirkung an der Grenzfläche von den Oberflächeneigenschaften des
Partikels bestimmt (z.B. Lead and Wilkinson, 2007). Typische anorganische Kolloide
im natürlichen Gewässern beinhalten Aluminiumsilikate, Eisen- und Mangan-Oxide,
während Kohlehydrate und Fulvin-/ Huminstoffe zu den wichtigsten organischen
Kolloiden zählen (Filella, 2007). Aluminiumsilikate können z.B. Cadmium und Blei
adsorbieren (z.B. Kozar, et al., 1992) und Polysaccharide sind in der Lage, Kupfer,
Nickel und Blei aufzunehmen (z.B. Gérente, et al., 2000). Darüber hinaus ist in
Kapitel1: Einführung
- 15 -
wässriger Phase eine Reihe von Spurenmetallen mit Eisen-/Mangan-Oxiden und
Fulvin-/Huminstoffen assoziiert (z.B. Tessier, et al., 1996).
Nach IUPAC-Definition sollte ein Kolloid zumindest in einer Dimension eine Länge
zwischen typischerweise 1 nm und 1 μm haben (z.B. Everett, 1988). In Abb. 1 sind
typische umweltrelevante Kolloide und deren Größen aufgelistet. Auf Grund
methodischer Probleme wurden Kolloide lange Zeit zur gelösten Phase gezählt und
Prozesse auf kolloidaler Ebene, wie das Entfernen gelöster Spurenmetalle durch die
Ausflockung von Kolloiden in Ästuaren, nicht erkannt oder falsch interpretiert (z.B.
Bach, 2012). Die Verwendung neuartiger Aufbereitungs- und Analysentechniken
ermöglicht heutzutage die Separation und Bestimmung von Kolloiden. Jedoch ist
eine exakte Separation von Kolloiden nach physikalischer Größe wegen technischer
Schwierigkeiten nicht möglich (z.B. Buesseler et al., 1996). Dies wird in einem
späteren Kapitel diskutiert.
Abbildung 1. Korngrößenverteilung von verschiedenen Kolloiden und Partikeln in der Umwelt (Lead and Wilkinson, 2007).
Kapitel1: Einführung
- 16 -
1.3.5. Redoxpotential und pH-Wert
Das Redoxpotential ist von zentraler Bedeutung bei der ästuarinen Verteilung von
Spurenmetallen. So soll z.B. im Danshuei Ästuar kolloidales Cadmium nur unter
anoxischen Bedingungen relevant sein und nur relativ wenig unter oxischen
Bedingungen (Jiann et al., 2005). Bei Anwesenheit von Sauerstoff favorisieren Eisen
und Mangan die Festphase und können andere Spurenmetalle fixieren und mit in das
Sediment nehmen (z.B. Turner et al., 2008).
Der pH-Wert ist eine weitere Mastervariable in Ästuaren. Das Flusswasser hat in der
Regel einen pH-Wert von 5 bis 8, wohingegen das Oberflächen-Meerwasser einen
durchschnittlichen pH-Wert von 8.2 besitzt (Chester, 1990). Diese pH-Unterschiede
führen zu einem Gradient des pH-Werts in Ästuaren. Für die meisten Spurenmetalle
steigt die Löslichkeit mit fallendem pH-Wert (z.B. Hatje et al., 2003a). Außer direkter
Konkurrenz (zwischen H+ und Metallkationen) um die Bindungsstellen kann eine
Veränderung des pH-Werts u.a. die Umwandlungen von Metallspezies und die
Modifikationen von metall-bindenden Liganden beeinflussen (z.B. Kola and Wilkinson,
2005). Kinetische Studien offenbarten, dass mit steigendem pH-Wert die
Adsorptionsquoten von Eisen, Cobalt und Cadmium an das SPM zunehmen (Hatje et
al., 2003b). Auch die Bildung von Metall-Hydroxiden und -Carbonaten, die häufig in
wässrigen Systemen vorkommen (z.B. Turner et al., 1981), kann dabei entscheidend
gestört werden.
1.3.6. Biologische Aktivität
Da viele Spurenmetalle biologisch aktiv sind (z.B. Norisuye et al., 2007), kann deren
Verhalten im Ästuar auch von biologischen Aktivitäten abhängen, die wiederum u.a.
vom Sauerstoffgehalt (z.B. Beck and Sañudo-Wilhelmy, 2007) und der Residenzzeit
(z.B. Crump et al., 2004) beeinflusst werden. So kann z.B. die Adsorption von Blei an
der SPM-Oberfläche durch Algen und Bakterien gefördert werden (z.B. Nelson et al.,
1999). Die Bioakkumulation von u.a. Nickel (z.B. Hong et al., 2009), Cobalt (z.B.
Nolan et al., 1992) und Kupfer (z.B. Zamuda and Sunda, 1982) ist nachgewiesen
worden. Außerdem kann eine Algenblüte den pH-Wert steigen lassen und darüber
hinaus - wie eben diskutiert - das Adsorptionsverhalten der Spurenmetalle
beeinflussen (Hatje et al., 2003a).
Allerdings ist die biologische Verfügbarkeit der Spurenmetalle durch ihre
Ionenaktivität beschränkt (z.B. Zamuda and Sunda, 1982); anders ausgedrückt, nicht
alle Spurenmetallspezies sind biologisch verfügbar. Abgesehen von elementaren
Kapitel1: Einführung
- 17 -
Unterschieden sind die unkomplexierten Metallionen generell eher biologisch
verfügbar als die organisch komplexierten (z.B. Helmers, 1994).
1.4. Ästuarine Mischungsprozesse für Spurenmetalle
Aufgrund von Änderungen bei den oben genannten ästuarinen Variablen spielen sich
unterschiedliche chemische, physikalische und biologische Prozesse im Ästuar ab.
Dadurch lässt sich der Aggregatszustand der Spurenmetalle verändern.
Eisenkolloide koagulieren beispielsweise gern mit organischen Materien und flocken
unter ästuarinen Bedingungen aus (z.B. Mylon et al., 2004). Auch Blei wird leicht aus
der Lösung entfernt als Konsequenz der Adsorption an Eisen- oder Mangan-Kolloide
sowie an das SPM (z.B. Waeles et al., 2008b). Außerdem können z.B. Nickel und
Cadmium (Luoma et al., 1998) sowie Kupfer (Sharp et al., 1982) durch biologische
Aktivitäten aus der wässrigen Phase entfernt werden. Auf der anderen Seite ist auch
eine Freisetzung von Spurenmetallen aus dem Sediment oder aus dem SPM möglich.
Nickel verliert z.B. mit steigender Präsenz von Calcium- und Magnesiumionen
zunehmend seine Bindungsstellen auf der organischen Materie (z.B. Mandal et al.,
2002). Cobalt kann reduktiv aus dem Sediment ins Wasser freigesetzt werden (z.B.
Audry et al., 2006).
Außer zu Wechseln beim Aggregatszustand kann es in Ästuaren auch zu
Modifikationen bei der Komplexbildung der Spurenmetalle geben. Kupfer und
Cadmium können beispielsweise durch Abbau der organischen Substanz ihren
kolloidalen Charakter verlieren (z.B. Masson et al., 2011; Waeles et al., 2005). Der
Anteil des organisch gebundenen Nickels kann in Ästuaren auch abnehmen (Turner
and Martino, 2006).
Kapitel 2: Material und Methoden
- 18 -
2. Material und Methoden
2.1. Studiengebiete
Die Feldarbeiten wurden im Rahmen der Projekte LANCET und POLCAMAR auf der
Südchinesischen Insel Hainan bzw. an der Küste Nordostbrasiliens durchgeführt.
2.1.1. Ästuare auf der südchinesischen Insel Hainan
Abbildung 2. Die Studiengebiete: (a) Wenchang-Wenjiao Ästuar und (b) Wanquan Ästuar in Ost-Hainan.
Die tropische Insel Hainan liegt zwischen 18°09’-20°10’N und 108°03’-111°03’E im
südchinesischen Meer. Es herrschen warme und feuchte klimatische Bedingungen.
Das Wanquan Ästuar und das Wenchang-Wenjiao Ästuar liegen beide an der
Ostküste der chinesischen Insel (Abb. 2). In diesem Studiengebiet fallen etwa 80%
der ganzjährigen Niederschläge von 1740 mm in der Regensaison zwischen Mai und
November (Ma et al., 2007; Liu et al., 2011).
Der Wanquan Fluss ist mit einer Länge von 156.6 km, einem Einzugsgebiet von
3693 km2 und einem durchschnittlichen Wassereintrag von 163.9 m3/s der drittgrößte
Fluss auf Hainan (Ge et al., 2003). Der Wenchang Fluss (Länge 37 km,
Einzugsgebiet 381 km2, durchschnittlicher Wassereintrag 9.1 m3/s) und der Wenjiao
Fluss (Länge 56 km, Einzugsgebiet 522 km2, durchschnittlicher Wassereintrag 11.6
m3/s) münden beide in die Bamen Bucht (Wasseroberfläche ca. 40 km2; Herbeck et
al., 2013) und stellen gemeinsam das Wenchang-Wenjiao Ästuar dar (Zeng and
Zeng, 1989). Je nach Gezeiten kann das Salzwasser jeweils von der Yudai
Kapitel 2: Material und Methoden
- 19 -
Sandbarriere und von der Bamen Bucht 5 km flussaufwärts in die Ästuare eindringen.
Die Lage der Mischzone für Frisch- und Salzwasser im Wanquan Ästuar ist
angesichts des Zusammenwirkens von Frischwassereintrag sowie Tiden- und
Windeinfluss äußerst variabel. Hingegen bleibt im seichten Wenchang-Wenjiao
Ästuar (< 3 m, die Schiffroute ausgenommen; Fu et al., 2013) der Salinitätsgradient
relativ stabil. Die Residenzzeit des Wassers beträgt 0.2-4.7 Tage im Wanquan
Ästuar (Li et al., 2013) und 7.8 Tage im Wenchang-Wenjiao Ästuar (Liu et al., 2011).
Im Vergleich zu dem tourismusorientierten und umweltgeschützten Wanquan Ästuar
(z.B. Gao et al., 2004) ist das Wenchang-Wenjiao Ästuar von der Aquakultur geprägt.
Die gesamte von der Aquakultur bewirtschaftete Fläche in der Region wird auf ca.
21.6 km2 mit einem jährlichen Abwassereintrag ins Ästuar von ca. 210·106 m3
geschätzt (Herbeck et al., 2013; Unger et al., 2013). Dadurch ist das Wenchang-
Wenjiao Ästuar mit einiger Wahrscheinlichkeit umweltbelastet.
2.1.2. Lagunen-System in Nordost-Brasilien
Abbildung 3. Das Studiengebiet Mundaú-Manguaba Lagunen-System in Nordost-Brasilien.
Das Mundaú-Manguaba Lagunen-System liegt im nordostbrasilianischen
Bundesstaat Alagoas (Abb. 3). Seine tropische Lage deutet darauf hin, dass das
Kapitel 2: Material und Methoden
- 20 -
Klima dort warm und feucht ist. Es herrscht eine klar definierte Regensaison von
April bis August und eine Trockensaison von September bis März (Oliveira and
Kjerfve, 1993; Brockmeyer and Spitzy, 2011). Die durchschnittliche Temperatur
beträgt etwa 24°C (Diegues, 1994).
Das Mundaú-Manguaba Lagunen-System besteht aus zwei Lagunen. Die nördlich
liegende Mundaú Lagune hat den Rio Mundaú (Einzugsgebiet 2135 km2) als
Hauptfrischwasserzufuhr, eine Wasseroberfläche von 24 km2 und eine
durchschnittliche Wassertiefe von 1.5 m (Melo-Magalhães et al., 2009). Der Rio
Paraíba do Meio mit einem Einzugsgebiet von 3299 km2 ist der Hauptzufluss für die
südlich liegende Manguaba Lagune, welche eine Wasseroberfläche von 43 km2 und
eine durchschnittliche Wassertiefe von 2 m hat (z.B. De Souza et al, 2002; Melo-
Magalhães et al., 2009; Costa et al, 2011). Die beiden Lagunen sind miteinander
durch mehrere schmale Kanäle verbunden. Durch eine einzige Öffnung fließt das
Wasser aus den Kanälen in den Atlantik. Durch das Kanalsystem wird das
Eindringen von Salzwasser weitgehend gedämpft. Solche sogenannten
„choked“ Lagunen sind durch lange Residenzzeiten und die Anfälligkeit gegenüber
anthropogenen Beeinträchtigungen charakterisiert (Knoppers et al., 1991). Vor allem
in der Regensaison sind die beiden Lagunen vom Eintrag aus den Flüssen dominiert
(Spörl, 2011). Die durchschnittliche Residenzzeit beträgt 1-2 Wochen für die Mundaú
Lagune und 5-7 Wochen für die Manguaba Lagune (Oliveira and Kjerfve, 1993).
In den letzten Jahrzehnten ist das Mundaú-Manguaba Lagunen-System von der
Zuckerrohrmonokultur in den Einzugsgebieten seiner Zuflüsse stark geprägt worden.
Dies führt zu erhöhten Nährstoffeinträgen in die Lagunen und kann saisonale
Unterschiede in den Nährstoffkonzentrationen verursachen (Spörl, 2011). Darüber
hinaus liegt die Millionenstadt Maceío direkt am Ufer der Mundaú Lagune, was das
System zusätzlich belastet.
2.2. Maßnahmen zur Kontaminationsminderung
Die Spurenmetallkonzentration im Ästuar, insbesondere im Meerwasser-Endglied, ist
meistens sehr gering (z.B. Bruland, 1983). Während die analytischen Methoden
immer empfindlicher geworden sind, müssen mögliche Kontaminationen während der
Aufarbeitung als ein einschränkender Faktor gesehen werden (z.B. Kosta, 1982).
Aufgrund dessen fanden außer Probenahme und Filtration/Ultrafiltration alle Labor-
Aktivitäten in einem Reinraumlabor (Class 1000) statt; die Probenbehandlungen
wurden unter einer zusätzlichen Clean-Bench (Class 100) durchgeführt. Zum Einsatz
Kapitel 2: Material und Methoden
- 21 -
kamen ausschließlich supra-pur®-Chemikalien, außer Salpetersäure und "Freon"
(1,1,2-Trichlor-trifluorethan), die unter Reinraumbedingung destilliert wurden (sub-
boiled HNO3). Alle verwendeten Materialien (u.a. Gefäße) wurden zur Entfettung für
drei Tage in ein 2%-iges Mucasol-Bad (anionisches Tensid) eingelegt, danach mit
warmem Leitungswasser und dann mit Milli-Q®-Wasser gründlich gespült. Um die
Metalle an der Oberfläche der Materialien zu entfernen, wurden die Materialien
anschließend in Salzsäure (3 mol/L) und später in Salpetersäure (3 mol/L) für jeweils
drei Tage eingelegt. Dazwischen und am Ende wurden die Materialien mit Milli-Q®-
Wasser gründlich gespült und dann unter einer Clean-Bench getrocknet ("supra-
sauber"). Die Pipettenspitzen und Autosamplercups für das Massenspektrometer mit
induktiv gekoppeltem Plasma wurden für ca. eine Woche in Salpetersäure (1 mol/L)
eingelegt, mit Milli-Q®-Wasser gründlich gespült und unter einer Clean-Bench
getrocknet (Salpetersäure-sauber). Für einige Materialien, wie z.B. die Gefäße für
Mikrowellen-Aufschlüsse und Querstrom-Ultrafiltrations-Kassetten waren
methodenspezifische Reinigungsverfahren gefordert, welche in den jeweiligen
Kapiteln behandelt werden.
2.3. Probenahme
Die Probenahme fand in den jeweiligen Studiengebieten der südchinesischen Insel
Hainan bzw. in Nordost-Brasilien statt.
Die Wanquan und Wenchang-Wenjiao Ästuare wurden in der Trockensaison 2006
(Dezember), in der Regensaison 2007 (August) und in der Regensaison 2008
(Juli/August) beprobt.
Die Probenahme im Mundaú-Manguaba Lagunen-System erfolgte in der
Regensaison 2007 (September/Oktober), der Trockensaison 2008 (Februar) und der
Trockensaison 2009 (März).
Oberflächenwasserproben aus einer Tiefe von ca. 0.3 m wurden flussaufwärts von
einem Boot aus direkt in eine Polyethylenflasche (LDPE, Nalgene®) gefüllt. Vor der
Behandlung wurden die Proben gekühlt gelagert.
2.4. Filtration und Querstrom-Ultrafiltration
Im Labor vor Ort wurden die Oberflächenwasserproben innerhalb von 24 Stunden
filtriert und anschließend ultra-filtriert. Polycarbonatfilter (Nuclepore®, Porengröße
0.45 μm; vorher mit Salpetersäure gereinigt) wurden benutzt, um die Proben zu
Kapitel 2: Material und Methoden
- 22 -
filtrieren. Die mit Filterkuchen beladenen Filter wurden in einer Petrischale und das
Filtrat in einer Polyethylenflasche (LDPE, Nalgene®) aufbewahrt. Bevor ein Filtrat mit
konzentrierter Salpetersäure (sub-boiled, Verhältnis zur Probe 1:1000) angesäuert
wurde, wurden 10 mL davon für die Bestimmung von gelöstem organischem
Kohlenstoff (dissolved organic carbon; DOC) für ein anderes Projekt in eine
Glassampulle abgefüllt, mit Phosphorsäure (Verhältnis zur Probe 1:100) angesäuert
und durch Abschmelzen verschlossen. Etwa 1 L des Filtrats wurde nicht angesäuert
und stand für die Querstrom-Ultrafiltration zur Verfügung.
Die Querstrom-Ultrafiltration (Cross-Flow-Filtration, Tangential-Flow-Filtration) wird
seit den 90-er Jahren des letzten Jahrhunderts vermehrt in der Ästuarforschung
eingesetzt, um die Kolloide nach ihrer physikalischen Größe zu trennen, bzw. um die
kolloidalen Spezies in Ästuaren zu studieren (z.B. Whitehouse et al., 1990; Buesseler
et al., 1996; Guo and Santschi, 1996; Gustafsson et al., 1996; Wen et al., 1996; Guo
and Santschi, 2007). Diese Technik beruht auf einen Membran-Trennverfahren, bei
dem durch die Anströmung quer zur Filtrationsrichtung an der Membranoberfläche
eine hohe Strömungsgeschwindigkeit erzeugt wird, um die Ablagerung von Kolloiden
bzw. die Bildung von Deckschichten möglichst weitgehend zu verhindern. Da die
Eigenschaften der dennoch auftretenden Ablagerungen sich im Verlauf der Filtration
ändern können (z.B. Gustafsson et al., 1996), spielt der Konzentrationsfaktor CF eine
gewisse Rolle, welcher gegeben ist durch:
VVV
VVCF
R
PR
R
+== 0
Dabei ist VR das Volumen des Retentats, VP das Volumen des Permeats und V0 das
Ausgangsvolumen. Zur Überprüfung der Effizienz von Querstrom-Ultrafiltrationen
wird der Begriff der „Wiederfindung“ verwendet, welche dem Prozentsatz der Summe
der Gehalte aller erhaltenen Fraktionen bezogen auf die Ausgangsmenge entspricht
(z.B. Whitehouse et al., 1990):
100(%)00
××
×+×=
VCVCVCungWiederfind RRPP
CP, CR und V0 stehen jeweils für die Konzentration des Permeats, die des Retentats
und die Ausgangskonzentration. Bei einer Wiederfindung größer als 100% kommt
Kontamination in Betracht (z.B. Martin et al., 1995), eine kleinere Wiederfindung als
100% deutet auf einen Verlust während der Querstrom-Ultrafiltration hin. Um einen
hohen Konzentrationsfaktor zu erreichen, braucht die Querstrom-Ultrafiltration zwar
viel Probe (Waeles et al., 2008a); vorteilhaft ist jedoch, dass dadurch die späteren
Kapitel 2: Material und Methoden
- 23 -
Analyten im Retentat angereichert werden und so die Anforderungen an die
Messinstrumente reduziert werden.
Um die Bedeutung des Konzentrationsfaktors gibt es erhebliche Uneinigkeiten. Für
Meerwasser wurde z.B. ein Konzentrationsfaktor von über 40 empfohlen (Guo and
Santschi, 1996). Auf der anderen Seite kann ein hoher Konzentrationsfaktor jedoch
zum Durchbruch des Retentats durch die Membran führen (z.B. Dai et al., 1998).
Auch Matrix-Effekte (z.B. Whitehouse et al., 1990) und Aggregation von Kolloiden
(z.B. Waeles et al., 2008a) können vermehrt auftreten. In manchen Literaturstellen
wird ein Konzentrationsfaktor von 5-10 als optimaler Konzentrationsfaktor angesehen
(z.B. Wen et al., 1996). Außerdem verändert sich die Performance der
Ultrafiltrations-Membran je nach Konzentration und Natur der Probe (z.B. Baker and
Strathmann, 1970). Eine Übereinstimmung über einen optimalen
Konzentrationsfaktor ist noch nicht gegeben (Guo et al., 2000b).
Außer der Querstrom-Ultrafiltration ist die Frontal-Ultrafiltration (z.B. Jiann et al.,
2005; Guo et al., 2011) eine gängige Methode, um die Kolloide nach ihrer
physikalischen Größe zu trennen. Diese Technik ist ausgezeichnet vor allem durch
einfache Bedienung (Waeles et al., 2008a), führt jedoch leicht zu Überladung des
Filters, welche die effektive Porengröße des Filters vermindern und eine einheitliche
Fraktionierung von Kolloiden verhindern können (z.B. Benoit and Rozan, 1999).
Deswegen hat man sich in meisten Kolloidstudien für die Querstrom-Ultrafiltration
entschieden.
Verschiedene Querstrom-Ultrafiltrations-Systeme wie z.B. Amicon® (z.B. Benoit et al.,
1994; Tang et al., 2001), Millipore Pellicon® (z.B. Whitehouse et al., 1990; Dai et al.,
1995) und Vivaflow® (z.B. Gaffney et al., 2008; Schlosser and Croot, 2008) wurden
entwickelt. Im Wesentlichen unterscheiden sich diese Systeme hinsichtlich des
Designs, des Membranmaterials und der Membranoberfläche (z.B. Guo and Santschi,
2007). Ein System aus Edelstahl ist gut geeignet für Fraktionierung organischer
Materien, während bei der Spurenmetallanalyse Kunststoffmaterialien einzusetzen
sind. Eine Tiefenfilter-Membran kann beispielsweise nur 60% Poren haben, die der
effektiven Filtrationsrate entsprechen, während die anderen 40% entweder größer
oder kleiner sind (Guo and Santschi, 2007). Zudem ist die Sorptionskapazität der
Membran auch von Material zu Material unterschiedlich. Eine große aktive
Membranoberfläche ermöglicht zwar die schnelle Ultrafiltration, damit vergrößert sich
jedoch gleichzeitig das Totvolumen des Systems.
Da die wichtigen kolloidalen Spurenmetallenträger, wie Huminstoffe und
Kapitel 2: Material und Methoden
- 24 -
Eisen(hydr)oxide, unterschiedliche Größen und -bereiche aufweisen (z.B. Lead and
Wilkinson, 2007), wird manchmal eine Mehrfachfraktionierung von Kolloiden
gewünscht. Dafür wird der Begriff NMWCO (nominal molecular weight cut-off) als
Maßstab eingesetzt, um die Membranen zu charakterisieren. Bei einem definierten
NMWCO werden rund 90% der spezifischen globularen Moleküle, die größer als die
Poren der Membran sind, zurückgehalten (z.B. Guo and Santschi, 2007). Da
natürliche Kolloide selten kugelförmig sind, kann die reale Filtrationsrate jedoch von
der idealen abweichen.
Konzeptionell kommen Kaskaden- und Parallel-Prozeduren in Betracht. Bei der
Kaskaden-Prozedur werden die Membranen nach Porengrößen bzw. NMWCOs
sukzessiv zugeschaltet. Je nach der Anforderung werden aus einer Probe mehrere
Subproben mit unterschiedlichen Kolloidgrößen. Das Permeat der jeweils
vorausgegangenen Ultrafiltration wird als Ausgangsprobe des nachfolgenden
Ultrafiltrationsvorgangs eingesetzt. Diese Prozedur hat den Vorteil, dass die Kolloide
nach Größe nacheinander entfernt werden. Dadurch wird die Konzentration des
Retentats möglichst klein gehalten und die Aggregation von Kolloiden an der
Membranoberfläche minimiert (Waeles et al., 2008a). Da bei der Querstrom-
Ultrafiltration nicht nur größere Kolloide in konzentrierter Form, sondern auch
kleinere Kolloide im Retentat vorhanden ist, wird die Konzentration der größeren
Kolloide CCol wie folgt berechnet:
CFCCC PR
Col−
=
Dabei stellt die Fortpflanzung von analytischen Fehlern das größte Problem dar.
Besonders bei der kleinsten "Kolloid"-Fraktion, die das Permeat (sog. "echt gelöste"
Fraktion) des letzten Querstrom-Ultrafiltrationsvorgangs darstellt und nicht
aufkonzentriert ist, kann eine Kontamination schwerwiegende Folgen für die
Berechnung der Konzentrationen größerer Kolloid-Fraktionen haben. Vorteilhafter in
Hinblick darauf ist die Parallel-Prozedur (z.B. Laxen and Chandler, 1983), bei der
mehrere Aliquote einer einzigen Probe durch Filter unterschiedlicher Porengröße
parallel ultrafiltriert werden, um Fraktionen unterschiedlicher Kolloidgröße zu erhalten.
Dadurch wird die Fehlerfortpflanzung zwar verhindert, aber besonders bei der
Fraktionierung kleiner Kolloidgrößen kommt es leicht zur Aggregation an der
Membranoberfläche und die Modifikation des Retentats ist schwer zu vermeiden.
Kapitel 2: Material und Methoden
- 25 -
Abbildung 4. Schematische Darstellung der Querstrom-Ultrafiltrations-Kassette Vivaflow® von Sartorius (www.sartorius-stedim.com).
Abbildung 5. Ablauf der Querstrom-Ultrafiltrations-Prozedur.
In dieser Arbeit wurden Querstrom-Ultrafiltrations-Kassetten aus Polycarbonat
(Vivaflow® 200, Sartorius; Abb. 4) mit einer Polyethersulfon-Membran und einer
Kapitel 2: Material und Methoden
- 26 -
aktiven Oberfläche von 200 cm2 eingesetzt; sie wurden mit Porengrößen von 30 kDa,
10 kDa und 5 kDa kaskadisch betrieben. Das System ist vor allem ausgezeichnet
durch hohe Fluss- und gute Wiederfindungsraten. Die Querstrom-Ultrafiltrations-
Kassetten wurden vor Gebrauch jeweils 30 Minuten lang erst mit Salzsäure (1%) und
dann mit EDTA (10 mmol/L) durchspült. Am Anfang, zwischen den beiden
Reinigungsprozeduren und am Ende wurden die Querstrom-Ultrafiltrations-Kassetten
jeweils mit großen Mengen an Milli-Q®-Wasser gespült. Ein schematischer Ablauf der
Querstrom-Ultrafiltrations-Prozedur einschließlich der Filtrationsprozedur ist in Abb. 5
dargestellt. Nach der konventionellen Filtration (0.45 μm Filter) wurden die Filtrate
zuerst durch eine Membran mit einem NMWCO von 30 kDa ultrafiltriert. Die
Fliessgeschwindigkeit der Probe durch die Querstrom-Ultrafiltrations-Kassette betrug
dabei etwa 500 mL pro Minute bei einem Druck von ca. 2 bar. Das Permeat wurde
als Ausgangsprobe für nachfolgende Fraktionierung von Kolloiden mit einem
NMWCO von 10 kDa verwendet. Schließlich wurde das daraus entstandene Permeat
weiter als Ausgangsprobe für die Fraktionierung der 5 kDa Kolloide eingesetzt. Nach
der Querstrom-Ultrafiltration wurden die drei Retentate und das Permeat (< 5 kDa)
jeweils in eine Polyethylenflasche (Nalgene®) abgefüllt und mit konzentrierter
Salpetersäure auf einen pH-Wert von ca. 2 angesäuert. Zur Volumenbestimmung
wurden die drei Retentate und das Permeat sowie die Ausgangsprobe vorher
gewogen. Jeweils 10 mL von den Retentaten und dem Permeat wurden ebenfalls vor
der Ansäuerung für die DOC-Bestimmung in eine Glassampulle abgefüllt, mit
Phosphorsäure angesäuert und abgeschmolzen. Die durchschnittlichen
Konzentrationsfaktoren betrugen ca. 7.5 für die Fraktion mit einer Größe zwischen 30
kDa und 0.45 μm, ca. 6 für die Fraktion mit einer Größe zwischen 10 kDa und 30 kDa,
und ca. 4.5 für die Fraktion mit einer Größe zwischen 5 kDa und 10 kDa. Um die
Membranen zu konditionieren, wurden die Kassetten vor jeder Ultrafiltration mit der
zu filtrierenden Probe gespült.
Bei der Untersuchung hat sich herausgestellt, dass die Kolloidfraktion mit einer
Größe zwischen 10 kDa und 30 kDa im Hinblick auf die ästuarine Modifikation relativ
unrelevant ist. In der folgenden Diskussion wurde sie daher mit der Kolloidfraktion mit
einer Größe zwischen 30 kDa und 0.45 μm zusammen betrachtet und HMW-Fraktion
(high molecular weight) genannt. Die Kolloidfraktion mit einer Größe zwischen 5 kDa
und 10 kDa und die echt gelöste Fraktion (< 5 kDa) wurden als LMW- (low molecular
weight) bzw. TD-Fraktion (truly dissolved) bezeichnet.
Die Querstrom-Ultrafiltrations-Technik - und die gegenwärtigen Techniken zur
Kolloidfraktionierung überhaupt - basieren hauptsächlich auf der physikalischen
Kapitel 2: Material und Methoden
- 27 -
Abtrennung von Kolloiden (z.B. Lyvén et al., 2003). Dabei sind die Struktur und das
Verhalten von Kolloiden nicht unbedingt von ihren physikalischen Größen abhängig
(Lead and Wilkinson, 2007). Neben der chemischen Zusammensetzung sind
Faktoren wie Beladung, funktionelle Gruppen und spezifische Oberfläche ebenfalls
von Bedeutung. Wie oben schon erwähnt, können die Ablagerung und der
Durchbruch von Kolloiden die Separation beeinträchtigen. Eine Ablagerung kann
jedoch unter Umständen effektiv bekämpft werden (z.B. Schlosser and Croot, 2008),
während der Durchbruch durch Verringerung des Konzentrationsfaktors weitgehend
zu dämpfen ist (Guo et al., 2000b). Gleichzeitig wurden auch neue Techniken wie z.B.
Flow-Field-Flow-Fraktionierung (z.B. Stolpe et al., 2010) entwickelt. Solche Verfahren
können zum besseren Verstehen von ästuarinen Kolloiden beitragen.
2.5. UV-Aufschluss
Organische Moleküle und organische Partikel können bekanntlich mit Metallionen
Komplexe bilden, die im Wasser sehr stabil und zum Teil resistent gegen Säure sind
(z.B. Golimowski and Golimowska, 1996). Infolgedessen wird traditionell ein UV-
Aufschluss-Verfahren unter Anwesenheit von Oxidationsmitteln bei der Analyse
natürlicher Wässer angewendet, um die organischen Matrix zu zerstören bzw. die zu
untersuchenden Metallionen zu befreien (z.B. Achterberg et al., 2001). Im Reinraum
wurden alle wässrigen Proben (im Wesentlichen nach: Danielsson et al., 1978) unter
Einsatz von Wasserstoffperoxid (30%) als Oxidationsmittel (Verhältnis zur Probe
1:100) in Quarzglasgefäßen für 2 Stunden mit UV-Licht bestrahlt. Nach der
Behandlung wurden die Proben wieder in eine vorgereinigte LDPE-Flasche gefüllt
und waren bereit für den nächsten Verarbeitungsschritt.
Zur Reinigung wurden die Quarzgefäße zuerst in Mucasol (5%) und anschließend in
Salpetersäure (3 mol/L) für jeweils zwei Stunden mit Ultraschall behandelt. Nach
dem Mucasol-Bad wurden sie mit warmem Leitungswasser und dann mit Milli-Q®-
Wasser gründlich gespült; nach dem Salpetersäure-Bad erfolgte ebenfalls ein
Spülvorgang mit Milli-Q®-Wasser. Schließlich wurden sie unter einer Clean-Bench
getrocknet.
2.6. Flüssig-Flüssig-Extraktion
Die Flüssig-Flüssig-Extraktion ist eine gebräuchliche Methode zur Untersuchung von
Spurenmetallen in Meerwasser bzw. in ästuarinen Wässern (z.B. Komjarova and
Kapitel 2: Material und Methoden
- 28 -
Blust, 2006), da der hohe Salzanteil im Meerwasser und Ästuarwasser die
Auswertung der analytischen Signale stören und den hochempfindlichen Detektor
des Messinstruments beschädigen kann. Darüber hinaus treten in solchen Medien
häufig Matrixeffekte auf, was zu Verfälschungen des Messergebnisses führen kann
(z.B. Rodushkin and Ruth, 1997). Durch die Flüssig-Flüssig-Extraktion werden die
störenden Alkali- und Erdalkali-Ionen ausgeschlossen und Matrixeffekte weitgehend
minimiert. Die zu untersuchenden Spurenmetalle werden zudem aufkonzentriert.
Trotz aller Vorteile ist das Verfahren jedoch relativ aufwendig und benötigt eine
Reihe von Instrumenten und Chemikalien. Eine Kontamination während der
Verarbeitung gilt es zu verhindern.
Im Rahmen dieser Arbeit wurde eine Flüssig-Flüssig-Extraktions-Methode im
Wesentlichen nach Bruland et al. (1985) entwickelt. Zuerst wurden je nach
Einschätzung der Probekonzentration 40-100 g Probe in einen fluorierten Ethylen-
Propylen Scheidetrichter (FEP, Nalgene®) eingewogen. Die Lösung wurde mit
Ammoniumacetat-Puffer (selbst hergestellt aus Ammoniak und Essigsäure) auf einen
pH-Wert von 4.0-4.5 gebracht. Als Chelatbildner wurden 500 μL einer Mischung von
Natriumdiethyldithiocarbamat und Ammoniumpyrrolidindithiocarbamat (Verhältnis 1:1,
jeweils 0.05 mol/L) zugesetzt und durchmischt. Nach Zugabe von 10 mL Freon
wurde der Scheidetrichter für 15 Minuten geschüttelt. Nach der Phasenseparation
wurde die untere organische Phase in ein Polypropylenröhrchen abgelassen und mit
100 μL konzentrierter Salpetersäure versetzt. Das Polypropylenröhrchen wurde
einschließend kräftig geschüttelt und für kurze Zeit ruhig gestellt, bevor 1900 μL Milli-
Q®-Wasser zugegeben und das Polypropylenröhrchen erneut 3 Minuten lang
geschüttelt wurde. Die obere wässrige Phase wurde schließlich mit einer Pipette
sorgfältig abgezogen und in einem Polypropylenröhrchen aufbewahrt.
Um die Effizienz des Verfahrens zu überprüfen, wurden einige Süßwasserproben
direkt vermessen. Ein Vergleich der Ergebnisse mit flüssig-flüssig-extrahierten
Proben wies nur auf geringfügige Abweichungen hin (< 5%).
2.7. Mikrowellen-Aufschluss-Verfahren
Das Mikrowellen-Aufschluss-Verfahren kommt in der analytischen Chemie und
Umweltchemie häufig zum Einsatz (z.B. Zlotorzynski, 1995), da es in der
Spurenmetallanalyse eine schnelle und effiziente Methode zur Überführung der
Festproben in die lösliche Phase darstellt (z.B. Sandroni et al., 2003). Dieses
Verfahren wird in die Gruppe der „Nassaufschlussmethoden unter Verwendung
Kapitel 2: Material und Methoden
- 29 -
lösender und/oder oxidierender Aufschlussreagenzien im geschlossenen
System“ einordnet, die dadurch die Anwendung höherer Temperaturen erlauben
(Stoeppler, 1994).
Tabelle 1. Mikrowellen-Programme einschließlich der Parameter.
Mikrowellen-Programm Zeit maximale Temperatur
maximale Leistung
Filterzersetzen mit Zusatz von 5 mL HNO3 (65%) 30 Min 60 °C 250 W
Druckaufschluss mit Zusatz von 5 mL HNO3 (65%), 1 mL HF (40%), 0.5 mL HCl (30%) und
0.5 mL HClO4 (70%)
5 Min 60 °C 250 W
1 Min 60 °C 0 W
5 Min 90 °C 300 W
5 Min 90 °C 400 W
4 Min 90 °C 600 W
3 Min 90 °C 400 W
10 Min 90 °C 350 W
erstes Abrauchen 30 Min 100 °C 300 W
zweites Abrauchen mit Zusatz von 3 mL HNO3 (65%) 30 Min 100 °C 300 W
Lösen mit Zusatz von 5.00 mL HNO3 (0.5 mol/L) 15 Min 65 °C 320 W
Reinigung mit Zusatz von 5 mL HNO3 (65%), 1 mL HF (40%) und 1 mL HClO4 (70%)
5 Min 80 °C 300 W
1 Min 80 °C 0 W
2 Min 80 °C 400 W
1 Min 100 °C 0 W
2 Min 80 °C 500 W
Im dieser Arbeit wurde die SPM-Probe einschließlich Filter nach Schüßler et al.
(2005) zunächst gewogen und dann in einem Teflon-Gefäß mit entsprechendem
Mikrowellen-Programm aufgeschlossen. Die Parameter des Mikrowellen-Programms
sind in Tab. 1 aufgelistet. Sobald die eingestellte maximale Temperatur im Gefäß
erreicht wurde, wurde die Mikrowellen-Leistung heruntergefahren, um die
Temperatur konstant zu halten. Am Ende wurde die Probe in genau 5.00 mL
Salpetersäure (0.5 mol/L) aufgelöst und war damit bereit für die Analyse
Kapitel 2: Material und Methoden
- 30 -
2.8. Massenspektrometrie mit HR-ICP-MS
Die hochauflösende Massenspektrometrie mit induktiv gekoppeltem Plasma (high-
resolution inductively-coupled-plasma mass-spectrometry; HR-ICP-MS) ist eine
relativ neue instrumentelle Technik zur Untersuchung von Spurenelementen in
Umweltmatrices (z.B. Liang et al., 2000). Dabei wird die Probe unter Einsatz von
Argon durch einen hochfrequenten Strom auf 5000 °C bis 10000 °C erhitzt bzw.
ionisiert. Die dabei entstandenen Ionen werden massenspektrometrisch getrennt und
gemessen. Mit dieser Technik ist man in der Lage, Spurenmetallanalysen schnell
und präzise durchzuführen (z.B. Jenner et al., 1990). Im Rahmen dieser Arbeit wurde
ein HR-ICP-MS-Gerät (Element 2, Thermo Scientific) verwendet. Rhodium wurde als
interner Standard eingesetzt. Vor jeder Messung wurde zuerst ein Tuning des ICP-
MS durchgeführt. Zum Einsatz gekommen sind dabei eine Gasflussrate von ca. 16
L/min für cool gas, ca. 1 L/min für sample gas und ca. 1 L/min für auxiliary gas sowie
eine Generatorleistung von 1200 W. Anschließend wurde das Mass-Offset bestimmt.
Die Konzentrationen von Eisen, Nickel, Cobalt, Kupfer, Cadmium und Blei wurden
mittels externer Kalibrierung quantifiziert. Für die externe Kalibrierung wurden 7
Kalibrationspunkte verwendet. Mit einem Bestimmtheitsmaß R2 ≥ 0.999995 wurden
die Probekonzentrationen anhand der Kalibrationsgleichung berechnet. Überschritt
eine gemessene Konzentration den Kalibrationsbereich, erfolgte eine erneute
Bestimmung nach einer Verdünnung der Probe. Innerhalb einer Messung wurden
routinemäßig mehrfach die Konzentrationen der Blanks (0.5 mol/L HNO3) vermessen.
Bei der Berechnung wurden die Mittelwerte von den Blanks für die jeweiligen Metalle
ermittelt und von den Probekonzentrationen abgezogen.
2.9. Referenzmaterialien
Spurenmetalle sind in der Hydrosphäre nur in geringen Mengen vorhanden (z.B.
Chester, 1990). Bezogen auf diese Arbeit schwankten die Konzentrationen in Lösung
meist im Bereich von nmol/L. Entsprechend musste großer Wert auf saubere und
präzise Arbeit gelegt werden. Mögliche Kontaminationsquellen wurden bereits
mehrfach erwähnt. Gegenmaßnahmen, wie z.B. die sorgfältige Probenahme, die
Aufarbeitung, die gründliche Reinigung von Chemikalien und Gefäßen sowie das
Arbeiten im Reinraum und unter einer Clean-Bench, sollten kontaminationsarme
Ergebnisse ermöglichen.
Wenn die Konzentration eines bestimmten Analyts sich der Konzentration Null nähert,
droht das Signal im Grundrauschen zu verschwinden. Man ist ständig konfrontiert mit
Kapitel 2: Material und Methoden
- 31 -
der Frage, ob solche Signale auswertbar sind. In der analytischen Chemie wurde der
Begriff „Nachweisgrenze (detection limit)“ eingeführt, um die Aussagekraft solcher
Signale quantitativ zu bewerten (z.B. Willard et al., 1988). Häufig ist die
Nachweisgrenze NG gegeben durch:
σ3+= XNG
wobei X für den durchschnittlichen Blank-Wert und σ für die relative
Standardabweichung steht (z.B. Schwedt, 2008). Die Messwerte, die unterhalb der
Nachweisgrenze liegen, weisen statistisch betrachtet eine große Ungenauigkeit auf,
und werden als „nicht nachweisbar“ bezeichnet.
Für jede Messung durch ICP-MS wurde eine instrumentelle Nachweisgrenze für alle
sechs untersuchten Metalle ermittelt. Im Vergleich zu den Probekonzentrationen
waren die Nachweisgrenzen so klein, dass bis auf einige wenige Kolloid-
Konzentrationen für Cadmium und Eisen alle Metallkonzentrationen deutlich darüber
lagen. Für die Berechnungen der kolloidalen Konzentrationen wurden die Werte
unterhalb der Nachweisgrenze auf Null gesetzt.
Referenzmaterialien finden in der analytischen Chemie regelmäßig Verwendung, z.B.
um eine Messmethode zu beurteilen oder um Apparaturen zu kalibrieren. Mit Hilfe
solcher Referenzmaterialien können die Einflüsse der Matrix auf das analytische
Signal eingeschätzt werden. Zum Einsatz gekommen sind in dieser Arbeit die von
NRCC (National Research Council of Canada) zertifizierten Referenzmaterialien.
SLRS-5 (Flusswasser), SLEW-3 (ästuarines Wasser) sowie CASS-4 und NASS-5
(Meerwasser) wurden für den UV-Aufschluss bzw. für die Flüssig-Flüssig-Extraktion
verwendet. Für den Mikrowellen-Aufschluss kamen MESS-1 und MAG-1 zum Einsatz.
Die Metallkonzentrationen für alle Referenzmaterialien wurden mindestens dreifach
gemessen. Alle Messergebnisse lagen im zertifizierten Wertebereich.
Für die Ultrafiltration gibt es keine Referenzmaterialien. Die Berechnung von
Massenbilanzen (Vergleich der direkt gemessenen Lösungs-Gesamtkonzentration in
einer Probe mit der Summe der Konzentrationen der einzelnen Kolloidfraktionen)
kann jedoch Aufschluss geben über die Effizienz der Prozeduren. Bei Annahme einer
fehlerfreien Gesamtmessung ist eine Massenbilanz kleiner als 100% wahrscheinlich
auf einen Verlust der entsprechenden Metalle durch Adsorption an den Membranfilter
oder an die Oberfläche des Probengefäßes zurückzuführen, während bei einer
Massenbilanz > 100% eine Kontamination oder analytische Schwankungen bei den
Kolloid-Messungen zu vermuten sind (z.B. Martin et al., 1995; Powell et al., 1996).
Kapitel 2: Material und Methoden
- 32 -
Tabelle 2. Massenbilanzen der Metalle (N: die gesamte Anzahl der Proben; n: die verwendete Anzahl der Proben).
Studiengebiet Fe Ni Co Cu Cd Pb
Ost-Hainan 75% ± 30%
80% ± 20%
100% ± 30%
80% ± 20%
75% ± 30%
75% ± 30%
N=20 n=12 n=12 n=17 n=18 n=15 n=11
Nordost-Brasilien (Regensaison 2007)
80% ± 40%
110% ± 40%
100% ± 30%
110% ± 30%
90% ± 40%
70% ± 40%
N=15 n=7 n=5 n=9 n=10 n=9 n=8
Nordost-Brasilien (Trockensaison
2009)
100% ± 50%
100% ± 30%
100% ± 20%
100% ± 30%
100% ± 40%
100% ± 50%
N=16 n=7 n=15 n=12 n=14 n=11 n=7
Da die ermittelten Massenbilanzen stark variierten, wurden Niveaus und
Bereichsgrenzen festgelegt (Tab. 2). Nur die Proben, deren Massenbilanz innerhalb
der Bereichsgrenzen lag, wurden weiter verwendet. Obwohl bei den beiden
Projekten nach gleichen Kriterien und mit identischen Materialien gearbeitet wurde,
sind unterschiedliche Ergebnisse herausgekommen. Bei dem LANCET Projekt waren
die Massenbilanzen meist kleiner als 100%. Gleichzeitig wiesen die Massenbilanzen
aus dem POLCAMAR Projekt hohe Wiederfindungsraten auf. Die dafür eingesetzten
Filtrationskartuschen waren zwar von der gleichen Firma, doch wurden sie für die
jeweilige Feldarbeit separat bestellt. Qualitätsunterschiede der Versuchsmaterialien
könnten die unterschiedlichen Ergebnisse verursacht haben. Allerdings traten beim
POLCAMAR Projekt während der Regensaison 2007 teilweise erhöhte
Massenbilanzen auf. Neben der Fortpflanzung analytischer Fehler, die bei der
Berechnung kolloidaler Fraktionen eine Rolle spielen (siehe oben), sind
Kontaminationen trotz aller getroffenen Maßnahmen nicht auszuschließen.
2.10. Andere Parameter
Die Parameter wie z.B. die Salinität, der pH-Wert und die Wassertemperatur wurden
bei den Probenahmen mit einer Multisensor-Einheit (MultiLine® P3, WTW) vor Ort
vermessen. Die Bestimmung von DOC erfolgte durch eine katalytische
Hochtemperatur-Oxidation (Total Organic Carbon Analyzer; Apollo 9000; Tecmar)
mit einer Nachweisgrenze von 0.1 mg/L.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 33 -
3. Vergleichende Diskussion der Ergebnisse
Das physikalische Zusammenmischen von Süßwasser und Salzwasser ist einer der
grundlegenden Prozesse in Ästuaren, die die fluvialen chemischen Signale
modifizieren (Chester, 1990). Bei Abwesenheit jeglicher biologischen, geologischen,
chemischen und anthropogenen Prozesse würde die Konzentration einer bestimmten
chemischen Komponente in Ästuaren entlang des Salinitätsgradients linear verlaufen.
In der Realität treten solche Prozesse zumeist auf; Auswirkungen können nicht
ausgeschlossen werden. Wird Linearität beobachtet, sind Einwirkungen gering oder
nicht vorhanden oder sie gleichen sich aus; man spricht von einem konservativen
Mischen. Eine Abweichung der Konzentrationsverteilung von der idealen
Verdünnungslinie wird als non-konservatives Verhalten bezeichnet. Dabei kann eine
Verteilungskurve sowohl über (positive Abweichung) als auch unter (negative
Abweichung) der idealen Verdünnungslinie verlaufen, was einen Eintrag in die
wässrige Phase bzw. eine Entfernung aus der wässrigen Phase bedeutet. Ein
Eintrag kommt zustande, wenn eine Komponente aus der partikulären Phase (SPM
und/oder Sediment) sich zum Teil löst und ins Wasser übertritt, während bei einer
Entfernung die gelöste Komponente an das SPM adsorbiert wird oder Kolloide zu
größeren Partikeln koagulieren. Wegen der Bedeutung von Ästuaren für die
Materialzufuhr zum Ozean hat sich eine Reihe von wissenschaftlichen
Veröffentlichungen mit dem ästuarinen Verhalten von Spurenmetallen
auseinandergesetzt. In Tab. 3 sind einige wichtige Ergebnisse aufgelistet. Sowohl
Eintrag in die wässrige Phase als auch Entfernung aus der wässrigen Phase können
bei allen genannten Metallen festgestellt werden. Außer bei Cobalt wurde
konservatives Verhalten bei allen untersuchten Metallen beobachtet. In manchen
Ästuaren wie z.B. dem Changjiang Ästuar (Wang and Liu, 2003) und der Galveston
Bucht (Wen et al., 1999) sind sogar saisonale Unterschiede beim
Spurenmetallverhalten erkennbar. Diese Vielfalt im Verhalten der Metalle ist das
Resultat einer Verflechtung von ästuarinen Variablen, und deutet auf die Komplexität
der ästuarinen Chemie hin, die sich im Wanquan Ästuar (im Folgenden: WR) und im
Wenchang-Wenjiao Ästuar (im Folgenden: WWR) an der Ost-Küste Hainans bzw. in
der Mundaú-Lagune (im Folgenden: MD) und in der Manguaba-Lagune (im
Folgenden: MG) Nordostbrasiliens widerspiegelt. In den nachfolgenden Kapiteln wird
das Verhalten der sechs verschmutzungsrelevanten Metalle Eisen, Nickel, Cobalt,
Kupfer, Cadmium und Blei in den tropischen Studiengebieten WR, WWR, MD und
MG diskutiert.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 34 -
Tabelle 3. Ästuarines Verhalten von gelösten Spurenmetallen (P: Positive Abweichung von der idealen Verdünnungslinie; N: Negative Abweichung von der idealen Verdünnungslinie; K: Konservatives Verhalten; --: nicht untersucht).
Studiengebiet Fe Ni Co Cu Cd Pb Literatur
Amazonas Ästuar -- K -- K P -- Boyle et al., 1982
Bristol Kanal und outer Severn Ästuar -- -- -- K K -- Harper, 1991
Changjiang Ästuar N P N P P N Wang and Liu, 2003
Changjiang Ästuar -- -- -- K P K Wang et al., 2009
Danshuei Ästuar -- P P P P -- Fang and Lin, 2002
Delaware Ästuar N N N N N -- Sharp et al., 1982
Donau Ästuar K K P K P -- Guieu et al., 1998
Forth Ästuar N P -- P K N Balls et al., 1994
Galveston Bucht Juli 1993 -- K -- K -- N Wen et al., 1999
Galveston Bucht Mai 1994 -- P -- P -- N Wen et al., 1999
Galveston Bucht Juli 1995 N P P P P N Wen et al., 1999
Gironde Ästuar N P -- P P K Kraepiel et al., 1997
Göta Ästuar N K -- K N N Danielsson et al., 1983
Humber Ästuar -- P -- P P -- Comber et al., 1995
Loire Ästuar -- -- -- P P -- Waeles et al., 2004
Mersey Ästuar -- P P K P P Martino et al., 2002
Mississippi Ästuar K K -- K P -- Shiller and Boyle, 1991
Nerbioi-Ibaizabal Ästuar P K -- K -- -- Fernández et al., 2008
Ob Ästuar N K -- K P N Dai and Martin, 1995
Ochlockonee Ästuar N K -- K K -- Powell et al, 1996
Penzé Ästuar -- -- -- P P N Waeles et al., 2008a
Rhône Ästuar -- K -- K P K Elbaz-Poulichet et al., 1996
Scheldt Ästuar -- P P P P -- Chaudry and Zwolsman, 2008
Seine Ästuar -- P P P P N Chiffoleau et al., 1994
Venedig Lagune -- K -- N P K Martin et al., 1995
Weser Ästuar N P -- P -- N/P Turner et al., 1992
Yenisey Ästuar N P -- K P N Dai and Martin, 1995
3.1. Eisen
Dass Eisen in Ästuaren bei niedrigen Salinitäten ausflockt, ist seit langem etabliert
(z.B. Boyle et al., 1977; Sharp et al., 1982; Danielsson et al., 1983; Turner et al.,
1992; Balls et al., 1994; Dai and Martin, 1995; Powell et al, 1996; Kraepiel et al.,
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 35 -
1997; Wen et al., 1999; Wang and Liu, 2003). Ein signifikanter Anteil (> 80%) des in
den Flüssen transportierten gelösten Eisens wird dadurch aus der wässrigen Phase
entfernt und erreicht nicht den Ozean (z.B. Sholkovitz et al., 1978). Lediglich im
Mississippi Ästuar (Shiller and Boyle, 1991) und im Donau Ästuar (Guieu et al., 1998),
wo gelöstes Eisen im Fluss-Endglied extrem niedrig ist (20-40 nmol/L), ist dieser
Prozess nicht deutlich zu erkennen. Interessanterweise wurde im Nerbioi-Ibaizabal
Ästuar in Spanien eine Erhöhung des gelösten Eisens mit steigender Salinität
festgestellt, wo gelöstes Eisen vergleichsweise sehr hoch ist (1.5-7.6 μmol/L;
Fernández et al., 2008). Die anoxischen Bedingungen ermöglichen vermutlich die
Freisetzung aus partikulärem Eisen (z.B. Pakhomova et al., 2007).
WR
0
5
10
15
0 10 20 30Salinity
Fe (μ
mol
/L)
200620072008
WWR
0
5
10
15
0 10 20 30Salinity
Fe (μ
mol
/L)
MD
0
5
10
15
0 10 20 30Salinity
Fe (μ
mol
/L)
200720082009
MG
0
5
10
15
0 10 20 30Salinity
Fe (μ
mol
/L)
Abbildung 6. Salinitäts-Abhängigkeit des gelösten Eisens.
Mit bis zu 14 μmol/L an der Ost-Küste Hainans bzw. bis zu 6 μmol/L in Nordost-
Brasilien ist die Konzentration des gelösten Eisens in den Studiengebieten ebenfalls
sehr hoch. Eine Anreicherung gelösten Eisens ist jedoch nicht sichtbar. Wie in Abb. 6
dargestellt, sinkt die Konzentration gelösten Eisens bei niedriger Salinität schnell auf
ein Niveau von 1-2 μmol/L, und weiterhin auf unterhalb von 0.5 μmol/L bei hohen
Salinitäten. Gleichzeitig sinkt die partikuläre Konzentration von Eisen von
typischerweise 1.5 mmol/g (schnell v. a. von S = 0 bis S = 10) auf 0.5 mmol/g mit
steigender Salinität (Abb. 7). Die Analyse der Kolloidfraktionen weist darauf hin, dass
bei niedrigen bis mittleren Salinitäten über 80% des gelösten Eisens aus HMW-
Kolloiden besteht. Bei hohen Salinitäten nimmt der HMW-Anteil mehr oder weniger
ab, wobei der Anteil der Eisenkolloide in der LMW-Fraktion und der TD-Fraktion
zunimmt (Abb. 8). Es liegt wahrscheinlich zum einen an der Salz-induzierten
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 36 -
Aggregation (z.B. Nowostawska et al., 2008), durch die die HMW-Fraktion bevorzugt
aus der wässrigen Phase entfernt wird, zum anderen an der möglichen Dissoziation
vom HMW-Eisen wie z.B. Eisenhuminstoffen (z.B. Batchelli et al., 2010).
WR
0.0
0.5
1.0
1.5
2.0
2.5
0 10 20 30Salinity
Fe (m
mol
/g)
200620072008
WWR
0.0
0.5
1.0
1.5
2.0
2.5
0 10 20 30Salinity
Fe (m
mol
/g)
MD
0.0
0.5
1.0
1.5
2.0
2.5
0 10 20 30Salinity
Fe (m
mol
/g)
200720082009
MG
0.0
0.5
1.0
1.5
2.0
2.5
0 10 20 30Salinity
Fe (m
mol
/g)
Abbildung 7. Salinitäts-Abhängigkeit des partikulären Eisens.
WR 2008
0
25
50
75
100
0 10 20 30Salinity
Fe (%
) HMWLMWTD
WWR 2008
0
25
50
75
100
0 10 20 30Salinity
Fe (%
)
MD 2007
0
25
50
75
100
0 10 20 30Salinity
Fe (%
)
HMWLMWTD
MG 2007
0
25
50
75
100
0 10 20 30Salinity
Fe (%
)
MD 2009
0
25
50
75
100
0 10 20 30Salinity
Fe (%
)
HMWLMWTD
MG 2009
0
25
50
75
100
0 10 20 30Salinity
Fe (%
)
Abbildung 8. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Eisen.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 37 -
3.2. Nickel
Nickel ist relativ unreaktiv und verhält sich in Ästuaren zumeist konservativ (z.B.
Boyle et al., 1982; Danielsson et al., 1983; Shiller and Boyle, 1991; Dai and Martin,
1995; Martin et al., 1995; Elbaz-Poulichet et al., 1996; Powell et al, 1996; Guieu et al.,
1998; Wen et al., 1999; Fernández et al., 2008). Dennoch konnten sowohl positive
(z.B. Campbell et al., 1988; Turner et al., 1992; Balls et al., 1994; Comber et al.,
1995; Dai and Martin, 1995; Kraepiel et al., 1997; Wen et al., 1999; Fang and Lin,
2002; Martino et al., 2002; Wang and Liu, 2003; Chaudry and Zwolsman, 2008) als
auch negative (z.B. Sharp et al., 1982; Turner et al., 1998) Abweichungen von der
theoretischen Verdünnungslinie beobachtet werden. Ein Überschuss des gelösten
Nickels kann sowohl aus dem Sediment (z.B. Martino et al., 2004) als auch aus dem
SPM (z.B. Edmond et al., 1985) stammen. Kinetikstudien bestätigen auch, dass die
Dissoziation des Organonickels durch Konkurrenz zwischen Calcium-/Magnesium-
und Nickelionen um die Bindungsstellen an organischen Materialien begünstigt wird
(z.B. Mandal et al., 2002). Andererseits kann das in die Lösung freigesetzte Nickel
durch die Präsenz von organischen Molekülen wie z.B. Ethylendiamintetraacetat
(Bedsworth and Sedlak, 1999) stabilisiert und die Adsorption von Nickel dadurch
effektiv reduziert werden (Turner et al., 1998; Martino et al., 2003). Außerdem kann
Nickel auch mit Huminstoffen und Eisenoxiden koagulieren (z.B. Sholkovitz, 1978),
oder beispielsweise durch Bioakkumulation aus der wässrigen Phase entfernen
werden (z.B. Hong et al., 2009).
WR
0
5
10
15
20
0 10 20 30Salinity
Ni (
nmol
/L)
200620072008
WWR
0
5
10
15
20
0 10 20 30Salinity
Ni (
nmol
/L)
MD
0
5
10
15
20
0 10 20 30Salinity
Ni (
nmol
/L)
200720082009
MG
0
5
10
15
20
0 10 20 30Salinity
Ni (
nmol
/L)
Abbildung 9. Salinitäts-Abhängigkeit des gelösten Nickels.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 38 -
WR
0.0
0.5
1.0
1.5
2.0
0 10 20 30Salinity
Ni (μm
ol/g
)
200620072008
WWR
0.0
0.5
1.0
1.5
2.0
0 10 20 30Salinity
Ni (μm
ol/g
)
MD
0.0
0.5
1.0
1.5
2.0
0 10 20 30Salinity
Ni (μm
ol/g
)
200720082009
MG
0.0
0.5
1.0
1.5
2.0
0 10 20 30Salinity
Ni (μm
ol/g
)
Abbildung 10. Salinitäts-Abhängigkeit des partikulären Nickels.
WR 2008
0
25
50
75
100
0 10 20 30Salinity
Ni (
%)
HMWLMWTD
WWR 2008
0
25
50
75
100
0 10 20 30Salinity
Ni (
%)
MD 2007
0
25
50
75
100
0 10 20 30Salinity
Ni (
%)
HMWLMWTD
MG 2007
0
25
50
75
100
0 10 20 30Salinity
Ni (
%)
MD 2009
0
25
50
75
100
0 10 20 30Salinity
Ni (
%)
HMWLMWTD
MG 2009
0
25
50
75
100
0 10 20 30Salinity
Ni (
%)
Abbildung 11. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Nickel.
Trotz Fluktuationen bei niedriger Salinität verhält sich gelöstes Nickel in allen vier
Studiengebieten nahezu konservativ (Abb. 9). Die Konzentrationen bei niedrigen
Salinitäten variieren zwischen 18 nmol/L in WWR und kleiner als 1 nmol/L in
Nordostbrasilien. Bei einer Salinität von S = 10 bis S = 15 vereinigen sich die
Konzentrationen des gelösten Nickels typischerweise auf ein Niveau von etwa 5
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 39 -
nmol/L, und sinken seewärts weiter ab. Die Modifikationen bei niedrigen Salinitäten
können aufgrund der Calcium-/Magnesium-induzierten Auslösung aus dem
partikulären Nickel (z.B. Mandal et al., 2002) oder der Koagulation mit u.a.
Eisenoxiden (z.B. Sholkovitz, 1978) zustande gekommen sein. Das partikuläre Nickel
verhält sich im Wesentlichen ebenfalls konservativ, wobei der Gehalt in Ost-Hainan
mit steigender Salinität abnimmt und in Nordost-Brasilien konstant bleibt (Abb. 10).
Die Unterschiede in den mittleren Niveaus des partikulären Nickels zwischen Hainan
und Nordost-Brasilien sind wohl auf die jeweiligen geologischen Randbedingungen
zurückzuführen. Angesichts der Beobachtung, dass der TD-Anteil mit zunehmender
Salinität steigt, während die HMW- und LMW-Anteile sinken (Abb. 11), und
gleichzeitig die Konzentration des gelösten Nickels sich nahezu konservativ verhält,
kann man vermuten, dass entweder eine interne Konversion von HMW- und LMW-
zur TD-Fraktion aufgetreten ist oder die Dissoziation von Koagulation begleitet wird.
3.3. Cobalt
Auch wenn bei Feldarbeiten (Sharp et al., 1982; Wang and Liu, 2003) und
Mischungsexperimenten (Sholkovitz, 1978) auch negative Abweichung festgestellt
werden konnten, weicht die Verteilungskurve von Cobalt im Ästuar überwiegend
positiv von der theoretischen Verdünnungslinie ab (z.B. Chiffoleau et al., 1994; Guieu
et al., 1998; Wen et al., 1999; Fang and Lin, 2002; Martino et al., 2002; Tovar-
Sánchez et al., 2004; Chaudry and Zwolsman, 2008). Das Verhalten von Cobalt im
Ästuar wird häufig mit Mangan in Verbindung gebracht (z.B. Takara et al., 2010).
Ästuarinen Untersuchungen von z.B. Chaudry and Zwolsman (2008) zufolge findet
unter oxischen Bedingungen eine gemeinsame Sedimentation von Cobalt und
Mangan statt. Cobalt kann in gelöster Form aus dem Sediment freigesetzt werden,
wenn partikuläres Mangan (IV) während der Mineralisierung von organischer
Substanz reduktiv gelöst wird (z.B. Moffett and Ho, 1996; Audry et al., 2006).
Allerdings ist ein gewisser Anteil von gelöstem Cobalt im Ästuar organisch gebunden,
und diese Bindung scheint sehr stabil zu sein (z.B. Zhang et al., 1990).
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 40 -
WR
0
3
6
9
0 10 20 30Salinity
Co
(nm
ol/L
)
200620072008
WWR
0
3
6
9
0 10 20 30Salinity
Co
(nm
ol/L
)
MD
0
3
6
9
0 10 20 30Salinity
Co
(nm
ol/L
)
200720082009
MG
0
3
6
9
0 10 20 30Salinity
Co
(nm
ol/L
)
Abbildung 12. Salinitäts-Abhängigkeit des gelösten Cobalts.
WR
0.00
0.25
0.50
0.75
1.00
0 10 20 30Salinity
Co
(μm
ol/g
)
200620072008
WWR
0.00
0.25
0.50
0.75
1.00
0 10 20 30Salinity
Co
(μm
ol/g
)
MD
0.00
0.25
0.50
0.75
1.00
0 10 20 30Salinity
Co
(μm
ol/g
)
200720082009
MG
0.00
0.25
0.50
0.75
1.00
0 10 20 30Salinity
Co
(μm
ol/g
)
Abbildung 13. Salinitäts-Abhängigkeit des partikulären Cobalts.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 41 -
WR 2008
0
25
50
75
100
0 10 20 30Salinity
Co
(%) HMW
LMWTD
WWR 2008
0
25
50
75
100
0 10 20 30Salinity
Co
(%)
MD 2007
0
25
50
75
100
0 10 20 30Salinity
Co
(%)
HMWLMWTD
MG 2007
0
25
50
75
100
0 10 20 30Salinity
Co
(%)
MD 2009
0
25
50
75
100
0 10 20 30Salinity
Co
(%)
HMWLMWTD
MG 2009
0
25
50
75
100
0 10 20 30Salinity
Co
(%)
Abbildung 14. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Cobalt.
Außer in WWR ist das Signal von gelöstem Cobalt im Flusswasser selten höher als 3
nmol/L. Nach anfänglichen Fluktuationen sinkt es kontinuierlich auf ein Niveau von
weniger als 1 nmol/L bei hohen Salinitäten (Abb. 12). Das gilt auch für das partikuläre
Cobalt, dessen Konzentration typischerweise von 0.5 μmol/g auf 0.2 μmol/g in Ost-
Hainan bzw. von 0.25 μmol/g auf 0.05 μmol/g in Nordost-Brasilien abnimmt (Abb. 13).
Besonderes in der Trockensaison von 2006 im WWR und in Nordost-Brasilien, ist
teilweise eine starke Erhöhung von gelöstem Cobalt bei niedriger Salinität (S < 5)
gegenüber dem Flusswasser zu beobachten. Da der Sauerstoffgehalt nicht
gemessen wurde, kann ein Zusammenhang mit anoxischen Bedingungen an der
Sediment-Wasser-Grenzfläche nicht ermittelt werden. Eventuell wurden die obersten
Sedimentschichten im Süßwasser-Zufuhrbereich aufgewirbelt, wodurch die
Freisetzung von Cobalt aus dem Sediment begünstigt würde. Steigende Anteile von
TD-Fraktion mit steigender Salinität (Abb. 14) liefern Hinweise dafür, dass das
freigesetzte Cobalt hauptsächlich hydratisierte Ionen und kleine lösliche Komplexe
mit Chlorid und Sulfat (z.B. Ćosović et al., 1982) und/oder mit kleineren organischen
Liganden (z.B. Zhang et al., 1990) bildet.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 42 -
3.4. Kupfer
Zahlreiche Studien haben sich mit dem ästuarinen Verhalten des Kupfers beschäftigt.
Demnach kann sich Kupfer sowohl konservativ (z.B. Boyle et al., 1982; Danielsson et
al., 1983; Harper, 1991; Shiller and Boyle, 1991; Dai and Martin, 1995; Elbaz-
Poulichet et al., 1996; Powell et al, 1996; Guieu et al., 1998; Wen et al., 1999;
Martino et al., 2002; Fernández et al., 2008; Wang et al., 2009) als auch non-
konservativ verhalten (z.B. Sharp et al., 1982; Turner et al., 1992; Balls et al., 1994;
Chiffoleau et al., 1994; Comber et al., 1995; Martin et al., 1995; Kraepiel et al., 1997;
Wen et al., 1999; Fang and Lin, 2002; Wang and Liu, 2003; Waeles et al., 2004;
Chaudry and Zwolsman, 2008; Waeles et al., 2008a). Dabei kommt eine Zufuhr zum
gelösten Kupfer meist durch Freisetzung aus dem SPM und/oder dem Sediment
zustande (z.B. Santos-Echeandia et al., 2008), möglicherweise jedoch auch infolge
des Abbaus organischen Substanz (z.B. Masson et al., 2011). Verluste hingegen
sind immer mit Adsorption (z.B. Sholkovitz, 1978) oder biochemischen Aktivitäten
(z.B. Sharp et al., 1982) verbunden. Kolloidales Kupfer besteht überwiegend aus Cu-
Komplexen mit organischen Liganden wie z.B. Huminstoffen (z.B. Kogut and Voelker,
2003). In organisch-reichen Ästuaren ist gelöstes Kupfer sogar fast ausschließlich
gebunden an organische Entitäten (z.B. Shank et al., 2004). Dies liegt zum einen an
der starken Neigung von Kupfer zu organischen Funktionsgruppen (z.B. Newell and
Sanders, 1986), zum anderen kann es auch eine Konsequenz der Primärproduktion
sein (z.B. Van den Berg and De Luca Rebello, 1986).
WR
0
5
10
15
20
0 10 20 30Salinity
Cu
(nm
ol/L
)
200620072008
WWR
0
5
10
15
20
0 10 20 30Salinity
Cu
(nm
ol/L
)
MD
0
5
10
15
20
0 10 20 30Salinity
Cu
(nm
ol/L
)
200720082009
MG
0
5
10
15
20
0 10 20 30Salinity
Cu
(nm
ol/L
)
Abbildung 15. Salinitäts-Abhängigkeit des gelösten Kupfers.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 43 -
WR
0.0
0.5
1.0
1.5
0 10 20 30Salinity
Cu
(μm
ol/g
)
200620072008
WWR
0.0
0.5
1.0
1.5
0 10 20 30Salinity
Cu
(μm
ol/g
)
MD
0.0
0.5
1.0
1.5
0 10 20 30Salinity
Cu
(μm
ol/g
)
200720082009
MG
0.0
0.5
1.0
1.5
0 10 20 30Salinity
Cu
(μm
ol/g
)
Abbildung 16. Salinitäts-Abhängigkeit des partikulären Kupfers.
WR 2008
0
25
50
75
100
0 10 20 30Salinity
Cu
(%) HMW
LMWTD
WWR 2008
0
25
50
75
100
0 10 20 30Salinity
Cu
(%)
MD 2007
0
25
50
75
100
0 10 20 30Salinity
Cu
(%)
HMWLMWTD
MG 2007
0
25
50
75
100
0 10 20 30Salinity
Cu
(%)
MD 2009
0
25
50
75
100
0 10 20 30Salinity
Cu
(%)
HMWLMWTD
MG 2009
0
25
50
75
100
0 10 20 30Salinity
Cu
(%)
Abbildung 17. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Kupfer.
In den vier Studiengebieten nimmt die Konzentration von gelöstem Kupfer
typischerweise von 10-15 nmol/L im Flusswasser auf weniger als 5 nmol/L im
Meerwasser ab (Abb. 15). Ein konservatives Verhalten ist deutlich zu erkennen. Auch
das partikuläre Kupfer verhält sich konservativ mit einer seewärts zumeist ebenfalls
fallenden Konzentration, typischerweise von 0.6 μmol/g auf 0.3 μmol/g in Ost-Hainan
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 44 -
bzw. mit relativ konstanter Konzentration von ca. 0.3 μmol/g in Nordost-Brasilien
(Abb. 16). Dabei fällt der Kupfergehalt auf Hainan in 2008 deutlich höher aus als
sonst üblich. Aus den Kolloid-Verteilungen ist jedoch ersichtlich, dass eine ästuarine
Modifikation trotzdem stattfindet: während die HMW- und LMW-Fraktionen
schrumpfen, steigt der Anteil der TD-Fraktion (schnell bei niedrigen Salinitäten und
langsam bei mittleren bis hohen Salinitäten) auf ein Niveau von 65% im
Küstenwasser an - unabhängig von der Natur des Flusswassers (Abb. 17).
Möglicherweise werden große kolloidale organische Kupfermoleküle zu kleineren
organischen Kupfer-Komplexen abgebaut als Konsequenz der Veränderung der
Ionenstärke (Waeles et al., 2004).
3.5. Cadmium
Manchmal wird konservatives ästuarines Verhalten von Cadmium behauptet; jedoch
scheint dabei entweder die Meerwasser-Endglied-Konzentration nicht berücksichtigt
worden zu sein (z.B. Balls et al., 1994) oder es wurde nur ein Ausschnitt von
mittleren bis zu hohen Salinitäten dargestellt (z.B. Powell et al, 1996). Die meisten
ästuarinen Untersuchungen zeigen eine Anreicherung von gelöstem Cadmium im
Verlauf der ästuarinen Mischung (z.B. Boyle et al., 1982; Elbaz-Poulichet et al., 1987;
Shiller and Boyle, 1991; Chiffoleau et al., 1994; Comber et al., 1995; Dai and Martin,
1995; Martin et al., 1995; Elbaz-Poulichet et al., 1996; Kraepiel et al., 1997; Guieu et
al., 1998; Wen et al., 1999; Fang and Lin, 2002; Martino et al., 2002; Wang and Liu,
2003; Waeles et al., 2004; Chaudry and Zwolsman, 2008; Waeles et al., 2008a;
Wang et al., 2009). Dabei ist die Mobilisierung von SPM- und/oder Sediment-
gebundenem Cadmium von großer Bedeutung (z.B. Comans and van Dijk, 1988).
Der Transfer von Cadmium in die Lösungsphase wird vor allem durch
Komplexbildung mit Chloridionen begünstigt (z.B. Dabrin et al., 2009). Darüber
hinaus zeigt die Speziesanalyse, dass Cadmium sowohl mit Chloridionen als auch
mit organischen Liganden stabile Komplexe bildet (z.B. Waeles et al., 2005). In
einigen Fällen ist aber auch Sedimentation (d.h. Entfernung aus der Lösungsphase)
bei niedriger Salinität nachgewiesen worden, wie z.B. auf der subtropischen Insel
Taiwan (Jiann et al., 2005), im kalten Schweden (Danielsson et al., 1983) und in den
moderat klimatischen Ästuaren des Mersey (Comber et al., 1995) und des Delaware
(Sharp et al., 1982). Obwohl sich ein solches Verhalten in den meisten Ästuaren
nicht deutlich bemerkbar macht, scheint es doch ein globales Phänomen zu sein. In
der Tat deuten Mischungsexperimente daraufhin, dass Cadmium durchaus
sedimentieren kann (Sholkovitz, 1978), und dass dies vor allem durch Adsorbieren
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 45 -
an die Oberfläche von Mangan- und/oder Eisenoxiden geschieht (z.B. Yang and
Sañudo-Wilhelmy, 1998; Turner et al., 2008).
WR
0.00
0.05
0.10
0.15
0 10 20 30Salinity
Cd
(nm
ol/L
)
200620072008
WWR
0.00
0.05
0.10
0.15
0 10 20 30Salinity
Cd
(nm
ol/L
)
MD
0.00
0.05
0.10
0.15
0 10 20 30Salinity
Cd
(nm
ol/L
)
200720082009
MG
0.00
0.05
0.10
0.15
0 10 20 30Salinity
Cd
(nm
ol/L
)
Abbildung 18. Salinitäts-Abhängigkeit des gelösten Cadmiums.
WR
0
4
8
12
0 10 20 30Salinity
Cd
(nm
ol/g
)
200620072008
WWR
0
4
8
12
0 10 20 30Salinity
Cd
(nm
ol/g
)
MD
0
4
8
12
0 10 20 30Salinity
Cd
(nm
ol/g
)
200720082009
MG
0
4
8
12
0 10 20 30Salinity
Cd
(nm
ol/g
)
Abbildung 19. Salinitäts-Abhängigkeit des partikulären Cadmiums.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 46 -
WR 2008
0
25
50
75
100
0 10 20 30Salinity
Cd
(%) HMW
LMWTD
WWR 2008
0
25
50
75
100
0 10 20 30Salinity
Cd
(%)
MD 2007
0
25
50
75
100
0 10 20 30Salinity
Cd
(%)
HMWLMWTD
MG 2007
0
25
50
75
100
0 10 20 30Salinity
Cd
(%)
MD 2009
0
25
50
75
100
0 10 20 30Salinity
Cd
(%)
HMWLMWTD
MG 2009
0
25
50
75
100
0 10 20 30Salinity
Cd
(%)
Abbildung 20. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Cadmium.
Die ästuarine Verteilung des gelösten Cadmiums in allen vier Studiengebieten ist in
Abb. 18 aufgetragen. In den meisten Fällen steigt die Konzentration mit steigender
Salinität von 0.01-0.05 nmol/L auf typischerweise 0.1 nmol/L im Küstenbereich. Eine
Mobilisierung aus dem partikulären Cadmium scheint aufzutreten, wobei dieser Effekt
in Ost-Hainan ausgeprägter ist als in Nordost-Brasilien. In der MG in 2007 z.B. bleibt
die Konzentration bei etwa 0.1 nmol/L stabil. In der MD vermindert sich die
Konzentration von gelöstem Cadmium von 0.1 nmol/L im Flusswasser sogar um zwei
Drittel auf 0.03 nmol/L bei niedrigen und mittleren Salinitäten. Auch in 2008 und 2009
ist auf ersten Blick nur eine mit der Salinität steigende konservative
Konzentrationskurve zu sehen. Wenn die weitaus kleinere Konzentration an der
Küste (in 2009; 0.03 nmol/L) im Betrag gezogen wird, ist eine Mobilisierung aus dem
partikulären Cadmium jedoch durchaus möglich, wie es im Huanghe (gelben Fluss)
Ästuar auch beobachtet worden ist (Elbaz-Poulichet et al., 1987). Das partikuläre
Cadmium verhält sich in allen vier Studiengebieten im Wesentlichen konservativ,
wobei zum Teil eine starke Abnahme bei niedriger Salinität vor allem in Ost-Hainan
zu beobachten ist (Abb. 19). Die Adsorption u.a. an Eisenoxide gefolgt von
gemeinsamer Ausfällung wird hier vermutlich eine wichtige Rolle gespielt haben. Die
Analysen der Kolloidfraktionen ergeben, dass bei niedriger Salinität die Anteile der
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 47 -
Kolloide (HMW und LMW) deutlich abnehmen (Abb. 20), was eventuell auf
zunehmende Chloro-Komplexierung des Cadmium zurückzuführen ist.
3.6. Blei
Im Allgemeinen wird Blei im Verlauf der ästuarinen Vermischung aus der wässrigen
Phase entfernt (z.B. Danielsson et al., 1983; Turner et al., 1992; Balls et al., 1994;
Chiffoleau et al., 1994; Dai and Martin, 1995; Wen et al., 1999; Wang and Liu, 2003;
Monbet, 2006; Waeles et al., 2008a). Allerdings ist konservatives Verhalten auch zu
beobachten (z.B. Harper, 1991; Martin et al., 1995; Elbaz-Poulichet et al., 1996;
Kraepiel et al., 1997; Wang et al., 2009). Zudem wird über Konzentrationsmaxima
von gelöstem Blei in der Turbiditäts-Maximum-Zone infolge einer Freisetzung von
Blei aus resuspendierten Sedimentpartikeln berichtet (Turner et al., 1992; Martino et
al., 2002); auch Stabilisierung durch Komplexbildung mit reduzierten Schwefel-
Spezies soll eine Rolle spielen (Tang et al., 2002). Dennoch kontrolliert in den
meisten Fällen die Adsorption an das SPM oder die Ausflockung mit Eisenkolloiden
die Ästuarchemie von Blei (z.B. Elbaz-Poulichet et al., 1984; Waeles et al., 2008b).
WR
0
1
2
3
0 10 20 30Salinity
Pb (n
mol
/L)
200620072008
WWR
0
1
2
3
0 10 20 30Salinity
Pb (n
mol
/L)
MD
0
1
2
3
0 10 20 30Salinity
Pb (n
mol
/L)
200720082009
MG
0
1
2
3
0 10 20 30Salinity
Pb (n
mol
/L)
Abbildung 21. Salinitäts-Abhängigkeit des gelösten Bleis.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 48 -
WR
0.0
0.1
0.2
0.3
0.4
0 10 20 30Salinity
Pb (μ
mol
/g)
200620072008
WWR
0.0
0.1
0.2
0.3
0.4
0 10 20 30Salinity
Pb (μ
mol
/g)
MD
0.0
0.1
0.2
0.3
0.4
0 10 20 30Salinity
Pb (μ
mol
/g)
200720082009
MG
0.0
0.1
0.2
0.3
0.4
0 10 20 30Salinity
Pb (μ
mol
/g)
Abbildung 22. Salinitäts-Abhängigkeit des partikulären Bleis.
WR 2008
0
25
50
75
100
0 10 20 30Salinity
Pb (%
) HMWLMWTD
WWR 2008
0
25
50
75
100
0 10 20 30Salinity
Pb (%
)
MD 2007
0
25
50
75
100
0 10 20 30Salinity
Pb (%
)
HMWLMWTD
MG 2007
0
25
50
75
100
0 10 20 30Salinity
Pb (%
)
MD 2009
0
25
50
75
100
0 10 20 30Salinity
Pb (%
)
HMWLMWTD
MG 2009
0
25
50
75
100
0 10 20 30Salinity
Pb (%
)
Abbildung 23. Salinitäts-Abhängigkeit der kolloidalen Verteilung von Blei.
Trotz einiger Ausreißer sinkt generell der Gehalt von gelöstem Blei in den beiden
vergleichsweise bleireichen Ästuaren Hainans (bis zu 2.2 nmol/L), und zwar um mehr
als 50% bereits bei niedriger Salinität (S < 10); eine solche Abnahme ist im
bleiarmen ästuarinen Wasser in Nordost-Brasilien (< 1 nmol/L) nicht klar erkennbar
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 49 -
(Abb. 21). Ab S > 10 sinkt die gelöste Bleikonzentration nur noch wenig bis auf ein
Niveau von ca. 0.5 nmol/L im Küstenwasser. Partikuläres Blei folgt im Allgemeinen
der konservativen Mischungslinie (Abb. 22). Die Erhöhungen bei höherer Salinität in
WWR könnten durch die Aufwirbelung von Blei-haltigen Sedimenten hervorgerufen
worden sein. Dass die kolloidale Verteilung von gelöstem Blei der von Eisen ähnelt
(Abb. 23), legt eine Affinität von Blei zu Eisenkolloiden nahe. Offensichtlich ist Blei im
Salinitätsbereich S < 20 primär mit der HMW-Fraktion assoziiert. Jedoch steigt auch
beim Blei die relative Bedeutung des "echt gelösten" Pb (TD-Fraktion) mit
zunehmendem Salzgehalt.
3.7. Verteilungskoeffizienten
Der Verteilungskoeffizient (distribution coefficient; partition coefficient), der die
relative Affinität eines Stoffes zur Festphase widerspiegelt (z.B. Chiffoleau et al.,
1994), ist ein Schlüsselparameter zur Interpretation des chemischen Transports in
Ästuaren (Turner et al., 1992). Der Verteilungskoeffizient ist gegeben durch:
CC
Ks
pD =
wobei Cp (in: mol/kg) die partikuläre und CS (in: mol/L) die gelöste Konzentration
darstellen (z.B. Benoit et al., 1994). Cadmium und Kupfer z.B. neigen zum
bevorzugten Auftreten in gelöster Form und haben daher einen kleineren
Verteilungskoeffizient, während Eisen und Blei wegen ihrer geringen Löslichkeit im
Wasser größere Verteilungskoeffizienten aufweisen.
Häufig wird der Verteilungskoeffizient in Abhängigkeit von der Salinität diskutiert (z.B.
Balls et al., 1994). Demnach ist eine Zunahme des Verteilungskoeffizienten mit
steigender Salinität ein Anzeichen für die Dominanz von Metalladsorption und eine
Abnahme für das Überwiegen der Desorption. Eine steigende Kurve in Diagrammen
KD vs. Salinität deutet auf eine zunehmende Entfernung aus der wässrigen Phase im
Ästuar hin und eine fallende Kurve offenbart Eintrag in die wässrige Phase.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 50 -
Fe in East-Hainan
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
WRWWR
Fe in Northeast-Brazil
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
MDMG
Ni in East-Hainan
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
Ni in Northeast-Brazil
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
Co in East-Hainan
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
Co in Northeast-Brazil
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
Cu in East-Hainan
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
Cu in Northeast-Brazil
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
Cd in East-Hainan
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
Cd in Northeast-Brazil
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
Pb in East-Hainan
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
Pb in Northeast-Brazil
3
4
5
6
7
8
0 10 20 30Salinity
LogK
D (L
/kg)
Abbildung 24. Salinitäts-Abhängigkeit des log KD.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 51 -
In Abb. 24 ist ersichtlich, dass Cadmium das einzige Element ist, dessen
Verteilungskoeffizient negativ mit der Salinität korreliert ist. Die Bildung von stabilen
Cadmium-Chloro-Komplexen und die erfolgreiche Konkurrenz von Meerwasser-
kationen mit Cadmiumionen um Sorptionsstellen am SPM ist vermutlich dafür
verantwortlich (z.B. Turner, 1996; Dabrin et al., 2009). Die KD-Werte von Nickel,
Cobalt und Kupfer können annähernd als konstant betrachtet werden, während die
von Eisen und Blei mit steigender Salinität teilweise deutlich (logarithmische Skala)
zunehmen, was als eine Folge von Ausfällungen der beiden Metalle aus der gelösten
Phase gewertet werden kann (z.B. Dai and Martin, 1995).
Vergleiche der Höhe der Verteilungskoeffizienten unterschiedlicher Metalle ergeben
zumeist die folgende Reihung: Fe > Pb > Co > Ni > Cd ≥ Cu (z.B. Balls, 1989; Turner
et al., 1992; Chiffoleau et al., 1994; Munksgaard and Parry, 2001; Tang et al., 2002).
Mit durchschnittlichen log KD-Werten von
Fe : Pb : Co : Ni : Cd : Cu = 6.0 : 5.7 : 5.3 : 5.1 : 4.8 : 4.8 auf Hainan bzw.
Fe : Pb : Co : Ni : Cd : Cu = 6.1 : 5.8 : 5.0 : 4.8 : 4.6 : 4.5 in Nordost-Brasilien
stimmen die Ergebnisse dieser Arbeit im Wesentlichen mit den Literatur-Reihungen
überein (z.B. Balls, 1989; Turner et al., 1992; Balls et al., 1994; Munksgaard and
Parry, 2001; Fang and Lin, 2002; Tang et al., 2002; Guo et al., 2000b; Jiann et al.,
2005). Allerdings muss man sich dabei im Klaren sein, dass je nach den
angewendeten analytischen Methoden die Verteilungskoeffizienten sehr stark
variieren können (z.B. Turner et al., 1992): beispielsweise werden in Hinblick auf die
SPM-Bestimmung sowohl - wie in der vorliegenden Arbeit - die totalen
Metallkonzentrationen (z.B. Zhang, 1999) als auch die - deutlich niedrigeren -
säureverfügbaren Metallkonzentrationen (acid leachable; z.B. Fang and Lin, 2002)
verwendet, um den Verteilungskoeffizient zu berechnen.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 52 -
3.8. Vergleich der Studiengebiete und ihrer Kontaminationsniveaus
Die durchschnittlichen Konzentrationen in den vier Studiengebieten von gelösten und
partikulären Metallen sowie des DOC und des SPM werden in Tab. 4 und Tab. 5
dargestellt.
Tabelle 4. Durchschnittliche Konzentrationen der gelösten Spurenmetalle und des DOC in den vier Studiengebieten.
Studiengebiet Fe Ni Co Cu Cd Pb DOC
μmol/L nmol/L nmol/L nmol/L nmol/L nmol/L mg/L
WR
Trockensaison 2006 0.1 4.8 3.2 17.6 0.06 0.27 1.4
Regensaison 2007 2.6 5.6 1.2 7.5 0.06 0.92 1.7
Regensaison 2008 2.6 4.2 1.1 6.3 0.06 0.71 2.1
gesamt 2.1 4.9 1.6 9.3 0.06 0.70 1.8
WWR
Trockensaison 2006 0.3 8.4 3.3 8.4 0.11 0.09 1.9
Regensaison 2007 1.5 7.7 2.4 7.2 0.07 0.27 2.1
Regensaison 2008 5.2 8.8 3.0 11.5 0.06 0.99 5.6
gesamt 2.8 8.4 2.9 9.4 0.07 0.53 3.8
MD
Regensaison 2007 1.3 5.6 1.0 7.5 0.04 0.16 3.0
Trockensaison 2008 1.3 2.0 2.0 5.6 0.02 0.23 4.3
Trockensaison 2009 1.2 7.7 2.0 7.7 0.03 0.34 5.4
gesamt 1.3 4.9 1.8 6.8 0.03 0.25 4.4
MG
Regensaison 2007 1.7 6.8 0.9 10.7 0.01 0.45 3.8
Trockensaison 2008 1.3 3.7 2.0 6.3 0.04 0.21 4.5
Trockensaison 2009 0.9 7.7 1.9 6.7 0.04 0.45 6.6
gesamt 1.3 5.8 1.6 7.9 0.03 0.35 4.8
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 53 -
Tabelle 5. Durchschnittliche Konzentrationen der partikulären Spurenmetalle und des SPM in den vier Studiengebieten.
Studiengebiet Fe Ni Co Cu Cd Pb SPM
mmol/g μmol/g μmol/g μmol/g nmol/g μmol/g mg/L
WR
Trockensaison 2006 1.02 0.52 0.29 0.30 4.2 0.23 10.9
Regensaison 2007 0.91 0.58 0.29 0.46 2.8 0.21 2.0
Regensaison 2008 0.92 0.71 0.30 0.71 4.4 0.28 3.0
gesamt 0.94 0.61 0.29 0.52 3.7 0.24 4.3
WWR
Trockensaison 2006 1.12 0.91 0.29 0.46 3.2 0.13 9.1
Regensaison 2007 0.86 0.94 0.36 0.35 3.5 0.17 2.7
Regensaison 2008 0.73 0.94 0.33 0.86 3.5 0.17 1.4
gesamt 0.87 0.93 0.33 0.60 3.4 0.16 3.8
MD
Regensaison 2007 0.89 0.27 0.16 0.22 2.3 0.17 20.8
Trockensaison 2008 0.72 0.25 0.16 0.21 0.7 0.12 6.5
Trockensaison 2009 0.67 0.28 0.16 0.21 0.9 0.13 9.1
gesamt 0.75 0.27 0.16 0.21 1.2 0.13 11.1
MG
Regensaison 2007 0.79 0.29 0.15 0.21 0.6 0.13 24.1
Trockensaison 2008 0.57 0.21 0.10 0.18 0.6 0.11 9.0
Trockensaison 2009 0.87 0.22 0.14 0.17 0.8 0.14 20.4
gesamt 0.73 0.24 0.13 0.19 0.7 0.13 17.1
Der DOC-Gehalt in Nordost-Brasilien mit einer durchschnittlichen Konzentration von
4.6 mg/L ist zwar mehr als 50% höher als in Ost-Hainan (2.9 mg/L), unterscheidet
sich jedoch nicht wesentlich von anderen Ästuaren, wie z.B. dem europäischen
Mersey Ästuar (1.2-9.6 mg/L; Martino et al., 2004), dem asiatischen Danshuei Ästuar
(2.6 mg/L; Jiann et al., 2005), dem nordamerikanischen Ochlockonee Ästuar (2.3-
18.8 mg/L; Powell et al, 1996) oder dem südamerikanischen Amazonas Ästuar (0.8-
3.7 mg/L; Sholkovitz et al., 1978).
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 54 -
Die SPM-Beladung in Nordost-Brasilien mit einer durchschnittlichen Konzentration
von 14.0 mg/L ist ebenfalls deutlich größer als in Ost-Hainan (4.0 mg/L). Dennoch
gehören die vier Studiengebiete zu den am wenigsten SPM-beladenen Ästuaren der
Welt (vgl.: Sholkovitz et al., 1978; Munksgaard and Parry, 2001; Martino et al., 2002;
Wang and Liu, 2003; Tovar-Sánchez et al., 2004).
In Ost-Hainan ist der Metallgehalt sowohl in der gelösten als auch in der partikulären
Phase generell höher als in Nordost-Brasilien. Dies könnte zum Teil auf die stärkeren
anthropogenen Aktivitäten auf Hainan zurückgeführt werden, liegt aber vermutlich
primär an der unterschiedlichen geologischen Natur der beiden Regionen. Der
höhere SPM-Gehalt kann in geringem Umfang zu stärkerer Adsorption der Metalle
beitragen (z.B. Turner, 1996). Das WWR weist zumeist die höchsten
Metallkonzentrationen von den vier Studiengebieten auf. Mit Ausnahme von Kupfer
und Cadmium, die zur gelösten Phase neigen, nehmen die Metallkonzentrationen
jedoch durch Prozesse im Ästuar mit steigender Salinität schnell ab. Ansonsten
liegen die Metallkonzentrationen bei hohen Salinitäten in Ost-Hainan und Nordost-
Brasilien etwa auf dem gleichen Niveau.
Erwähnenswert ist noch, dass in der Trockensaison von 2009 ein
Konzentrationsmaximum (S ≈ 10) in der MD bei allen gelösten Metallen mehr oder
weniger deutlich auftritt, was vielleicht auf eine Verschmutzungsquelle aus der
naheliegenden Stadt Maceío hindeutet. Wie zuvor ausgeführt, werden die erhöhten
Konzentrationen durch ästuarine Einwirkungen bald gemildert; die filternde Wirkung
des Ästuars scheint einwandfrei zu funktionieren.
In Tab. 6 und Tab. 7 sind typische Konzentrationen für gelöste bzw. partikuläre
Spurenmetalle in einigen Ästuaren weltweit aufgelistet. Ein Vergleich macht deutlich,
dass die gelösten und die partikulären Spurenmetallenkonzentration auf Hainan und
in Nordost-Brasilien sich nicht viel von denen anderer, nicht-industriell geprägter
Ästuare in der Welt unterscheiden. Vielmehr gehören die gelösten und die
partikulären Konzentrationen aller Spurenmetalle außer Eisen auf Hainan und in
Nordost-Brasilien zu den niedrigsten, die in Tab. 6 und Tab. 7 gelistet sind. Auch im
Vergleich zu den weniger industriell belasteten Ästuaren, wie z.B. den Ästuaren auf
Sumatra (Bach, 2012), fallen die beiden Studiengebiete nicht durch höhere Werte auf.
Was die verschmutzungsrelevanten Spurenmetalle anbelangt, scheint die Aquakultur
auf Hainan und die Zuckerrohrmonokultur in Nordost-Brasilien keine bzw. nur
begrenzte Auswirkungen auf die Umweltkonzentrationen zu haben.
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 55 -
Tabelle 6. Konzentrationen der gelösten Spurenmetalle in Ästuaren.
Studiengebiet Fe Ni Co Cu Cd Pb
Literatur μmol/L nmol/L nmol/L nmol/L nmol/L nmol/L
WR 2.1 4.9 1.6 9.3 0.06 0.70
diese Arbeit WWR 2.8 8.4 2.9 9.4 0.07 0.53
MD 1.3 4.9 1.8 6.8 0.03 0.25
MG 1.3 5.8 1.6 7.9 0.03 0.35
Bristol Kanal und outer Severn Ästuar -- -- -- 26.8-
74.0 0.01-1.25
0.01-48.3 Harper, 1991
Changjiang Ästuar 0.1-0.6 18-28 1.4-2.9 19-26 0.02-0.36 2.3-3.4 Wang and
Liu, 2003
Danshuei Ästuar Oktober 2001
0.01-1.8 138 -- 60 0.2 0.9 Jiann et al.,
2005
Forth Ästuar 1.4 13.6 -- 25 1.1 1.0 Balls et al., 1994
Galveston Bucht 0.01-1.1
4.7-32.3 0.4-9.6 3.5-
18.6 0.03-0.17 0.1-0.8 Wen et al.,
1999
Huanghe Ästuar -- -- -- 1.6-70 0.9-29 1.1-6.4 Tang et al., 2010
Mersey Ästuar -- 10-150 0.1-40 20-90 0.1-4.5 0.1-6 Martino et al., 2002
Mississippi Ästuar 0.03 1-30 -- 1-28 0.01-0.3 -- Shiller and
Boyle, 1991
Nordaustralische Küste und ästuarines
Meerwasser
0.001-0.6 2.0-9.4 0.1-1.6 2.3-16 0.01-
0.31 0.01-0.3
Munksgaard and Parry,
2001
Ob Ästuar 0.4-0.7 20-24 -- 29-38 0.005-0.008
0.06-0.08
Dai and Martin, 1995
Ochlockonee Ästuar 0.04-7.0 2.6-5.5 -- 2.0-5.4 0.04-
0.05 -- Powell et al, 1996
Penzé Ästuar -- -- -- 1.0-4.5 0.1-0.4 0.04-0.30
Waeles et al., 2008a
Rhône Ästuar -- 4-18 -- 3-23 0.1-0.3 -- Dai et al., 1995
Venedig Lagune 0.001-1.1
6.7-20.9 -- 5.6-
15.8 0.01-0.08 0.1-1.1 Martin et al.,
1995
Weser Ästuar 0.03 64 -- 47 0.22-0.62 0.3 Turner et al.,
1992
Yenisey Ästuar 0.3 8.8-9.4 -- 22-29 0.01-0.02 0.03 Dai and
Martin, 1995
Gironde Ästuar 0.05 7 -- 15 0.5 0.2 Kraepiel et al., 1997
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 56 -
Tabelle 7. Konzentrationen der partikulären Spurenmetalle und des SPMs in Ästuaren.
Studiengebiet Fe Ni Co Cu Cd Pb SPM
Literatur mmol/g μmol/g μmol/g μmol/g nmol/g μmol/g mg/L
WR 0.94 0.61 0.29 0.52 3.7 0.24 4.3
diese Arbeit
WWR 0.87 0.93 0.33 0.60 3.4 0.16 3.8
MD 0.75 0.27 0.16 0.21 1.2 0.13 11.1
MG 0.73 0.24 0.13 0.19 0.7 0.13 17.1
Changjiang Ästuar 0.18-1.01
0.22-9.58
0.12-0.88
0.32-2.48
0.18-9.79
0.08-0.58 70 Wang and
Liu, 2003
Conwy Ästuar 0.57 2.98 -- 2.58 370 2.64 < 140 Zhou et al., 2003
Danshuei Ästuar 0.19-1.2 0.1-1.8 0.03-
0.32 0.35-7.9 -- -- -- Fang and
Lin, 2002
Elbe Ästuar 0.09-0.15
0.36-0.68
0.25-0.32
0.16-0.49 -- 0.26-
0.31 13-289 Turner et al., 1991
Galveston Bucht 0.09 -- -- 0.31 -- 0.97 10 Benoit et al., 1994
Gironde Ästuar 0.89 0.88 -- 0.58 4.8 0.27 -- Kraepiel et al., 1997
Humber Ästuar 0.09-0.52
0.32-0.53
0.19-0.49 0.2-3.8 -- 0.14-
0.33 35-693 Turner et al., 1991
Mersey Ästuar 0.30 0.38 0.02 0.41 82 0.45 < 700 Martino et al., 2002
Patos Bucht < 0.2 -- -- < 0.22 -- < 0.07 < 261
Niencheski and
Baumgarten, 2000
Rhône Ästuar 9.8 -- -- 0.72 -- 0.21 -- Elbaz-
Poulichet et al., 1996
Scheldt Ästuar 0.01-0.07
0.20-0.58
0.14-0.29
0.05-0.69 -- 0.11-
0.29 8-239 Turner et al., 1991
Seine Ästuar -- 0.22-1.02
0.07-0.26 0.7-4.0 14-59 0.24-
0.82 < 300 Chiffoleau et al., 1994
Sepetiba Bucht -- -- -- 0.74 2.4 0.07 160 Lacerda et al., 1987
Thames Ästuar 0.07-0.15
0.26-0.75
0.10-0.25
0.25-0.71 -- 0.22-
0.55 19-424 Turner et al., 1991
Venedig Lagune 0.14-0.30
0.26-0.64 -- 0.29-
3.8 3.2-11.0
0.038-0.075
3.5-23.9
Martin et al., 1995
Weser Ästuar 0.73 1.18 -- 0.65 5.2-64.1 0.55 < 400 Turner et
al., 1992
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 57 -
Bemerkenswert ist in Tabelle 6, dass die gelöste Konzentration von Cadmium im
Weser Ästuar deutlich höher ausgefallen ist als in den meisten anderen Ästuaren
weltweit. Die Daten von Turner (1992) in Tab. 6 stammen aus einer Expedition im
August 1989. Infolge des seit dem 10. Jahrhundert nach unserer Zeitrechnung
bestehenden Bergbaus im Harz sind die Harzer Oberböden zum Teil stark mit
Cadmium angereichert (Umweltbericht Niedersachsen, 2013). Durch Erosion gelangt
das Cadmium in die Aller, die einen Nebenfluss der Weser darstellt. Über die
Schwermetallfracht der Aller wird schließlich die Wasserqualität der Weser
entscheidend u.a. mit Cadmium belastet (FGG Weser, 2012). Jedoch deuten die
veröffentlichten Messdaten an der FGG-Weser-Messstelle Bremerhaven zwischen
den Jahren 1979 und 2000 auf einen Rückgang des gelösten Cadmiums in der
Weser hin (Abb. 25); allerdings sind keine Daten nach dem Jahre 2000 vorhanden.
Die durchschnittliche Konzentration des gelösten Cadmiums an dieser Messstelle ist
von ca. 19 nmol/L in 1980 auf ca. 7 nmol/L in 1998, d.h. um knapp zwei Drittel,
gesunken.
Abbildung 25. Gelöstes Cadmium in der Weser an der FGG-Weser-Messstelle Bremerhaven (Datenquelle: Weser-Datenbank, in: FGG Weser, 2012).
3.9. Offene Fragen und Perspektive
In dieser Arbeit wurden die gelösten, partikulären und kolloidalen Konzentrationen
von sechs Spurenmetallen im Wanquan Ästuar und Wenchang-Wenjiao Ästuar in
Ost-Hainan sowie in der Mundaú Lagune und der Manguaba Lagune in Nordost-
Brasilien bestimmt. Die ästuarinen Verteilungs- und Verhaltensmuster der gelösten,
partikulären und kolloidalen Spurenmetalle wurden diskutiert. Daneben wurden die
Konzentrationen und das Verhalten der sechs untersuchten Spurenmetalle in den
genannten Studiengebieten mit anderen Ästuaren weltweit verglichen und die
Kapitel 3: Vergleichende Diskussion der Ergebnisse
- 58 -
Kontaminationsniveaus in Bezug auf die untersuchten Spurenmetalle beurteilt. Einige
Fragen bedürfen jedoch weiterer Bearbeitung:
1. Welche Rolle spielt das Auftreten von oxischen/anoxischen Bedingungen in den
Oberflächensedimenten für das Spurenmetallverhalten in den genannten
Studiengebieten?
2. Wie groß ist der Anteil an den partikulären Spurenmetallen, der sich aktiv an den
ästuarinen Prozessen beteiligt, d.h. welches Verhältnis besteht zwischen den
Spurenmetallen an den Partikeloberflächen zu denen in Gitterpositionen?
3. In welchem Umfang sind die "echt gelösten" bzw. die kolloidalen Spurenmetalle
mit organischen Liganden bzw. organischen Partikeln assoziiert?
Um diese Fragen zu beantworten wäre es sinnvoll:
1. bei einer evtl. erneuten Beprobung der Studiengebiete den gelösten
Sauerstoffgehalt nahe der Sedimentoberfläche zu vermessen. Anhand der
Beziehung zwischen dem gelösten Sauerstoffgehalt und den Spurenmetall-
Konzentrationen könnte das Verhalten mancher Sauerstoff-empfindlicher
Spurenmetalle wie Cobalt und Cadmium besser verstanden werden.
2. zusätzlich andere Methoden zur Untersuchung der partikulären Spurenmetalle
wie z.B. die Festphasen-Extraktion (z.B. Danielsson et al., 1983; Fang and Lin,
2002) zu verwenden. Damit könnten Hinweise auf die Menge der tatsächlich
beteiligten partikulären Spurenmetalle gewonnen werden.
3. zusätzlich die physikalische Größe der Kolloide durch alternative Methoden, wie
z.B. mittels Field-Flow-Fraktionierung (z.B. Stolpe et al., 2010) zu bestimmen, um
operationelle Artefakte zu mindern.
4. die chemische Zusammensetzung unterschiedlicher Kolloidfraktionen durch
alternative Methoden, wie z.B. Ionenaustausch-Chromatographie (z.B. Wen et al.,
2011) oder "hyphenated methods" wie HPLC-ICP-MS zu untersuchen.
Kapitel 4: Publikation 1
- 59 -
4. Publikation 1:
erschienen als: Continental Shelf Research 57 (2013): 59-72.
Estuarine modification of dissolved and particulate trace metals in major rivers of East-Hainan
Jun Fu a, Xiao-Liang Tang a, Jing Zhang b, Wolfgang Balzer a, * a University of Bremen, FB2, Marine Chemistry, D- 28334 Bremen, Germany
b State Key Laboratory of Estuarine and Coastal Research, East China Normal University,
3663 Zhongshan Road North, Shanghai 200062, People’s Republic of China
* Corresponding author: e-mail address: balzer@mch.uni-bremen.de
Abstract
Dissolved and particulate cadmium, copper, iron, lead, cobalt and nickel were
analyzed in surface waters of the Wanquan River estuary and the
Wenchang/Wenjiao River estuary in East-Hainan Island during the dry season
(December 2006) and two wet seasons (August 2007 and July/August 2008). A major
difference to other Chinese rivers was the very low concentration of suspended
particles in these tropical Hainan estuaries. In the dissolved phase, a positive
deviation from the theoretical dilution line was observed for Cd during different
expeditions. Dissolved Cu and Ni essentially behaved conservatively, while Fe, Pb
and partly also Co correlated in their negative deviation from simple mixing. Strong
seasonal variability was observed only for dissolved Fe, Pb and Cd: sorption by the
much higher loading with suspended particles during the dry season lead to a strong
lowering of dissolved Fe and Pb, while the opposite was observed for dissolved Cd.
In both estuaries all six metals in particulate form showed almost constant values
with a tendency for slight decreases along the salinity profile. The normalization to
particulate Al revealed some specific particle properties during the different
expeditions. The dynamics of Fe chemistry dominated the distribution of Pb in all
forms. The distribution coefficients KD showed a general decrease in the order Fe >
Pb > Co > Ni > Cu ≈ Cd. There was no ‘‘particle concentration effect’’; rather the KD’s
of Fe and Pb exhibited slightly positive correlations with the suspended particle
loadings. Elevated concentrations levels in the Wenchang/Wenjiao River estuary,
especially during the wet season 2008, were ascribed to diffuse inputs from
Kapitel 4: Publikation 1
- 60 -
aquaculture ponds which girdle the upper estuary. In comparison to major Chinese
rivers, the tropical Hainan estuaries (S > 0) showed similar levels for Cd, Cu, Pb, Co
and Ni in particles and solution, while Fe was enriched in both matrices. On a global
scale, neither in the Wanquan River estuary nor in the Wenchang/Wenjiao River
estuary significant trace metal contamination was observed.
Keywords: Trace metals; Particulate; Dissolved; Estuary; Distribution coefficient;
Hainan
Kapitel 4: Publikation 1
- 61 -
4.1. Introduction
River-transported signals are subjected to a variety of physical, chemical and
biological processes in the estuarine mixing zone, in which the boundary conditions
are extremely variable in both space and time; estuaries can be thought of as acting
as filters of the river-transported chemical signals, which can often emerge from the
mixing zone in a form that is considerably modified with respect to that which entered
the system (Chester, 1990). This concept of the estuarine filter is based on the fact
that the mixing of the two very different end-member waters will result in the setting
up of strong physico-chemical gradients in an environment. It is these gradients
which are the driving force behind the filter. In addition, also organic matter
production (Louis et al., 2009), the oxygenation state (Khalid et al., 1978; Zwolsman
and van Eck, 1999), possible inputs from anoxic sediments (Salomons et al., 1987)
and/or other processes may have influence.
Trace metals are a particularly interesting aspect of estuarine chemistry because
their differing physical chemistries lead to a variety of geochemical behaviours
(Shiller and Boyle, 1991). For example, previous studies have indicated an affinity of
cobalt to manganese oxide phases and a removal of lead through adsorption onto
the suspended particulate matter (Chiffoleau et al., 1994), the desorption of cadmium
from suspended particles and the flocculation of iron colloids (Roux et al., 1998), the
removal of copper from the dissolved phase at low salinities (Comber et al., 1995)
and non-conservative behaviour of nickel (Wang and Liu, 2003). The trace metal
behaviour in the large rivers and major estuaries of China at mid latitudes such as
the Changjiang (Yangtze River) and the Huanghe (Yellow River) has been well
studied (e.g., Zhang, 1995; Zhang and Liu, 2002; Wang et al., 2009; Tang et al.,
2010), but about the trace metal chemistry in estuaries of tropical China, especially of
Hainan Island, very little is known.
Adsorption/desorption is one of the most significant factors that affect the solid-
solution interaction (O’Connor and Connolly, 1980). Under the assumption that
equilibrium conditions prevail, the solid-solution-interaction can be described in terms
of the conditional distribution coefficient (or partition coefficient) KD, which is defined
as the ratio of the adsorbed or the total particulate concentration (CP, w/w) to the
dissolved concentration (CS, w/v) of a chemical constituent: KD = CP / CS. KD’s are of
fundamental significance to geochemical modelling and pollution impact assessment
(e.g., Wood et al., 1995). The distribution coefficients provide empirical information
on the combined effect of heterogeneous reactions of an element at the solid-solution
interface. An elevated KD value may indicate that an element is associated and
Kapitel 4: Publikation 1
- 62 -
transported with the solid phase, which eventually may become part of the sediment
and may never reach the ocean. From the dependence of the distribution coefficients
on salinity it is possible to identify whether or not there is a tendency for the release
of the respective metal from the particulate matter, when proceeding from fresh water
through brackish to seawater conditions. Estuarine cycling of trace metals covering
all these aspects and including also changes over decades has been studied
extensively in the Scheldt estuary; it is characterised as a system with strong
physico-chemical gradients, large anthropogenic inputs, a very long mixing zone of
fresh- and saltwater (up to 100 km) and a long residence time of about 3 months of
water in the upper estuary (e.g., Paucot and Wollast, 1997; Zwolsman et al., 1997;
Zwolsman and van Eck, 1999).
In this paper we investigate the environmentally relevant trace metals cadmium,
copper, iron, lead, cobalt and nickel in dissolved and particulate form. In addition to
contributing to the general knowledge about the trace metal behaviour in tropical
estuaries, a specific objective was to estimate the contamination state of the
Wanquan River and the Wenchang/Wenjiao River estuaries of Hainan Island (~
20°N).
4.2. Materials and Methods
4.2.1. Study area
Fig. 1. Estuaries of East Hainan under investigation: (a) Wenchang/Wenjiao River estuary (WWR) and (b) Wanquan River estuary (WR).
The Wanquan River (in the following: WR) and the Wenchang/Wenjiao River system
Kapitel 4: Publikation 1
- 63 -
(in the following: WWR) in East-Hainan are both located at the northern part of the
tropical zone with a humid warm climate (Fig. 1). The Wanquan River, with a
drainage area of 3693 km2, a total length of 156.6 km and a mean annual discharge
of 163.9 m3/s, is the third largest river in Hainan Island (Ge et al., 2003). The
Wenchang River (drainage area 381 km2, total length 37 km, mean annual discharge
9.1 m3/s) and the Wenjiao River (drainage area 522 km2, total length 56 km, mean
annual discharge 11.6 m3/s) empty into the Bamen Bay (Zeng and Zeng, 1989) and
form the Wenchang/Wenjiao River estuary (areal extent: ~ 40 km2; Herbeck et al.,
2013). Both estuaries have a micro-tidal, irregular diurnal tidal regime with a mean
range of about 0.7-0.8 m (Zhu et al., 2005). The salinity intrusion may extend 5 km
from the Yudai Sand Barrier into the WR and from the Bamen Bay end into the WWR,
respectively. The WWR is a shallow estuary ( < 3 m; except for the shipping channel)
with a low flow rate, while in the WR the position of the mixing zone is highly variable
due to changes of the river discharge rate, tides and winds. About 80% of the annual
rainfall (1740 mm/yr; Liu et al., 2011) and of the water flow in this region occurs in the
wet period from May to November (Ma et al., 2007). Typhoons with heavy rainfall and
elevated input of soil erosion products, nutrients and suspended particles (Herbeck et
al., 2011) did not occur during the sampling for this study. The water residence time
in the estuaries was estimated to be 7.8 day in WWR (Liu et al., 2011) and 0.2-4.7
day in WR (Li et al., 2013). Comprehensive nutrient investigations of the WWR
system (Liu et al., 2011) revealed that the tributaries are enriched in dissolved
inorganic nitrogen and depleted in dissolved inorganic phosphorus and have major
contributions of dissolved organic forms to the total N and P dissolved concentrations;
these elements - in contrast to Si - show non-conservative behaviour. The WWR
system is characterized by lower silicate than average in tropical systems (Liu et al.,
2011), while the Wanquan River estuary (WR) is enriched in silicate (Li et al., 2013).
Because of the generally very low depths of about 3 m except for a few more meters
in the shipping channels, oxygen depletion or anoxic conditions with their far
reaching effects on redox-dependent trace metals (e.g., Balzer, 1982) were not
observed and could not be expected. Ma et al. (2007) investigated major and trace
elements in laterites of Northern Hainan which are in-situ weathering products
developed from basalts. Typical soils of the study area are Oxisols which are
especially rich in Fe2O3 (Li et al., 2012) which explains the comparatively high Fe
values in the estuary found during the present study. The Shilu Iron Mine as China’s
largest iron mine is located in the western part of the Hainan Island (Hsieh and
Zhong, 1990), but is outside the drainage area of either estuary studied.
Nevertheless, the high Fe content of the soil and possibly also some atmospheric
Kapitel 4: Publikation 1
- 64 -
transport from this mine might affect aquatic processes of East-Hainan. In contrast to
the WR, the upper WWR estuarine system is girdled by shrimp and fish ponds with
an estimated area of ~ 21.6 km2 and estimated effluents of ~ 210 • 106 m3/yr
(Herbeck et al., 2013; Unger et al., 2013). Minor contamination of both systems may
also originate from municipal discharge of small towns. Hainan has a population
density of 255 inhabitants/km2.
4.2.2. Sampling
Surface water samples were collected during the dry season (Dec. 2006) and the wet
seasons (Aug. 2007 and Jul./Aug. 2008) in the WR and the WWR (exact positions
may be obtained from the corresponding author upon request). Upstream of the ship,
low-density polyethylene bottles (LDPE, 1 L, Nalgene®) were filled by hand with
surface water from a depth of about 0.3 m. Within 20 h the water samples were
drawn through acid-cleaned 0.4 mm polycarbonate filters (pre-weighed, Nuclepore®,
Whatman) into LDPE bottles (Nalgene®). A 10 mL aliquot of each sample was filled
into a glass ampoule for the determination of the dissolved organic carbon (DOC)
followed by the addition of phosphoric acid to remove the inorganic carbon and for
sterilization. After acidification to pH 2 with sub-boiled concentrated nitric acid,
sample bottles for trace elements were wrapped in polyethylene bags and stored at
room temperature. The filter cakes were transferred to Petri dishes and stored at -
10 °C.
4.2.3. Pretreatment and measurement of dissolved metals
Sample work-up was performed in a clean-air room (class 1000) with additional clean
benches of class 100. The filtrates were first irradiated in a UV-digester at 90 °C for 2
h with addition of hydrogen peroxide in order to destroy the organic complexes. Then
the trace metals were preconcentrated by an extraction/backextraction procedure
developed by our workgroup based on the work of Danielsson et al. (1978). A 40-100
g sample (depending on the metal concentration) was weighed into a fluorinated
ethylene propylene separatory funnel (FEP, Nalgene®) and after adding (i)
ammonium acetate buffer to adjust the pH value of the sample to 4.0-4.5, (ii) 500 μL
APDC/NaDEDTC solution (0.06 mol/L) and (iii) 10 mL Freon, the sample had been
well shaken (15 min). After phase separation, the lower organic phase was drawn off
into an acid-cleaned polypropylene tube and 100 μL concentrated nitric acid were
added, followed by a further addition of 1900 μL Milli-Q® water. After phase
Kapitel 4: Publikation 1
- 65 -
separation, the upper aqueous extract was removed into an acid-cleaned sample
tube by using a pipette. The extract was analyzed by high-resolution inductively-
coupled-plasma mass-spectrometry (HR-ICP-MS; Element 2, Thermo Scientific).
Four certified reference materials from the National Research Council of Canada
(NRCC) were investigated: SLRS-5 (river water), SLEW-3 (estuarine water), CASS-4
and NASS-5 (seawater); the results for all metals in all reference materials were
inside the certified ranges. The detection limits (3σ; n = 13) for Milli-Q® water being
subjected to UV-digestion and liquid/liquid-extraction were 0.004, 0.016, 2.63, 0.067,
0.004, 0.127 for Cd, Pb, Fe, Ni, Co, Cu, respectively. The mean precisions of
replicates during the analysis of the certified reference materials were 0.7%, 2.6%,
10.2%, 5.7%, 2.6%, 5.3% for Cd, Pb, Fe, Ni, Co and Cu, respectively.
4.2.4. Pretreatment and analysis of particulate metals
The filter cakes were weighed and completely decomposed using a microwave-
assisted pressure digestion technique after adding a mixture of concentrated nitric,
hydrochloric and hydrofluoric acid (for details, see: Balzer et al., 2013). At regular
intervals the certified reference material MESS-1 (marine sediment) from the NRCC
was analyzed and all values were found to be in the certified ranges. The protocol
implies a total destruction of the particles and includes trace metals in surface layers
and in lattice positions. Therefore, results cannot be directly compared to many other
studies of estuarine particles, during which the metals are leached from the surface
layers by employing e.g., acetic acid or dilute hydrochloric acid. The metal
concentrations were determined by HR-ICP-MS.
4.2.5. Other parameters
Environmental parameters such as salinity and pH were measured on the spot using
a multi-sensor-device (MultiLine P3, WTW). Dissolved organic carbon (DOC) in the
sealed glass ampoules was determined by high temperature catalytic oxidation (Total
Carbon Analyzer, Apollo 9000; Tecmar) with a detection limit of 0.1 mg/L. Replicates
from 19 samples gave an average relative standard deviations of 2.5% (n = 5).
Before use, the glass ampoules were pre-combusted at 450 °C for 4 h.
Kapitel 4: Publikation 1
- 66 -
4.3. Results and discussion
4.3.1. Distribution of suspended particulate matter and environmental conditions
WR
0
5
10
15
20
0 10 20 30
SPM
(mg/
L)
WWR
0
5
10
15
20
0 10 20 30
SPM
(mg/
L)
WR
0
2
4
6
8
0 10 20 30Salinity (psu)
DO
C (p
pm)
WWR
0
2
4
6
8
0 10 20 30Salinity (psu)
DO
C (p
pm)
Fig. 2. Distribution of SPM and DOC vs. salinity (open symbols for the dry season and solid symbols for the wet seasons; : WR 2006; : WWR 2006; : WR 2007; : WWR 2007; ▲: WR 2008; : WWR 2008).
The distributions of the suspended particulate matter (SPM) and the dissolved
organic carbon (DOC), which were obtained during the three expeditions in the two
estuaries, are shown in Fig.2.
Except for a slight minimum at mid salinities, the SPM concentration was essentially
constant in both estuaries during the wet seasons (Aug. 2007: on average 2.3 mg/L
and Jul./Aug. 2008: on average 2.0 mg/L). The SPM values during the dry season
(Dec. 2006: on average 9.8 mg/L) showed more variability and were distinctly higher
than during both wet seasons, possibly originating from higher resuspension loadings
at lower water levels and/or from enhanced dust deposition. Also Li et al., 2013 report
higher SPM values for the dry season (Dec. 2006, Mar./Apr. 2009) in comparison to
the wet seasons (Aug. 2007, Jul/Aug. 2008) for the Wanquan estuary (WR). The
SPM concentrations were much lower than in other Chinese estuaries e.g., in the
Daliao River (8-1001 mg/L; Guo et al., 2009), in the Changjiang estuary (70-350
Kapitel 4: Publikation 1
- 67 -
mg/L; Koshikawa et al., 2007), or in the Zhujiang (Pearl River) Delta (70-247 mg/L; Ni
et al., 2008). Summarizing the studies in several Chinese estuaries, Zhang and Liu
(2002) reported a typical SPM range from 10 mg/L to 20000 mg/L. The SPM content
in both study areas is obviously at the lower limit of this range.
The DOC content generally decreased with increasing salinity with concentration
ranges of 1-3 ppm during the dry season and 1-8 ppm during the two wet seasons.
Comparable data was determined in the Amazon estuary (1-4 ppm; Sholkovitz et al.,
1978) and in the Zhujiang estuary (1-4 ppm; Dai et al., 2000). The DOC content in
the WWR was considerably higher than in the WR and its distribution showed much
more scatter especially during the wet season of 2008. In accordance with the fact
that the WWR is an aquacultural region with some biological activity, anthropogenic
sources from the aquacultural activities and overflow during storms were suspected
to be responsible for the higher DOC concentration in the WWR. In contrast to WR,
both the river and the upper estuary (Bamen Bay) of the WWR system are
surrounded by aquaculture ponds as shown in detail by Unger et al., 2013. Also the
different DOC levels between the two wet seasons in WWR are probably related to
different aquaculture inputs during the specific sampling campaigns.
The pH increased with increasing salinity (not shown). It ranged between 7.0 and 8.3
in the WR and between 6.8 and 8.5 in the WWR which is comparable with other
Chinese estuaries: e.g., in the Zhujiang estuary (7.0-8.0; Dai et al., 2000) and in the
Changjiang estuary (7.5-8.3; Zhang and Zhang, 2003).
Kapitel 4: Publikation 1
- 68 -
WR
0.00
0.05
0.10
0.15
0 10 20 30
Cd
(nm
ol/L
)
WWR
0.00
0.05
0.10
0.15
0 10 20 30
0
5
10
15
20
0 10 20 30
Cu
(nm
ol/L
)
0
5
10
15
20
0 10 20 30
0
5
10
15
0 10 20 30
Fe (μ
mol
/L)
0
5
10
15
0 10 20 30
0
1
2
3
0 10 20 30
Pb (n
mol
/L)
0
1
2
3
0 10 20 30
0
3
6
9
0 10 20 30
Co
(nm
ol/L
)
0
3
6
9
0 10 20 30
0
5
10
15
20
0 10 20 30Salinity (psu)
Ni (
nmol
/L)
0
5
10
15
20
0 10 20 30Salinity (psu)
Fig. 3. Dissolved trace metals (left column: WR, right column: WWR) vs. salinity (open symbols: dry season, solid symbols: wet seasons; : WR 2006; : WWR 2006; : WR 2007; : WWR 2007; ▲: WR 2008; : WWR 2008); for further details: see text.
Kapitel 4: Publikation 1
- 69 -
4.3.2. Dissolved trace metals
Tab. 1. Dissolved trace metals (average values ± 1 standard deviation; n = number of values) in freshwater (S = 0), in the estuarine mixing zone (0 < S < 28) and at high salinities (S > 28) of the WR and the WWR during the dry season (2006) and the wet seasons (2007 and 2008).
Expedition Cd Cu Fe Pb Co Ni
nmol/L nmol/L μmol/L nmol/L nmol/L nmol/L
WR 2006
S=0 (n=3) 0.02±0.01 17±16 0.23±0.09 0.25±0.26 0.27±0.08 3.2±1.0
0<S<28 (n=3) 0.11±0.02 19±19 0.06±0.05 0.30±0.13 6.1±9.0 6.5±2.2
S>28 (n=0) -- -- -- -- -- --
WWR 2006
S=0 (n=2) 0.07±0.09 8.2±0.5 0.44±0.39 0.06±0.06 2.6±3.0 8.1±0.5
0<S<28 (n=3) 0.10±0.01 8.7±2.1 0.59±0.79 0.12±0.13 6.8±1.7 12.1±2.6
S>28 (n=4) 0.14±0.01 8.4±2.9 0.04±0.03 0.08±0.04 0.98±0.34 5.9±0.7
WR 2007
S = 0 (n=2) 0.06±0.02 9.7±0.7 4.2±0.2 1.4±0.2 1.6±0.2 6.6±0.8
0<S<28 (n=8) 0.06±0.01 7.6±0.9 2.5±0.8 0.81±0.18 1.2±0.4 5.6±1.2
S>28 (n=1) 0.06 2.3 0.28 0.15 0.41 3.8
WWR 2007
S = 0 (n=2) 0.02±0.01 8.3±1.1 3.7±2.6 0.51±0.51 2.4±0.2 11.9±8.3
0<S<28 (n=8) 0.07±0.03 7.1±1.0 1.1±1.3 0.23±0.17 2.6±1.4 7.0±2.3
S>28 (n=1) 0.09 6.2 0.16 0.06 0.91 4.6
WR 2008
S = 0 (n=4) 0.03±0.00 6.6±1.8 3.7±0.7 1.0±0.2 1.6±0.4 2.8±1.3
0<S<28 (n=5) 0.08±0.02 7.0±3.0 2.2±2.4 0.60±0.58 0.90±0.59 5.3±2.6
S>28 (n=1) 0.10 1.7 0.16 0.01 0.09 4.2
WWR 2008
S = 0 (n=4) 0.02±0.01 11.4±1.1 9.6±1.6 1.6±0.6 4.2±0.6 6.8±2.8
0<S<28 (n=10) 0.07±0.03 12.0±0.9 4.0±4.1 0.85±0.61 2.7±1.3 10.2±3.0
S>28 (n=1) 0.09 7.6 0.35 0.25 0.67 2.8
The dissolved trace metal concentrations obtained during the three expeditions in the
two estuaries are plotted as a function of salinity in Fig. 3. The dissolved metal
concentrations depicted for salinity zero, which show considerable variability for most
metals, are not always representative for the geographical freshwater entrance to the
estuaries but also include samples which were taken upstream the rivers at different
dates. The average concentrations for freshwater, in the estuarine mixing zone (0 < S
< 28) and at high salinities of the WR and the WWR during the three expeditions are
shown in Table 1.
Kapitel 4: Publikation 1
- 70 -
Cadmium: Dissolved Cd concentrations generally increased with increasing salinity
from 0.01 to 0.05 nmol/L in the freshwater to about 0.14 nmol/L in the saline waters
during the dry season (only WWR) and to about 0.10 nmol/L at salinities above S =
28 during the wet seasons. During the dry season of 2006 and the wet season of
Jul./Aug. 2008, the dissolved Cd showed a clear positive deviation from the
theoretical dilution line in both estuaries. A similar non-conservative behaviour of Cd
was also observed in many other studies, including the Changjiang estuary (Edmond
et al., 1985; Elbaz-Poulichet et al., 1987), the Hudson River estuary (Yang and
Sañudo-Wilhelmy, 1998), the Loire estuary (Waeles et al., 2004) and during
laboratory experiments (e.g., Roux et al., 1998). This behaviour indicates a
mobilization of dissolved Cd from the SPM into the solution phase due to the
formation of highly stable Cd-chloro-complexes (Van der Weijden et al., 1977;
Kraepiel, et al., 1997). Although showing some scatter, the dissolved Cd distribution
during the wet season of August 2007 seems to be conservative. Such a behaviour,
which was also observed e.g., in the outer Severn estuary (Harper, 1991) and in the
Rhône estuary in July 1987 (Elbaz-Poulichet et al., 1996), might be related to low
particulate Cd per volume and a short residence time of water and particles in the
estuary as suggested by Kraepiel et al. (1997). While the residence times in the
Hainan estuaries are indeed short (see above), the actual concentrations of
particulate Cd per volume in 2007 and 2008 do not explain such a different behaviour
between the two years. Thus, the exact reason for it remains open. Dilution by
enhanced rainfall and the slightly lower pH during the dry season might have
contributed to the seasonally lower concentrations of dissolved Cd during the wet
seasons (Hatje et al., 2003). Because the dry season is also associated with
comparably high levels of SPM, very low concentrations of dissolved Fe and Pb (see
below), and essentially equal values for dissolved Cu and Ni, unknown properties of
the additional particles might have played a role for the removal of dissolved Fe and
Pb and the simultaneous release of Cd.
Copper: With increasing salinity dissolved Cu exhibited a slight but continuous and
essentially conservative decline, which is clearly evident in the WR during all
investigated seasons, but can be identified in the WWR estuary only as a tendency.
Elevated Cu concentrations were observed during the wet season 2008 in WWR,
coinciding with higher levels of DOC (Fig. 2). In combination with more scatter this
may be explained by irregular inputs from adjacent aquaculture ponds (Balzer et al.,
2013). Dissolved copper decreased from typically about 8-12 nmol/L in the
freshwater to about 3-7 nmol/L at higher salinities. A similar conservative behavior of
Kapitel 4: Publikation 1
- 71 -
dissolved Cu was also observed in the Amazon plume (Boyle et al., 1982), in the
Changjiang estuary (Edmond et al., 1985), in the Mississippi River Delta (Shiller and
Boyle, 1991) and in the tropical Bang Pakong estuary (Windom et al., 1988), and
seems to be related to the Cu speciation in form of stable complexes with organic
ligands (Shiller and Boyle, 1991; Roux et al., 1998). Strong positive deviations from
simple dilution of dissolved Cu as observed e.g., in the Scheldt estuary were
explained by sulfide precipitation under anoxic conditions at low salinities (Paucot
and Wollast, 1997; Zwolsman et al., 1997); however, oxygen depletion was not likely
to occur in the estuaries under investigation.
Iron: Dissolved Fe in both study areas and during both wet seasons decreased from
typically 5-10 μmol/L in the riverine water to 0.1-0.2 μmol/L at higher salinities.
Although the concentration of dissolved Fe in both estuaries during the dry season
was about 10 times lower than during both wet seasons, there was also a slight
decrease with salinity in both estuaries. Dissolved Fe showed a non-conservative
behavior and appeared to be removed from solution in both estuaries and during
both dry and wet seasons. Two elevated data points for Fe (and Pb) at low salinities
are considered outliers being due to either contamination or irregular input from
aquaculture ponds. Field and laboratory studies indicate that the main process
affecting Fe during estuarine mixing is flocculation of colloidal Fe forms and
subsequent sedimentation, globally resulting in a reduction of more than 90% of the
Fe river input during its way to the ocean (Boyle et al., 1977; Sholkovitz, 1978). In the
wet season of 2008, the dissolved Fe of the freshwater and at low salinities in the
WWR was two times higher than during all other expeditions and in the WR,
suggesting lateral inputs from aquaculture ponds as discussed before. The relatively
low concentrations of Fe (and Pb) in solution during the dry season 2006 coincide
with similar particulate Fe levels and similar Fe/Al ratios but an about six times higher
SPM concentration. This suggests enhanced adsorption onto unsaturated sites of the
additional particle load during 2006. However, the suspended particles form only a
fraction of the total particulate load with a major contribution from bed sediments in
addition to riverine particles (Kraepiel et al., 1997). Thus, the striking difference
between dissolved Fe and Pb during the dry and the wet seasons might also arise
from different ratios of the two particle types.
Kapitel 4: Publikation 1
- 72 -
0
1
2
0 5 10 15Fe (μmol/L)Pb
(nm
ol/L
)
Fig. 4. Dissolved Pb vs. dissolved Fe (open symbols: dry season, solid symbols: wet seasons; : WR 2006; : WWR 2006; : WR 2007; : WWR 2007; ▲: WR 2008; : WWR 2008). Broken and solid line: correlation for all WWR data (r2 = 0.849) and for all WR data (r2 = 0.923), respectively.
Lead: In the estuaries, dissolved Pb during both wet seasons exhibited (similar to Fe)
a very strong decline from riverine waters (1-2 nmol/L) to marine waters (0.02-0.3
nmol/L). Such a distribution was also observed in the Gironde estuary (Elbaz-
Poulichet et al., 1984), in the Bristol Channel (Harper, 1991), in the Bang Pakong
estuary (Windom et al., 1988) and in the Cochin estuary (Ouseph, 1992) and is often
explained by rapid sorption of dissolved Pb onto re-suspended particles. In contrast,
dissolved Pb during the dry season showed only a small decline with increasing
salinity and was 40-80% lower than during the wet seasons. Considering the good
correlation between dissolved Fe and dissolved Pb (r2 = 0.849 for all WR data, r2 =
0.923 for all WWR data; Fig. 4), a removal of dissolved Pb with freshly precipitated
Fe-oxy-hydroxides may be suspected (e.g., Balls, 1990; Wen et al., 2008).
0
2
4
6
0 5 10 15Fe (μmol/L)
Co
(nm
ol/L
)
Fig. 5. Dissolved Co vs. dissolved Fe during the wet seasons (: WR 2007; : WWR 2007; ▲: WR 2008; : WWR 2008). Broken and solid line: correlation for all WWR data (r2 = 0.887) and for all WR data (r2 = 0.515), respectively.
Cobalt: During both wet seasons, dissolved Co in WR decreased continuously with
Kapitel 4: Publikation 1
- 73 -
increasing salinity from 2 to 5 nmol/L to less than 1 nmol/L and behaved similar as Fe
and Pb. Also the differences between WR and WWR were similar to Fe and suggest
that the distribution of Co might also be controlled by co-precipitation with Fe-oxy-
hydroxides (Fig. 5). The good correlation of dissolved Co vs. Fe in WR (r2 = 0.887)
during the wet seasons supports the predominance of Fe-related processes. On the
other hand, the weak correlation for the WWR system (r2 = 0.515) provides evidence
for contribution from other processes; most probable are lateral inputs from
aquaculture ponds which girdle the WWR rivers and upper estuary in contrast to the
WR estuary (Herbeck et al., 2013; Unger et al., 2013). This behavior of Co during the
wet seasons contrasts with reports of dissolved Co maxima at low or mid salinities
(Windom et al., 1988; Martino et al., 2002; Takara et al., 2010). While the few
observations during the dry season in WR and WWR might be interpreted as low
salinity maxima or conservative behaviour (at S > 1), it is clear that additional work is
needed to understand the estuarine behavior of cobalt.
Nickel: During both wet seasons dissolved Ni concentration was essentially constant
at about 5 nmol/L over the whole salinity profile in the WR, while in the WWR Ni
decreased from 10 to 15 nmol/L to also about 5 nmol/L at higher salinities. When
excluding two higher values, dissolved Ni might be considered conservative in both
estuaries during the wet seasons, as also observed e.g., in the Göta River estuary
(Danielsson et al., 1983). A part of the Ni variability at low salinities may be explained
by fluctuations in the incoming river water (Chaudry and Zwolsman, 2008) and (in the
WWR) may also contain an anthropogenic contribution from the Wenchang town
being located at the Wenchang tributary. In addition, at salinities of about S = 20
diffuse inputs in WWR from aquaculture may be suspected. During the dry season,
dissolved Ni had a tendency for a positive deviation from conservative behavior,
which might be attributed to anthropogenic discharges (e.g., Campbell et al., 1988;
Martino et al., 2004).
Kapitel 4: Publikation 1
- 74 -
WR
0
4
8
12
0 10 20 30
Cd
(nm
ol/g
)
WWR
0
4
8
12
0 10 20 30
0
1
2
3
0 10 20 30
Cu
(μm
ol/g
)
0
1
2
3
0 10 20 30
0
1
2
3
0 10 20 30
Fe (m
mol
/g)
0
1
2
3
0 10 20 30
0.0
0.3
0.6
0.9
0 10 20 30
Pb
(μm
ol/g
)
0.0
0.3
0.6
0.9
0 10 20 30
0.0
0.3
0.6
0.9
0 10 20 30
Co
(μm
ol/g
)
0.0
0.3
0.6
0.9
0 10 20 30
0
1
2
3
0 10 20 30Salinity (psu)
Ni (μm
ol/g
)
0
1
2
3
0 10 20 30Salinity (psu)
Fig. 6. Particulate trace metals (left column: WR, right column: WWR) vs. salinity (open symbols: dry season, solid symbols: wet seasons; : WR 2006; : WWR 2006; : WR 2007; : WWR 2007; ▲: WR 2008; : WWR 2008); for further details: see text.
Kapitel 4: Publikation 1
- 75 -
4.3.3. Particulate trace metals
Table 2. Particulate trace metals (average values ± 1 standard deviation; n = number of values) in freshwater (S = 0), in the estuarine mixing zone (0 < S < 28) and at high salinities (S > 28) of the WR and the WWR during the dry season (2006) and the wet seasons (2007 and 2008).
Expedition Cd Cu Fe Pb Co Ni
nmol/g μmol/g mmol/g μmol/g μmol/g μmol/g
WR 2006
S=0 (n=3) 5.3±0.1 0.33±0.02 1.2±0.0 0.26±0.00 0.30±0.03 0.57±0.04
0<S<28 (n=3) 3.1±1.1 0.28±0.02 0.88±0.24 0.20±0.05 0.28±0.12 0.46±0.06
S>28 (n=0) -- -- -- -- -- --
WWR 2006
S=0 (n=2) 6.1±4.6 0.57±0.32 1.8±0.5 0.18±0.01 0.34±0.20 1.3±0.7
0<S<28 (n=3) 3.3±1.0 0.55±0.20 1.3±0.6 0.11±0.02 0.34±0.07 1.1±0.5
S>28 (n=4) 1.7±0.3 0.35±0.04 0.61±0.09 0.13±0.02 0.23±0.04 0.60±0.08
WR 2007
S = 0 (n=2) 4.6±0.7 0.46±0.02 1.1±0.0 0.26±0.01 0.44±0.01 0.81±0.06
0<S<28 (n=8) 2.2±0.6 0.48±0.25 0.91±0.19 0.21±0.04 0.27±0.05 0.54±0.10
S>28 (n=1) 3.2 0.25 0.57 0.17 0.22 0.45
WWR 2007
S = 0 (n=2) 3.2±1.7 0.52±0.30 1.9±0.3 0.23±0.06 0.91±0.69 1.2±0.3
0<S<28 (n=8) 3.5±1.1 0.32±0.20 0.68±0.55 0.17±0.09 0.25±0.20 0.9±1.2
S>28 (n=1) 4.2 0.26 0.26 0.09 0.11 0.32
WR 2008
S = 0 (n=4) 6.7±1.6 0.81±0.29 1.0±0.4 0.32±0.15 0.40±0.15 0.73±0.28
0<S<28 (n=5) 2.9±0.4 0.68±0.11 0.94±0.12 0.27±0.02 0.25±0.07 0.76±0.13
S>28 (n=1) 2.8 0.50 0.42 0.16 0.15 0.37
WWR 2008
S = 0 (n=4) 3.2±1.3 0.81±0.32 0.82±0.28 0.17±0.02 0.39±0.15 1.0±0.4
0<S<28 (n=10) 3.5±0.7 0.88±0.21 0.70±0.25 0.170.03 0.32±0.14 0.88±0.28
S>28 (n=1) 3.8 0.89 0.58 0.22 0.21 1.3
The concentrations of Cd, Cu, Fe, Pb, Co and Ni in the particles obtained during the
three expeditions in the two estuaries are plotted as a function of salinity in Fig. 6.
The average values are shown in Table 2. In the WR system, all particulate metals
show a very homogeneous distribution with constant or slightly decreasing levels
over the whole estuarine salinity profile. Except for some variability among the
freshwater samples, no difference occurred between the seasons. In contrast, the
salinity profiles of the particulate trace metals in the WWR system show much more
scatter in the data, but mainly at the same concentration levels as in WR. Some
Kapitel 4: Publikation 1
- 76 -
slightly positive deviations from simple endmember mixing suggest that in WWR
particles with different properties might have contributed to the composite signals.
Likely candidates with slightly differing properties are riverine particles entering the
estuary, resuspended bottom sediments and possibly also some aquacultural
particles.
In order to identify particles of potentially differing origin among the expeditions and
to obtain hints for biogenic and/or anthropogenic contributions (Zhang and Liu, 2002),
the trace metal contents of all SPM samples were normalized to aluminium as a
proxy for the mineral particle matrix (not shown). The particulate Al itself decreased
linearly along the salinity profile from 3.0 ± 1 mmol/g in the freshwaters and in the low
salinity region to 2.2 ± 0.3 mmol/g near the coastal endmember (S > 28). The only
exception were some low Al values during the wet season 2007 in WWR at S = 20-22;
concomitant high ratios to Al for Cd and Pb but not for Cu, Co, Ni suggests an
anthropogenic contribution; alternatively, it may originate from an unusually high
fraction of biogenic particles with elevated contents of particulate carbonates and/or
organic matter (e.g., Koshikawa et al., 2007).
From high riverine values in the WR particulate Cd rapidly decreased and remained
constant over the whole salinity range during both wet seasons, which was also
found in the Gironde estuary (Kraepiel et al., 1997). In the WWR system, the
particulate Cd appeared to slightly increase, but at a similar level as in the WR. The
Cd/Al ratio was constant along the salinity profile except for elevated ratios of Cd/Al
(and also of Pb/Al) at S = 20-28 during the wet season 2007 in WWR. Exponential
decreases of particulate Cd (and also Cu) over the whole salinity profile (i.e.,
negative deviations from simple dilution) being attributed to desorption into the
dissolved phase (e.g., in the highly polluted Scheldt estuary; Zwolsman et al., 1997;
Paucot and Wollast, 1997) cannot be detected in the Hainan estuaries. With few
other exceptions (e.g., enrichment of particulate Cd at mid salinities in the Danshuei
River estuary; Jiann et al., 2005), simple estuarine mixing seems to be the dominant
feature of particulate Cd worldwide (e.g., Balls, 1990; Paalman and Van der Weijden,
1992; Wen et al., 2008; Zwolsman and van Eck, 1999).
In both estuaries and during dry and wet seasons the particulate Cu showed no
considerable change except for a slight decrease towards the coast. Such
distributions are often found in estuaries, e.g., in several Texas estuaries (Benoit et
al., 1994) and in the Gironde estuary (Kraepiel et al., 1997). However, also a positive
relation of particulate Cu (and Pb) to salinity as a result of pH dependent sorption to
Fe-Mn-oxides has been reported (Hatje et al., 2001). In contrast to any other of the
Kapitel 4: Publikation 1
- 77 -
six particulate trace metals, elevated values of particulate Cu/Al were observed
during the wet season 2008 in WWR over the whole salinity range. In accordance
with comparatively high values for dissolved and particulate arsenic during the same
season in WWR (Balzer et al., 2013), a relatively high contribution of copper-rich
inputs from adjacent aquaculture ponds may be suspected. This presumption is
corroborated by a few measurements of aquaculture effluents which showed some
very high dissolved Cu values.
Particulate Fe, Ni and Co during all seasons decreased with increasing salinity,
generally with a larger slope in the WWR than in the WR. Such a distribution pattern
was often found, e.g., in the Clyde estuary (Balls, 1990), in the Rhine/Meuse estuary
(Paalman and Van der Weijden, 1992) and in the Tanshui River estuary (Fang and
Lin, 2002). Also particulate Pb showed a slight decrease with salinity but no temporal
variability. This essentially conservative salinity profile agrees with observations in the
Gironde estuary (Kraepiel et al., 1997) and - combined with a negative relation to
salinity - it was also observed in the Weser estuary (Turner et al., 1992). The ratios to
Al of particulate Fe, Pb, Co and Ni during all expeditions (and seasons) were rather
constant over the whole salinity range except for some high values of Fe/Al in WWR
at low to mid salinities in 2006 and 2007, and of Pb during the wet season 2008. It
cannot be ruled out that the high Fe/Al during the dry season 2006 partly correspond
to the extremely low dissolved Fe during that expedition.
A striking feature of the ratios to Al of all trace metals under investigation was their
rather constant lower limit over the whole salinity range. The ratios (in: mol/mol) of
Cd/Al = 6.0•10-7, Cu/Al = 1.0•10-4, Fe/Al = 2.4•10-1, Pb/Al = 6.0•10-5, Co/Al = 8.0•10-5
and Ni/Al = 1.8•10-4 might be representative for the lattice composition of the particles
from the drainage area. However, these ratios to Al for Cu, Fe, Co and Ni are lower
by 18-70% than the top layer of the East-Hainan laterites which are the products of
extreme tropical weathering (Ma et al., 2007; no data for Cd). In contrast, the Pb/Al of
the estuarine particles was considerably higher than in the weathered laterites. If the
above metal/Al ratios represent the lattice composition of the estuarine particles and
if these particles are the weathering products of basalts from which the laterites
formed, then the four metals Cu, Fe, Co and Ni were more refractory than Al in the
parent material.
On a volume basis, the percentages of particulate metal on the total metal
concentration (dissolved plus particulate) were essentially constant for Cd (~ 10 %),
Cu (~ 14 %), Co (~ 30 %) and Ni (~ 22 %) in the estuarine waters (S > 0), but with
generally higher values during the dry season. In contrast to this behavior of
Kapitel 4: Publikation 1
- 78 -
relatively constant particulate concentrations over the whole salinity range, the
particulate concentrations per volume of Fe, Al, Cu, Ni, Pb and Cd in Galveston Bay
all showed a general decrease (Wen et al., 2008).
0
20
40
60
80
100
0 20 40 60 80 100Fe (%)
Pb (%
)
Fig. 7. Ratio of the percentages of the particulate metal on the sum of the dissolved and particulate metal: Pb vs. Fe (open symbols: dry season, solid symbols: wet seasons; : WR 2006; : WWR 2006; : WR 2007; : WWR 2007; ▲: WR 2008; : WWR 2008).
The percentage of particulate Fe and Pb on the total concentration increased from
about 10 % to more than 80% along the salinity profile, which reflects the intense
conversion of Fe and Pb from solution to particles in the estuaries and the strong
preference of these two metals for particulate forms (Fig. 7). This also confirms the
assessment made before, that flocculation of Fe colloids and co-precipitation of Pb
were the dominant processes of particle-solution interaction for these metals in the
two study areas.
The freshwater concentrations of particulate metals as one endmember for the
mixing processes in the two estuaries were generally much lower than in typical
European and North-American estuaries with a more industrialized character (Balls,
1989, 1990), but they agreed with other Chinese estuaries (Table 3). For the marine
endmember with salinities S ≈ 30, neither considerable temporal nor spatial variability
was observed in the two estuaries, and the concentrations were comparable with
typical levels of particulate Fe, Cu, Co and Ni in the South China Sea (Ho et al.,
2007).
Kapitel 4: Publikation 1
- 79 -
Table 3. Particulate trace metals and SPM of East-Hainan estuaries (average values ± 1 standard deviation) in comparison to selected Chinese rivers (S = 0) and estuaries (S > 0) (averages or typical ranges; *: estimated from the figures; **: extract or leach).
Study area SPM Cd Cu Fe Pb Co Ni
Source mg/L nmol/g μmol/g mmol/g μmol/g μmol/g μmol/g
East Hainan
S=0 5.3±4.4 4.9±2.1 0.62±0.29 1.21±0.46 0.24±0.09 0.44±0.27 0.90±0.40
a 0<S<28 3.0±3.3 3.1±0.9 0.57±0.29 0.85±0.38 0.19±0.06 0.29±0.12 0.79±0.61
S>28 6.1±4.6 2.7±1.0 0.44±0.21 0.54±0.14 0.15±0.04 0.20±0.05 0.59±0.3
Daliaohe (S=0) 851 -- 0.80 0.93 0.69 1.42 2.91 b
Yalujiang (S=0) 127 5.6 0.62 0.69 0.27 0.23 0.63 c
Shuangtaizihe (S=0) 4651 6.9 0.60 0.78 0.40 0.35 0.86 c
Luanhe (S=0) 4633 7.4 0.86 0.90 0.33 0.29 0.88 c
Haihe (S=0) 3927 7.7 0.61 -- 0.37 -- 0.88 b
Huanghe (S=0) 25581 1.6 0.44 0.63 0.08 -- -- d
Huanghe (S=0) 2000-31500 1.5 0.42 0.67 0.08 0.24 0.69 e
Huanghe (S=0) 26829 1.5 0.42 0.67 0.08 0.24 0.69 b
Changjiang (S=0) 494 2.7 0.76 0.84 0.16 -- -- f
Changjiang (S=0) 539 2.6 0.98 0.93 0.24 0.32 2.11 b
Changjiang (S=0) 10-20000 3.6 0.98 0.93 0.19 0.32 1.09 c
Changjiang estuary (S>0) 100-1000 -- 0.8* -- 0.14* -- 1.0* g
Changjiang estuary (S>0)** 70 2* 1.1* 0.5* 0.2* 0.4* 2.6* h
Qiantangjiang (S=0) 187 2.5 1.41 -- 0.37 -- 1.58 b
Jiaojiang (S=0) 1254 6.7 0.57 0.65 0.26 0.29 0.79 c
Minjiang (S=0) 128 2.8 0.82 -- 0.30 -- -- b
Jiulongjiang (S=0) 211 4.4 0.62 1.03 0.29 0.39 1.38 b
Zhujiang (S=0) 348 5.3 0.77 -- 0.29 0.32 -- d
Zhujiang (S=0) 227 6.7 0.80 0.88 0.36 0.30 1.05 c
Tanshui River estuary (S>0)** 15* -- 0.35-7.9 0.19-1.2 -- 0.03-0.32 0.1-1.8 i
Tanshui River estuary (S>0) 10* 2-20* 1-11* -- 0.2-0.7* -- 1-6* j
Open South China Sea 0.4 -- 0.68 0.40 -- 0.25 0.46 k
(a): This study; (b): Zhang, 1995; (c): Zhang and Liu, 2002; (d): Qu and Yan, 1990; (e): Huang et al., 1992; (f): Martin and Meybeck, 1979; (g): Zhang, 1999; (h): Wang and Liu, 2003; (i): Fang and Lin, 2002; (j): Jiann et al., 2005; (k): Ho et al., 2007.
4.3.4. Distribution coefficients
The use of partition coefficients KD = CP / CS is a common approach for describing
solid-solution interactions; the KD, however, is not a true equilibrium coefficient, but
an empirical term which depends on various factors such as solution and particle
composition, speciation, analytical approach to particulate forms, ratio of colloidal
Kapitel 4: Publikation 1
- 80 -
forms to truly dissolved ions, DOC and SPM concentration, pH, temperature, and
others (Bourg, 1987; Comans and van Dijk, 1988; Balls, 1989; Turner, 1996; Hatje et
al, 2003). The situation is further complicated by the fact that suspended particles
represent mixtures of particles originating from the rivers entering the estuary and
from resuspended bottom sediments. For the samples of this study, the logarithms of
the distribution coefficient were plotted versus salinity in Fig. 8.
Cd
3
4
5
6
7
8
0 5 10 15 20 25 30 35Salinity
logK
D
Cu
3
4
5
6
7
8
0 5 10 15 20 25 30 35Salinity
Fe
3
4
5
6
7
8
0 5 10 15 20 25 30 35Salinity
logK
D
Pb
3
4
5
6
7
8
0 5 10 15 20 25 30 35Salinity
Co
3
4
5
6
7
8
0 5 10 15 20 25 30 35Salinity
logK
D
Ni
3
4
5
6
7
8
0 5 10 15 20 25 30 35Salinity
Fig. 8. Log10KD (L/kg) vs. salinity (open symbol: dry season, solid symbols: wet seasons; : WR 2006; : WWR 2006; : WR 2007; : WWR 2007; ▲: WR 2008; : WWR 2008).
Among the six elements under investigation, only the KD value of Cd clearly
decreased along the salinity profile in the estuary. This appears to result from
complexation of dissolved Cd by seawater anions forming highly stable Cd-chloro-
Kapitel 4: Publikation 1
- 81 -
complexes and from successful competition of seawater cations for sorption sites on
the SPM (e.g., Balls et al., 1994; Turner, 1996). In contrast, the KD’s of Cu, Ni and Co
showed neither temporal nor spatial variation nor any significant change with salinity.
Fe and Pb have the greatest affinity for the particulate phase with a log10KD increase
along the salinity profile from 5.4 to 8.0 L/kg for Fe and from 4.9 to 7.2 L/kg for Pb
during the wet seasons. Such a positive correlation of KD’s for Fe and Pb with salinity
and no relation to salinity for Cu and Ni was often observed in estuaries (e.g., Balls et
al., 1994). The higher KD’s for Fe and Pb in both estuaries during the dry season
were a consequence of its very low dissolved Fe and Pb and seemed to be related to
the specific properties of its much higher SPM loading. Several processes may
contribute to these peculiarities of the dry season, such as the additional input of
particles with differing surface properties from the drainage basin and/or from bottom
sediments, and the enhanced flocculation of colloidal Fe. An explanation is difficult,
because the dissolved Fe during the dry season is - per volume - more than one
order of magnitude lower than the Fe content in the particles which have a similar
composition at all seasons (Fig.6).
The general order of the log10KD for all samples was on average: Fe (6.0) > Pb
(5.7) > Co (5.3) > Ni (5.1) > Cu (4.8) ≈ Cd (4.8), and essentially agreed with the
results presented for the Seine estuary (Pb > Co > Cu > Zn > Ni > Cd; total digestion;
Chiffoleau et al., 1994) and for North Australian estuaries (Fe > Pb > Zn > Ni ≈ Cd ≈
Cu; extract at pH 2; Munksgaard and Parry, 2001). Although sorption processes of
the trace metals are sensitive for the source and the composition of the particles, and
although the KD is influenced by the method adopted for the particulate metal
determination (total destruction vs. partial extraction; Turner et al., 1992), the KD’s
obtained in this study differ from published data (e.g., Chiffoleau et al., 1994; Zhang,
1995; Kraepiel et al., 1997; Munksgaard and Parry, 2001; Tang et al., 2002) by less
than one order of magnitude.
In estuarine studies a so-called “particle concentration effect” is often observed,
which means that for a particular element the KD decreases with increasing SPM
loadings (e.g., Balls, 1989; Benoit and Rozan, 1999; Jiann et al., 2005). When the
KD’s of this study are plotted vs. SPM (not shown), the values for all investigated
metals form one cluster for the wet seasons at SPM loadings of 1-4 mg/L and a
second cluster for the dry season at SPM loadings of 8-12 mg/L (cf. Fig. 2). Possibly
owing to the overall relatively low SPM concentrations and the low particulate
concentrations in WR and WWR, no relationship at all was found between KD’s and
SPM for Cd, Cu, Co and Ni, while Fe and Pb even showed a slight increase. Also in
Kapitel 4: Publikation 1
- 82 -
the Changjiang KD’s of Cd, Cu and Ni did not change with increasing SPM
concentration between 40 and 350 mg/L, while the KD’s of Fe, Mn and Al showed
drastic declines as a consequence of their high ratio of particulate metal to dissolved
metal (Koshikawa et al., 2007). A stable concentration of particulate Cu, Ni and Pb in
the Changjiang relative to the SPM, which varied over several orders of magnitude,
was reported earlier by Zhang (1999).
4.3.5. Comparison of WR and WWR with other Chinese estuaries
This study asserted that all dissolved and particulate metals were similar or
somewhat lower (Co, Ni) in the WR with its low industrial and agricultural run-off than
in the WWR estuary, which is probably affected by aquacultural contamination plus
some municipal sewage along the river (Table1, Table 2). This is partly a
consequence of the exceptional characteristics during the wet season of 2008 in the
WWR, when not only DOC, but also dissolved Cu, Fe, Pb and Co at low salinities as
well as the particulate Cu, Co and Ni were pronouncedly higher than during the other
two expeditions. As also reported for dissolved arsenic (Balzer et al., 2013), this
points to extraordinary lateral inputs probably originating from adjacent aquaculture
ponds as shown for nutrients, dissolved organic matter and suspended particles by
Herbeck et al., 2013.
A comparison of the particulate metal concentrations with large Chinese rivers
demonstrates that the two Hainan estuaries contained similar levels of Cd, Cu, Pb,
Co and Ni (Table 3). The lower concentrations especially in the Huanghe and the
Changjiang probably originate from lower trace metal contents in soils of the northern
drainage areas (Zhang 1995). It is noteworthy that the particulate and dissolved Fe
concentrations in both study areas mostly exceeded those of the other estuaries
listed in Table 4. This is probably related to the generally high Fe content in the soils
of Hainan Island, although a small contamination - via atmospheric transport - by the
Shilu Iron Mine cannot be excluded,
The dissolved concentrations of the other trace metals in the freshwater part of the
WR and the WWR were similar (Cd, Ni, Cu) or higher (Fe, Pb, Co) than in the large
Chinese rivers (Table 4). The most striking difference between the Hainan estuaries
and the large rivers of mainland China is the orders of magnitude higher SPM loading
(Zhang, 1995; Zhang and Liu, 2002). Higher SPM, however, possibly fostered by
longer residence times only serve for a faster attainment of sorption equilibria, but
gives no general direction for concentration changes in the solution phase.
Kapitel 4: Publikation 1
- 83 -
Table 4. Dissolved trace metals of East-Hainan estuaries (average values ± 1 standard deviation) in comparison to selected Chinese rivers (S = 0) and estuaries (S > 0) (averages or typical ranges; *: estimated from the figures).
Study area SPM Cd Cu Fe Pb Co Ni
Source mg/L nmol/L nmol/L μmol/L nmol/L nmol/L nmol/L
East Hainan
S=0 5.3±4.4 0.03±0.03 10±6.8 4.1±3.6 0.88±0.65 2.2±1.6 5.9±4.0
a 0<S<28 3.0±3.3 0.07±0.03 9.6±5.7 2.2±2.7 0.59±0.49 2.7±3.0 7.7±3.2
S>28 6.1±4.6 0.11±0.03 6.4±3.4 0.14±0.13 0.10±0.08 0.75±0.40 4.9±1.3
Huanghe (S=0) 25600 0.02 22 0.35 0.14 0.21 6.9 b
Huanghe (S=0) 27000 0.03 20 0.24 0.13 0.30 7.5 c
Changjiang (S=0) 20-5000 0.03 26 0.55 0.26 -- 2.5 d
Changjiang estuary(S>0) 70 0.02-0.36 19-26 0.1-0.6 2.3-3.4 1.4-2.9 18-28 e
Minjiang (S=0) 128 0.04 14 -- 0.63 -- -- c
Jiulongjiang (S=0) 211 0.03 13 16 0.07 0.14 1.4 c
Tanshui River estuary (S>0) 15* -- 5-53 0.39-3.4 -- 0.3-6.1 7-310 f
Tanshui River estuary (S>0) 10* 0.20 60 -- 0.88 -- 138.0 g
Open South China Sea -- 0.01 0.8 0.19 -- -- 2.3 h
(a): This study; (b): Zhang et al., 1994; (c): Zhang, 1995; (d): Edmond et al., 1985; (e): Wang and Liu, 2003; (f): Fang and Lin, 2002; (g): Jiann et al., 2005; (h): Wen, et al., 2006.
With respect to the estuarine chemistry of the Chinese aquatic systems, detailed
information is only available for the Changjiang with its rather stable estuarine trace
metal distribution (Zhang, 1999; Wang and Liu, 2003) and about the subtropical
Tanshui (Fang and Lin, 2002; Jiann et al., 2005). Compared to these estuarine
systems, the tropical East-Hainan estuaries (S > 0) generally contained similar or
lower trace metal concentrations in the suspended particles (Table 3), and also
similar or lower concentrations for Cd, Cu, Pb and Ni, but higher levels of Fe in the
solution phase (Table 4). However, the concentration differences, the number of
regions for comparison and the amount of necessary subsidiary data are not large
enough to derive specific features for the estuarine chemistry of tropical East-Hainan.
The information compiled in Table 3 and Table 4 demonstrate, that on a global scale
the Chinese aquatic systems including the two Hainan estuaries have lower levels of
dissolved and particulate metals than the estuaries of countries with a longer history
of industrialization and with stronger signals of anthropogenic activities (Balls, 1989;
Zhang, 1995). The concentrations of the investigated trace metals in solution and in
Kapitel 4: Publikation 1
- 84 -
the particles suggest a low level of contamination especially in these two Hainan
estuaries.
4.4. Conclusions
This study provides data for dissolved and particulate cadmium, copper, iron, lead,
cobalt and nickel concentrations plus particulate aluminium along the salinity
gradients of two estuaries of the tropical East-Hainan Island during both the wet and
the dry season. The estuarine chemistry of the Wanquan River (WR) with its low
industrial and agricultural run-off and of the Wenchang/Wenjiao River (WWR), which
probably is affected by aquacultural contamination plus some municipal sewage, is
similar but not identical.
It was found that the dissolved trace metals mostly behaved conservatively. A positive
deviation from the theoretical dilution line was observed for dissolved cadmium, while
the converse situation applied for iron and lead. Seasonal variability was observed
during the dry season for dissolved iron and lead with a strong lowering and for
dissolved Cd with higher values. In both estuaries all six metals in particulate form
showed almost constant values with a tendency for slight decreases along the
salinity profile. The normalization to particulate Al, which itself decreased with salinity,
revealed specific particle properties during the different expeditions, e.g., Cu-rich
particles during the wet season 2008 in WWR, which coincided with very high DOC
levels and generally low percentages of particulate metals on the total metal
concentration (i.e., sum of dissolved and particulate) per unit volume. In general, the
dynamics of Fe chemistry dominated the distribution of Pb in all forms and partly also
of Co as revealed by correlations in the dissolved state and in their ratio of particulate
metal to the total metal per unit volume.
The distribution coefficients of the trace metals KD = CP / CS decreased in the order
Fe > Pb > Co > Ni > Cu ≈ Cd. Except for the expected KD decrease of Cd with
increasing salinity, the KD´s of Cu, Co and Ni were constant, and Fe and Pb exhibited
slightly positive correlations with salinity. There was no "particle concentration effect",
i.e., a negative dependence of the KD on the SPM level.
The most striking difference between the large Chinese rivers especially from
temperate latitudes and the two Hainan water systems is the extremely low SPM in
the latter ones. Together with its low levels of particulate trace metals, this entails the
possibility that equilibrium desorption might be so small in absolute amounts that the
transfer is hardly detectable in the solution phase.
Kapitel 4: Publikation 1
- 85 -
One of the aims of this study was to use the distributions of environmentally relevant
trace metals during estuarine mixing for assessing the pollution state of the East-
Hainan estuaries. In most cases, both particulate and dissolved metal concentrations
in the Wenchang/Wenjiao River estuary were higher than in the Wanquan River
estuary, which is probably due to diffuse inputs from aquaculture ponds to the WWR
system. The trace metal levels of these East-Hainan aquatic systems were similar to
large rivers in mainland China and showed no significant contamination when
compared to rivers and estuaries in Europe and North America. This is especially
important with respect to the recently observed health decline of the coral reefs at the
east coast of Hainan.
Acknowledgements
The authors wish to thank Dr. Dao-Ru Wang and Dr. Larissa Dsikowitzky for their
coordinating and managing efforts during the study. Appreciation is extended to the
colleagues from the Centre for Tropical Marine Ecology (ZMT) of Germany, the
Hainan Provincial Marine Development and Design Institute (HNMDDI) of China, the
East China Normal University (ECNU) and the Ocean University of China (OUC) for
their help in the field and during laboratory works. We thank Dr. Uwe Schüßler for his
help with the ICP-MS analyses. We are especially thankful for the constructive
criticism of two anonymous reviewers. We gratefully acknowledge the funding by the
"Bundesministerium für Bildung und Forschung" (BMBF, German Federal Ministry of
Education and Research) at Berlin under the contract number 03F0457 C and the
financial support from the Ministry of Science and Technology (MOST) through the
contract No. 2007DFB20380 and under the agreement between the State Oceanic
Administration of China (SOA) and the BMBF.
Kapitel 4: Publikation 1
- 86 -
References
Balzer, W., 1982. On the distribution of iron and manganese at the sediment/water interface: thermodynamic versus kinetic control. Geochimica et Cosmochimica Acta 46, 1153-1161.
Balzer, W., Boehler, E., Tang, X.-L., Ren, J.-L., Zhang, J., Wang, D.-R., 2013. Arsenic in solution, colloidal and particulate phases of East-Hainan estuaries. Continental Shelf Research 57, 73-81.
Balls, P.W., 1989. The partition of trace metals between dissolved and particulate phases in European coastal waters: a compilation of field data and comparison with laboratory studies. Netherlands Journal of Sea Research 23, 7-14.
Balls, P.W., 1990. Distribution and composition of suspended particulate material in the Clyde estuary and associated sea lochs. Estuarine, Coastal and Shelf Science 30, 475-487.
Balls, P.W., Laslett, R.E., Price, N.B., 1994. Nutrient and trace metal distributions over a complete semi-diurnal tidal cycle in the Forth estuary, Scotland. Netherlands Journal of Sea Research 33, 1-17.
Benoit, G., Oktay-Marshall, S.D., Cantu, A., II, Hood, E.M., Colemann, C.H., Corapcioglu, M.O., Santschi, P.H., 1994. Partitioning of Cu, Pb, Ag, Zn, Fe, Al, and Mn between filter-retained particles, colloids, and solution in six Texas estuaries. Marine Chemistry 45, 307-336.
Bourg, A.C.M., 1987. Trace metal adsorption modelling and particle-water interactions in estuarine environments. Continental Shelf Research 7, 1319-1332.
Boyle, E.A., Edmond, J.M., Sholkovitz, E.R., 1977. The mechanism of iron removal in estuaries. Geochimica et Cosmochimica Acta, 41, 1313-1324.
Boyle, E.A., Huested, S.S., Grant, B., 1982. The chemical mass balance of the Amazon plume - II. Copper, nickel, and cadmium. Deep-Sea Research / A, Oceanographic research papers 29, 1355-1364.
Campbell, J.A., Whitelaw, K., Riley, J.P., Head, P.C., Jones, P.D., 1988. Contrasting behaviour of dissolved and particulate nickel and zinc in a polluted estuary. The Science of the Total Environment 71, 141-155.
Chaudry, M.A. and Zwolsman, J.J.G., 2008. Seasonal dynamics of dissolved trace metals in the Scheldt Estuary: relationship with redox conditions and phytoplankton activity. Estuaries and Coasts 31, 430-443.
Chester, R., 1990. Marine Geochemistry. Unwin Hyman, London, 698 pp.
Chiffoleau, J.-F., Cossa, D., Auger, D., Truquet, I., 1994. Trace metal distribution, partition and fluxes in the Seine estuary (France) in low discharge regime. Marine Chemistry 47, 145-158.
Comans, R.N.J. and van Dijk, C.P.J., 1988. Role of complexation processes in cadmium mobilization during estuarine mixing. Nature, 336, 151-154.
Comber, S.D.W., Gunn, A.M., Whalley, C., 1995. Comparison of the partitioning of trace metals in the Humber and Mersey estuaries. Marine Pollution Bulletin 30, 851-860.
Dai, M.-H., Martin, J.-M., Hong, H.-S, Zhang, Z.-L., 2000. Preliminary study on the dissolved and colloidal organic carbon in the Zhujiang River estuary. Chinese Journal of Oceanology and Limnology 18, 265-273.
Danielsson, L.-G., Magnusson, B., Westerlund, S., 1978. An improved metal extraction procedure for the determination of trace metals in sea water by atomic absorption spectrometry with electrothermal atomization. Analytica Chimica Acta 98, 47-57.
Danielsson, L.-G., Magnusson, B., Westerlund, S., Zhang, K.-R., 1983. Trace metals in the Göta river estuary. Estuarine, Coastal and Shelf Science 17, 73-85.
Edmond, J.M., Spivack, A., Grant, B.C., Hu, M.-H., Chen, Z.-X., Chen, S., Zeng, X.-S., 1985. Chemical dynamics of the Changjiang estuary. Continental Shelf Research 4, 17-36.
Kapitel 4: Publikation 1
- 87 -
Elbaz-Poulichet, F., Holliger, P., Huang, W.-W., Martin, J.M., 1984. Lead cycling in estuaries, illustrated by the Gironde estuary, France. Nature 308, 409-414.
Elbaz-Poulichet, F., Martin, J.M., Huang, W.-W., Zhu, J.-X., 1987. Dissolved Cd behaviour in some selected French and Chinese estuaries. Consequences on Cd supply to the ocean. Marine Chemistry 22, 125-136.
Elbaz-Poulichet, F., Garnier, J.-M., Guan, D.-M., Martin, J.-M., Thomas, A.J., 1996. The conservative behaviour of trace metals (Cd, Cu, Ni and Pb) and As in the surface plume of stratified estuaries: example of the Rhône River (France). Estuarine, Coastal and Shelf Science 42, 289-310.
Fang, T.-H. and Lin, C.-L., 2002. Dissolved and particulate trace metals and their partitioning in a hypoxic estuary: the Tanshui Estuary in Northern Taiwan. Estuaries and Coasts 25, 598-607.
Ge, C.-D., Slaymaker, O., Pedersen, T.F, 2003. Change in the sedimentary environment of Wanquan River Estuary, Hainan Island, China. Chinese Science Bulletin 48, 2357-2361.
Guo, W., He, M.-C., Yang, Z.-F., Lin, C.-Y., Quan, X.-C., Men, B., 2009. Distribution, partitioning and sources of polycyclic aromatic hydrocarbons in Daliao River water system in dry season, China. Journal of Hazardous Materials 164, 1379-1285.
Hatje, V., Birch, G.F., Hill, D.M., 2001. Spatial and temporal variability of particulate trace metals in Port Jackson Estuary, Australia. Estuarine, Coastal and Shelf Science 53, 63-77.
Hatje, V., Payne, T.E., Hill, D.M., McOrist, G., Szymczak, R., 2003. Kinetics of trace element uptake and release by particles in estuarine waters: effects of pH, salinity, and particle loading. Environment International 29, 619-629.
Harper, D.J., 1991. The distribution of dissolved cadmium, lead and copper in the Bristol Channel and the outer Severn estuary. Marine Chemistry 33, 131-143.
Herbeck, L.S., Unger, D., Krumme, U., Liu, S.-M., Jennerjahn, T.C., 2011. Typhoon-induced precipitation impact on nutrient and suspended matter dynamics of a tropical estuary affected by human activities in Hainan, China. Estuarine Coastal and Shelf Science 93, 375–388.
Herbeck, L.S., Unger, D., Wu, Y., Jennerjahn, T.C., 2013. Effluent, nutrient and organic matter export from shrimp and fish ponds causing eutrophication in coastal and back-reef waters of NE Hainan, tropical China. Continental Shelf Research 57, 92-104.
Ho, T.-Y., Wen, L.-S., Y, C.-F., Lee, D.-C., 2007. The trace-metal composition of size-fractionated plankton in the South China Sea: biotic versus abiotic sources. Limnology and Oceanography 52, 1776-1788.
Hsieh, C.-M. and Zhong, G.-F., 1990. Hainan - the island of South Sea: a new province in China. GeoJournal 20, 385-391.
Huang, W.-W., Zhang, J., Zhou, Z.-H., 1992. Particulate element inventory of the Huanghe (Yellow River): a large, high-turbidity river. Geochimica et Cosmochimica Acta 56, 3669-3680.
Jiann, K.-T., Wen, L.-S., Santschi, P.H., 2005. Trace metal (Cd, Cu, Ni and Pb) partitioning, affinities and removal in the Danshuei River estuary, a macro-tidal, temporally anoxic estuary in Taiwan. Marine Chemistry 96, 293-313.
Khalid, R.A., Patrick, W.H., Gambrell, R.P., 1978. Effect of dissolved oxygen on chemical transformations of heavy metals, phosphorus, and nitrogen in an estuarine sediment. Estuarine and Coastal Marine Science 6, 21-35.
Koshikawa, M., K., Takamatsu, T., Takada, J., Zhu, M.-Y., Xu, B.-H., Chen, Z.-Y., Murakami, S., Xu, K.-Q., Watanabe, M., 2007. Distributions of dissolved and particulate elements in the Yangtze estuary in 1997-2002: background data before the closure of the Three Gorges Dam. Estuarine, Coastal and Shelf Science 71, 26-36.
Kraepiel, A.M.L., Chiffoleau, J.-F., Martin, J.-M., Morel, F.M.M., 1997. Geochemistry of trace
Kapitel 4: Publikation 1
- 88 -
metals in the Gironde estuary. Geochimica et Cosmochimica Acta 61, 1421-1436.
Li, J.-Y., Xu, R.-K., Zhang, H., 2012. Iron oxides serve as natural anti-acidification agents in highly weathered soils. Journal of Soils and Sediments 12, 876-887.
Li, R.-H., Liu, S.-M., Zhang, G.-L., Ren, J.-L., Zhang, J., 2013. Biogeochemistry of nutrients in an estuary affected by human activities: the Wanquan River estuary, eastern Hainan Island, China. Continental Shelf Research 57, 18-31.
Louis, Y., Garnier, C., Lenoique, V., Omanović, D., Mounier, S., Pižeta, I., 2009. Characterisation and modelling of marine dissolved organic matter interactions with major and trace cations. Marine Environmental Reseach 67, 100-107.
Ma, J.-L., Wei, G.-J., Xu, Y.-G., Long, W.-G., Sun, W.-D., 2007. Mobilization and re-distribution of major and trace elements during extreme weathering of basalt in Hainan Island, South China. Geochimica et Cosmochimica Acta 71, 3223-3237.
Martin, J.-M. and Meybeck, M., 1979. Elemental mass-balance of material carried by major world rivers. Marine Chemistry 7, 173-206.
Martino, M., Turner, A., Nimmo, M., Millward, G.E., 2002. Resuspension, reactivity and recycling of trace metals in the Mersey Estuary, UK. Marine Chemistry 77, 171-186.
Martino, M., Turner, A., Nimmo, M., 2004. Distribution, speciation and particle-water interactions of nickel in the Mersey Estuary, UK. Marine Chemistry 88, 161-177.
Munksgaard, N.C. and Parry, D.L., 2001. Trace metals, arsenic and lead isotopes in dissolved and particulate phases of North Australian coastal and estuarine seawater. Marine Chemistry 75, 165-184.
Ni, H.-G., Lu, F.-H., Luo, X.-L., Tian, H.-Y., Zeng, E.-Y., 2008. Riverine inputs of total organic carbon and suspended particulate matter from the Pearl River Delta to the coastal ocean off South China. Marine Pollution Bulletin 56, 1150-1157.
O'Connor, D.J. and Connolly, J.P., 1980. The effect of concentration of adsorbing solids on the partition coefficient. Water Research 14, 1517-1523.
Ouseph, P.P., 1992. Dissolved and particulate trace metals in the Cochin estuary. Marine Pollution Bulletin 24, 186-192.
Paalman, M.A.A. and Van der Weijden, C.H., 1992. Trace metals in suspended matter from the Rhine/Meuse estuary. Netherlands Journal of Sea Research 29, 311-321.
Paucot, H. and Wollast, R., 1997. Transport and transformation of trace metals in the Scheldt estuary. Marine Chemistry 58, 229-244.
Qu, C.-H. and Yan, R.-E., 1990. Chemical composition and factors controlling suspended matter in three major Chinese rivers. The Science of the Total Environment 97/98, 335-346.
Roux, L.L., Le Roux, S., Appriou, P., 1998. Behaviour and speciation of metallic species Cu, Cd, Mn and Fe during estuarine mixing. Marine Pollution Bulletin 36, 56-64.
Salomons, W., Rooij, N.M., Kerdijk, H., Bril, J., 1987. Sediments as a source for contaminants? Hydrobiologia 149, 13-30
Shiller, A.M. and Boyle, E.A., 1991. Trace elements in the Mississippi River Delta outflow region: behavior at high discharge. Geochimica et Cosmochimica Acta 55, 3241-3251.
Sholkovitz, E.R., 1978. The flocculation of dissolved Fe, Mn, Al, Cu, Ni, Co and Cd during estuarine mixing. Earth and Planetary Science Letters 41, 77-86.
Sholkovitz, E.R., Boyle, E.A., Price, N.B., 1978. The removal of dissolved humic acids and iron during estuarine mixing. Earth and Planetary Science Letters 40, 130-136.
Takara, H., Aono, T., Tagami, K., Uchida, S., 2010. Processes controlling cobalt distribution in two temperate estuaries, Sagami Bay and Wakasa Bay, Japan. Estuarine, Coastal and Shelf Science 89, 294-305.
Tang, D.-G., Warnken K.W., Santschi, P.H., 2002. Distribution and partitioning of trace metals
Kapitel 4: Publikation 1
- 89 -
(Cd, Cu, Ni, Pb, Zn) in Galveston Bay waters. Marine Chemistry 78, 29-45.
Tang, A.-K., Liu, R.-H., Ling, M., Xu, L.-G., Wang, J.-Y., 2010. Distribution characteristics and controlling factors of soluble heavy metals in the Yellow River Estuary and adjacent sea. Procedia Environmental Sciences 2, 1193-1198.
Turner, A., Millward, G.E., Schuchardt, B., Schirmer, M., Prange, A., 1992. Trace metal distribution coefficients in the Weser Estuary (Germany). Continental Shelf Research 12, 1277-1292.
Turner, A., 1996. Trace-metal partitioning in estuaries: importance of salinity and particle concentration. Marine Chemistry 54, 27-39.
Unger, D., Herbeck, L.S., Li, M., Bao, H.-Y., Wu, Y., Zhang, J., Jennerjahn, T.C., 2013. Sources, transformation and fate of particulate amino acids and hexosamines under varying hydrological regimes in the tropical Wenchang/Wenjiao Rivers and estuary, Hainan, China. Continental Shelf Research 57, 44-58.
Van der Weijden, C.H., Arnoldus, M. J. H. L., Meurs, C.J., 1977. Desorption of metals from suspended material in the Rhine estuary. Netherlands Journal of Sea Research 11, 130-145.
Waeles, M., Riso, R.D., Maguer, J.-F., Le Corre, P., 2004. Distribution and chemical speciation of dissolved cadmium and copper in the Loire estuary and North Biscay continental shelf, France. Estuarine, Coastal and Shelf Science 59, 49-57.
Wang, C.-Y., Wang, X.-L., Wang, B.-D., Zhang, X.-Y., Zhu, C.-J., 2009. Level and fate of heavy metals in the Changjiang estuary and its adjacent waters. Oceanology 49, 64-72.
Wang, Z.-L. and Liu, C.-Q., 2003. Distribution and partition behavior of heavy metals between dissolved and acid-soluble fractions along a salinity gradient in the Changjiang Estuary, eastern China. Chemical Geology 202, 383-396.
Wen, L.-S., Jiann, K.-T., Santschi, P.H., 2006. Physicochemical speciation of bioactive trace metals (Cd, Cu, Fe, Ni) in the oligotrophic South China Sea. Marine Chemistry 101, 104-129.
Wen, L.-S., Warnken, K.W., Santschi, P.H., 2008. The role of organic carbon, iron, and aluminium oxyhydroxides as trace metal carriers: comparison between the Trinity River and the Trinity River Estuary (Galveston Bay, Texas). Marine Chemistry 112, 20-37.
Windom, H., Smith, R.,Jr., Rawlinson, C., Hungspreugs, M., Dharmvanij, S., Wattayakorn, G., 1988. Trace metal transport in a tropical estuary. Marine Chemistry 24, 293-309.
Wood, T.M., Baptista, A.M., Kuwabara, J.S., Flegal, A.R., 1995. Diagnostic modeling of trace metal partitioning in south San Fransisco Bay. Limnology and Oceanography 401, 345-358.
Yang, M. and Sañudo-Wilhelmy, S.A., 1998. Cadmium and manganese distributions in the Hudson River estuary: interannual and seasonal variability. Earth and Planetary Science Letters 160, 403-418.
Zeng, Z.-X. and Zeng, X.-Z., 1989. Physicogeography of the Hainan Island. Science Press, Beijing. (in Chinese)
Zhang, J., 1995. Geochemistry of trace metals from Chinese river/estuary systems: an overview. Estuarine, Coastal and Shelf Science 41, 631-658.
Zhang, J., 1999. Heavy metal compositions of suspended sediments in the Changjiang (Yangtze River) estuary: significance of riverine transport to the ocean. Continental Shelf Research 19, 1521-1543.
Zhang, J., Huang, W.-W., Wang, J.-H., 1994. Trace-metal chemistry of the Huanghe (Yellow River), China. Examination of the data from in situ measurements and laboratory approach. Chemical Geology 114, 83-94.
Zhang, J. and Liu, C.-L., 2002. Riverine composition and estuarine geochemistry of particulate metals in China - weathering features, anthropogenic impact and chemical fluxes. Estuarine, Coastal and Shelf Science 54, 1051-1070.
Kapitel 4: Publikation 1
- 90 -
Zhang, E.-R. and Zhang, J., 2003. Effect of pH on adsorption of several metals to suspended sediment in the Changjiang River Estuary. Oceanologia et Limnologia Sinica 34, 267-273.
Zhu, D.-K., Yin, Y., Martini, I.P., 2005. Geomorphology of the Boao coastal system and potential effects of human activities - Hainan Island, South China. Journal of Geographical Sciences 15, 187-198.
Zwolsman, J.J.G., van Eck, G.T.M., van der Weijden, C.H., 1997. Geochemistry of dissolved trace metals (cadmium, copper, zinc) in the Scheldt estuary, southwestern Netherlands: impact of seasonal variability. Geochimica et Cosmochimica Acta 61, 1635–1652.
Zwolsman, J.J.G.and van Eck, G.T.M., 1999. Geochemistry of major elements and trace metals in suspended matter of the Scheldt estuary, southwest Netherlands. Marine Chemistry 66, 91–111.
Kapitel 5: Entwurf zur Publikation 2
- 91 -
5. Entwurf zur Publikation 2:
Transport of trace metals in colloidal form through estuarine systems of East-Hainan
Abstract
Cross flow ultrafiltration in conjunction with high resolution inductively coupled
plasma mass spectrometry was used to estimate the colloidal distribution of iron,
lead, cadmium, copper, cobalt and nickel in two estuarine systems of East Hainan.
Metal concentrations along the salinity gradient were determined in three operatically
defined size fractions: 0.45 μm - 10 kDa (large colloids), 5 kDa - 10 kDa (small
colloids) and < 5 kDa (assumed to be in true solution) in order to investigate the
changes of trace metals. Dissolved iron and lead were found to exist mainly as large
colloids, while dissolved cadmium, copper and nickel were present in low molecular
size fractions. Both the larger and the smaller size fractions were important for
dissolved cobalt. Modifications in size distribution during the estuarine mixing are
observed for all examined metals with the involvement of estuarine processes such
as coagulation, degradation and dissociation. For all investigated metals the truly
dissolved fraction becomes more important with increasing salinity. Furthermore, the
difference of colloidal size fraction between the two estuarine systems was discussed.
Keywords: Trace metals; Estuaries; Colloids; East Hainan
Kapitel 5: Entwurf zur Publikation 2
- 92 -
5.1. Introduction
An estuarine environment may be simply defined as one in which sea water is
substantially diluted with fresh water entering from land drainage. These two types of
water have different compositions, and as a result estuaries are very complex
environments in which the boundary conditions are extremely variable in both space
and time (Chester, 1990). It is now widely accepted that colloids are important in
controlling the cycling of trace metals that may either be scavenged e.g., onto Fe-
oxyhydroxides or complexed with large aggregates such as organic matter in the
estuarine environment (Powell et al., 1996), which can be scavenged in the estuary.
As a consequence, such colloidally mediated scavenging results in removal of both
bulk matter and associated trace components (Gustafsson et al., 1996).
The general pattern of colloidal partitioning of trace metals in the water column
depends primarily on the metal (Benoit et al., 1994; Farag, et al., 2007). Various size
fractionation studies indicate different proportions of colloidal forms for different trace
metals. For example, particle reactive elements such as Fe and Pb probably have
very small truly dissolved fractions (Ren et al., 2010). In contradistinction, cadmium
should have larger fractions in the truly dissolved pool via formation of strong chloride
complexes (Wen et al., 1999). The complexation of cobalt with dissolved organic
carbon was reported (Collins and Kinsela, 2010). Competition between different size
ligands may control the colloidal distribution of copper and nickel, which form strong
organic complexes (Wang et al., 2003).
Little is known about the trace metals in colloidal form in the estuarine systems of
East Hainan. The aims of this study were therefore to investigate the size distribution
of colloidal trace metals as well as their behavior in the Wanquan River and
Wenchang/Wenjiao River during estuarine mixing.
Although dissolved, colloidal and particulate phases are purely operational and are
not necessarily related to real differences in structure or to environmental or chemical
behavior (Lead and Wilkinson, 2007), present knowledge of natural water colloids is
largely based on the physical separation techniques, of which cross flow filtration
(tangential flow filtration; CFF) is currently widely used to provide concentrates of
colloidal material for further analyses. This approach has made it possible to
separate colloidal size material from true solutes (Buesseler et al., 1996; Guo et al.,
2000a). An exemplary representation of the CFF apparatus which was also used
during this study is shown in Fig. 1. The colloid concentration is given by: CCol = (CR-
CP)/CF, where CR is the retentate concentration, CP is the permeate concentration
Kapitel 5: Entwurf zur Publikation 2
- 93 -
and CF is the concentration factor. The CF is defined as: CF = (VR+VP)/ VR, where VR
is the volume of retentate and VP is the volume of permeate. A high CF can minimize
the retention at the membrane surface (Guo and Santschi, 1996), but a breakthrough
becomes more significant (Dai et al., 1998). On the other side, a low CF reduces the
coagulation of colloids at the membrane surface (Waeles et al., 2008), but the
recovery will be more sensitive for the system performance (Whitehouse et al., 1990).
There is still no general consensus as to what the optimal CF should be for
ultrafiltration (e.g., Guo et al., 2000b; Kottelat et al., 2008; Batchelli et al., 2009).
Fig. 1. Cross-flow filtration apparatus (source: www.sartorius-stedim.com).
Colloid size is defined in a range of 1 nm to 1 μm (Buffle and Leppard, 1995) with 1
nm ≈ 1000 Daltons (1 KDa; e.g., Wen et al., 1999). But the exact molecular weight or
size cut-off of the ultrafilter is not very sharp (Buesseler et al., 1996). To clarify the
notation for the present work, the size fraction between 10 kDa and 0.45 μm is
operationally defined as the high molecular weight colloid fraction (HMW), between 5
kDa and 10 kDa as the low molecular weight colloid fraction (LMW), and the
permeate from the ultrafiltration (< 5 kDa) as the truly dissolved fraction (UD).
5.2. Materials and Methods
5.2.1. Study area
The Wanquan River and the Wenchang/Wenjiao River estuaries (Fig. 2) are
important estuaries of East-Hainan which are both located at the northern part of the
tropical zone with a humid warm climate (Zhu et al., 2005). About 80% of the annual
rainfall in this region occurs in the period from May to November (Ma et al., 2007).
The Wanquan River (in the following: WR) is the third largest river in Hainan Island
Kapitel 5: Entwurf zur Publikation 2
- 94 -
with a drainage area of 3693 km2, a total length of 156.6 km and a mean annual
discharge of 163.9 m3/s (Ge, et al., 2003). The Wenchang River (drainage area 381
km2, total length 37 km, mean annual discharge 9.1 m3/s) and the Wenjiao River
(drainage area 522 km2, total length 56 km, mean annual discharge 11.6 m3/s) empty
into the Bamen Bay (Zeng and Zeng, 1989) and form the Wenchang/Wenjiao River
estuary (in the following: WWR). Both estuaries have a micro-tidal, irregular diurnal
tidal regime with a mean range of about 0.7-0.8 m (Zhu et al., 2005). The salinity
intrusion may extend 5 kilometers from the Yudai Sand Barrier into the WR and from
the Bamen Bay end into the WWR, respectively. The WWR is a shallow lagoon
region (< 2 m; except for the ship channel) with a low flow rate, as compared with the
WR, in which the position of the mix zone is highly variable due to changes of river
discharge rate, tides and wind.
Fig. 2. Wenchang/Wenjiao River estuary (a) and Wanquan River estuary (b).
These two estuarine systems of East Hainan are distinguished by their extremely low
suspended particulate matter and their low levels of particulate trace metals
compared to large Chinese rivers (< 15 mg/L; Fu et al., 2013). Also dissolved trace
metal levels of these two aquatic systems are comparable to the rivers and estuaries
in Europe and North America (< 15 μmol/L for Fe, < 2.5 nmol/L for Pb, < 0.15 nmol/L
for Cd, < 15 nmol/L for Cu except 2 abnormalities, < 9 nmol/L for Co except 1
abnormality and < 20 nmol/L for Ni; Fu et al., 2013). But the nutrient level in these
two estuaries, especially in the WWR where is dominated by aquaculture
contamination, is higher than the average global level (Herbeck et al., 2013; Li et al.,
2013). Also dissolved organic carbon (DOC) and dissolved trace metals in the WWR
Kapitel 5: Entwurf zur Publikation 2
- 95 -
are higher than in the WR (Fu et al., 2013).
5.2.2. Sampling and filtration
Surface water samples in the WR and the WWR were collected in July/August 2008.
Low-density polyethylene (LDPE) bottles (1 L, Nalgene®) were filled by hand with
surface water from a depth of 0.3 m upstream of the ship. In the laboratory, the water
samples were drawn through acid-cleaned 0.45 μm polycarbonate filters (pre-
weighed) within 24 hours before they were further separated into 30 kDa - 0.45 μm,
10-30 kDa, 5-10 kDa and < 5 kDa fractions using CFF with a flow rate of 500 mL/min
(2 bar). The cascade way was carried out to separate the colloid size fractions. This
method, in contrast to the parallel one, minimizes aggregation of colloids at the
membrane surface (Waeles et al., 2008). About 1 liter of the 0.45 μm filtrates were
first ultrafiltrated through a 30 kDa polyethersulfone membrane using a cross flow
filtration device (Vivaflow® 200, Sartorius; with an active membrane area of 200 cm2).
The permeate was used for the second ultrafiltration step employing a 10 kDa
polyethersulfone membrane. Eventually, the permeate from the second ultrafiltration
step was ultrafiltrated using a 5 kDa polyethersulfone membrane. A schematic
description of the filtration/ultrafiltration procedure is shown in Fig. 3. The average
CF’s ranged from 4.5 to 7.5. The filtrates, retentates and permeates were weighted
before the next step began. After 10 mL of each retentate/permeate was filled into a
pre-combusted glass ampoule with addition of phosphoric acid for dissolved organic
carbon measurements (Balzer et al., 2012), the sub-samples were filled into LDPE
bottles (Nalgene®) and acidified to pH 2 with concentrated nitric acid. The sample
bottles were wrapped in polyethylene bags and stored at room temperature. Before
use the CFF devices were washed with HCl (0.1 %, 30 min) and EDTA (0.02 mol/L,
30 min) and subsequently rinsed with Milli-Q® water. The environmental parameters
such as salinity and pH value were measured on the spot using a multi-sensor-
device (MultiLine P3, WTW).
Kapitel 5: Entwurf zur Publikation 2
- 96 -
Fig. 3. Flow diagram of the sample filtration procedure (SPM: suspended particular matter; Perm: permeat; Ret: retentat).
5.2.3. Pretreatment and measurement
Sample work-up was performed in a clean-air room (class 1000) with additional clean
benches of class 100. The sub-samples were first irradiated in a UV-digester for 2
hours with addition of hydrogen peroxide (1 mL per 100 mL sample) in order to
destroy the organic complexes, before the trace metals were extracted. The
extraction/backextraction procedure was developed by our workgroup based on the
work of e.g., Danielsson et al. (1978). At first, 40-100 g sample (depending on the
metal concentration) was weighed into a fluorinated ethylene propylene (FEP)
separatory funnel (Nalgene®). Ammonium acetate buffer (pH ~ 9) was added to
adjust the pH value of the sample to 4.0-4.5. The sample was well shaken (15 min)
after 500 μL APDC/NaDETC solution (0.06 mol/L) and 10 mL Freon were added.
Following the phase separation the lower organic phase was drawn off into an acid-
cleaned polypropylene tube. After addition of 100 μL concentrated nitric acid the tube
was shaken by hand for 3 min. Finally, 1900 μL Milli-Q® water was added and the
tube was again shaken by hand for 2 min for back-extraction. After the phase
Kapitel 5: Entwurf zur Publikation 2
- 97 -
separation, the upper aqueous extract was removed into an acid-cleaned sample
tube using a pipette. The metal concentrations were determined by high resolution
inductively coupled plasma mass spectrometry (HR-ICP-MS; Element 2, Thermo
Scientific).
5.2.4. CFF blanks and mass balance
Fe: y = 0.6422xR2 = 0.92
0
4
8
12
0 4 8 12TD
HM
W+L
MW
+UD
WRWWR
Pb: y = 0.5845xR2 = 0.8629
0.0
0.5
1.0
1.5
2.0
0.0 0.5 1.0 1.5 2.0TD
Cd: y = 0.6506xR2 = 0.678
0.00
0.04
0.08
0.12
0.00 0.04 0.08 0.12TD
HM
W+L
MW
+UD
Cu: y = 0.7884xR2 = 0.9025
0
5
10
15
0 5 10 15TD
Co: y = 0.8801xR2 = 0.944
0
1
2
3
4
5
0 1 2 3 4 5TD
HM
W+L
MW
+UD
Ni: y = 0.7924xR2 = 0.8184
0
5
10
15
0 5 10 15TD
Fig. 4. Colloidal and truly dissolved concentration vs. total dissolved concentration (nmol/L; for Fe μmol/L).
To produce CFF blanks, new CFF devices were pre-washed with HCl (0.1 %, 30 min)
and EDTA (0.02 mol/L, 30 min) and subsequently rinsed with Milli-Q® water. Milli-Q®
water was ultrafiltrated into 4 colloidal fractions (30 kDa - 0.45 μm; 10-30 kDa; 5-10
kDa and < 5 kDa). All sub-samples were analyzed using ICP-MS for trace metal
concentrations. Since elevated Cu concentration occurred, the devices were pre-
Kapitel 5: Entwurf zur Publikation 2
- 98 -
washed once again with HCl and EDTA and additional Milli-Q® samples were
ultrafiltrated and analyzed. All trace metal concentrations of the blanks were less than
1 % of the sample concentrations.
Since there is no standard reference material to quantify the accuracy of the
ultrafiltration analysis, a concentration mass balance between the sum of all size
fractions (HMW + LMW + UD) and the 0.45 μm filtrated sample (TD) was calculated.
For most samples, the mass balance was < 100% indicating a loss of material
probably due to scavenging onto the filter membrane or sample reservoir (Martin et
al., 1995; Powell et al., 1996). Also the propagation of sampling and analytical errors
of the sub-samples, of the UD samples in particular, may cause a deviation of the
mass balance from 100%. Especially for lower concentrations, a small deviation may
create a remarkable aberrance in the mass balance. The available data sets were
therefore operationally defined with a mass balance within a range of 75% ± 30%,
55% ± 20%, 65% ± 20%, 80% ± 20%, 100% ± 30% and 80% ± 20% for Fe, Pb, Cd,
Cu, Co and Ni, respectively. In total 20 analyzed samples, 12 for Fe, 11 for Pb, 15 for
Cd, 18 for Cu, 17 for Co and 12 for Ni have been selected, whose mass balances
were in the respective defined ranges. The calculated coefficients of determination
showed generally good correlation between the measured and the calculated total
dissolved concentrations (Fig. 4). Theoretically, the curves must have a slope of 1,
which indicates a mass balance of 100%. But indeed, all of the slopes resulted
smaller, which could be attributed to the sorption to the membrane as discussed.
5.3. Results and discussion
5.3.1. Iron
The dissolved Fe in the WR and the WWR was found mainly in the HMW fraction
with over 95% on average of the calculated TD in both study areas (Fig. 5). The LMW
and UD fractions with less than 5% in total of the calculated TD appear not to be
relevant in both study areas. The affinity of Fe to large colloids in the fresh water is
well known (Pham and Garnier, 1998; Ren et al., 2010). Also the loss of the colloidal
Fe through flocculation during the estuarine mixing was widely reported (e.g.,
Kraepiel et al., 1997; Nowostawska et al., 2008). This phenomenon bases on the fact,
that riverine colloidal Fe-oxyhydroxides closely associated with organic matter
aggregate to form larger particles due to interactions between the negatively charged
particles and Mg2+ and Ca2+ ions introduced by the seawater end member (e.g.,
Boyle et al., 1977; Nowostawska et al., 2008). As a consequence, the Fe ratios of the
Kapitel 5: Entwurf zur Publikation 2
- 99 -
LMW and the UD fraction increased with increasing salinity in both study areas. Iron
is also able to form stable organic complexes (Powell et al., 1996). Since the
solubility of inorganic iron in seawater is extremely low (e.g., Balzer, 1982), there
must be a significant part of dissolved Fe, which is associated with organic
compounds (Liu and Millero, 1999). Humic material in particular degrades only slowly,
and can protect Fe against coagulation by expanding the effective distance between
approaching colloids as well as by altering their net surface electrostatic charge
(Batchelli et al., 2010). For instance, humic ligands were identified to be a stabilizer
for the transport of riverine Fe to the ocean (Laglera and van den Berg, 2009). The
Fe-organic complexation may also be inferred by the correlation between DOC and
Fe concentrations in the HMW fraction (Fig. 6). The higher DOC content and
probably higher concentration of Fe-humic-substance in the WWR enhanced the
solubility of Fe and caused the difference in the slope between the WR and the WWR
in Fig. 6.
WR
0%
25%
50%
75%
100%
0 0 5 8 11 17 23 34Salinity
Fe ra
tio
WWR
0%
25%
50%
75%
100%
0 0 3 6 8 10 11 13 16 18 20 28Salinity
HMWLMWUD
0%
25%
50%
75%
100%
0 0 5 8 11 17 23 34Salinity
Pb r
atio
0%
25%
50%
75%
100%
0 0 3 6 8 10 11 13 16 18 20 28Salinity
Fig. 5. Trace metal ratios of various size fractions vs. salinity (left side: WR; right side: WWR).
Kapitel 5: Entwurf zur Publikation 2
- 100 -
0%
25%
50%
75%
100%
0 0 5 8 11 17 23 34Salinity
Cd
ratio
0%
25%
50%
75%
100%
0 0 3 6 8 10 11 13 16 18 20 28Salinity
0%
25%
50%
75%
100%
0 0 5 8 11 17 23 34Salinity
Cu
ratio
0%
25%
50%
75%
100%
0 0 3 6 8 10 11 13 16 18 20 28Salinity
0%
25%
50%
75%
100%
0 0 5 8 11 17 23 34Salinity
Co
ratio
0%
25%
50%
75%
100%
0 0 3 6 8 10 11 13 16 18 20 28Salinity
0%
25%
50%
75%
100%
0 0 5 8 11 17 23 34Salinity
Ni r
atio
0%
25%
50%
75%
100%
0 0 3 6 8 10 11 13 16 18 20 28Salinity
Fig. 5. (continued).
Kapitel 5: Entwurf zur Publikation 2
- 101 -
0
2
4
6
8
0 1 2 3DOC (ppm)
Fe (μ
mol
/L)
WR HMWWWR HMW
Fig. 6. Fe concentration vs. DOC in the HMW fraction.
5.3.2. Lead
The colloidal size distribution of dissolved Pb was comparable with dissolved Fe in
both study areas with Pb in the HMW fraction as the most important fraction (92%
and 95% on average in both estuaries, respectively). The ratios of the other size
fractions were less than 10% overall in both study areas and increased with
increasing salinity (Fig. 5). Dissolved Pb was widely found to associate with macro-
size Fe-rich colloids e.g., in the estuaries in Texas, USA (Benoit et al., 1994), in the
Penzé estuary (Waeles et al., 2008) and in the Mississippi Delta (Stolpe et al., 2010).
This is confirmed by the good correlations between HMW-Fe and HMW-Pb (Fig. 7).
Although labor experiment pointed out that Pb was taken up by both organic and Fe
colloids (Lyvén et al., 2003), the co-precipitation of Pb with Fe colloids appears to be
a typical characteristic in the estuarine chemistry of East Hainan (Fu et al., 2013).
0.0
0.5
1.0
1.5
0 2 4 6 8Fe (μmol/L)
Pb (n
mol
/L)
WR HMWWWR HMW
Fig. 7. Pb vs. Fe concentrations in the HMW fraction.
5.3.3. Cadmium
Fig. 5 shows the increase of both the concentration and the ratio of Cd in the UD
fraction with increasing salinity probably as a consequence of chloride induced
Kapitel 5: Entwurf zur Publikation 2
- 102 -
mobilization of particulate Cd (e.g., Comans and van Dijk, 1988). The Cd ratio in the
HMW and the LMW fraction decreases with increasing salinity, whereas their ratio
appears to be steady, perhaps with a small reduction of Cd in the HMW fraction at
low salinity due to adsorption of dissolved Cd on Mn oxide (Turner et al., 2008). It is
not surprising that Cd in the UD fraction was the greatest fraction with around 75% in
both estuaries due to formation of highly stable Cd-chloro-complexes (e.g., Van der
Weijden, 1977; Salomons, 1980; Comans and van Dijk, 1988). Such high
contribution of the Cd in the UD fraction was generally observed e.g., in the Rhône
delta (more than 90%; Dai et al., 1995), in the Venice Lagoon (65%; Martin et al.,
1995), in the Danube River estuary (60-70%; Guieu et al., 1998) and in the Seine
estuary (90-95%; Chiffoleau et al., 2001). Since the Cd ratio in the UD fraction in the
WR (81%) was higher than in the WWR (72%), the difference of the Cd ratio in the
UD fraction indicates a more complicated background than simple formation of highly
stable Cd-chloro-complexes. Organic complexes can also exert a significant control
of the Cd speciation in certain systems (Waeles et al., 2008). For instance, Cd in the
Penzé estuary was highlighted in winter and autumn to be constituted over 50% as
organic complexes due to its release from particle or sediment subsequent to
degradation of high molecular organic Cd (Waeles et al., 2005). Such degradation
could be inhibited by the relatively higher DOC content in the WWR (Fu et al., 2013)
and caused the difference in Cd speciation between the two study areas. On the
other hand, it also could be an evidence for the formation of organic Cd species
which aggregated in the HMW fraction. In conclusion, higher Cd ratio in the UD
fraction probably presents the dominance of Cd-chloro complexes. In systems with
relatively high colloidal Cd, both Cd-chloro and Cd-organic species may influence the
Cd behavior.
5.3.4. Copper
The Cu concentration in the UD fraction, with an average percentage of 40% and
52%, was the most significant size fraction, followed by the LMW fraction with 38%
and 25% and the HMW fraction with about 14% and 18% in the WR and the WWR,
respectively. Since dissolved Cu behaves generally conservatively (Fu et al., 2013), a
colloidal analysis suggested a conversion from the HMW fraction to the UD fraction
at low salinities with stronger tendency in the WR than in the WWR (Fig. 5). Copper
in the LMW fraction had a constant percentage along the whole salinity gradient. It is
believed that the competition between high and low molecular weight ligands may
control the Cu size distribution (Wen et al., 1999). For example, dissolved Cu was
Kapitel 5: Entwurf zur Publikation 2
- 103 -
mainly found in colloidal forms (5 kDa - 0.45 μm) which presented 94% of TD-Cu in
the Penzè estuary (Waeles et al., 2008). But the dominance of UD-Cu (< 10 kDa)
with more than 50% was observed overall in the Mississippi estuary (Wen et al.,
2011). These differences should be related to the nature of organic matter within the
systems of concern (Waeles et al., 2008). In this case, over 80% of dissolved Cu was
identified as low molecular weight ligands (< 10 kDa) in the saline water. In the
freshwater, however, most Cu occurred in the colloidal fraction (HMW and LMW),
especially in the WR with almost 100% of colloidal Cu. In association with the organic
nature of Cu in estuaries (e.g., Shank et al., 2004), the degradation of macro
molecular organic matter might cause such a drift of Cu from the HMW to the UD
fraction during the estuarine mixing (Masson et al., 2011). Such degradation was less
distinctive in the WWR probably because of the higher DOC content as discussed
before.
5.3.5. Cobalt
The investigation in the Galveston Bay demonstrated a permanently strong UD
fraction of dissolved Co (about 80%) with the argument of metabolical interaction with
Zn (Wen et al., 1999), while over 50% of dissolved Co was reported to be controlled
by Fe colloids (HMW) in Northwest Russia (Pokrovsky and Schott, 2002). In this
study, the HMW fraction as the largest size fraction of dissolved Co decreased
seawards from 75-80% to typically 35% in the seawater in both study areas (Fig. 11).
Conversely, the content of Co in the UD fraction increased from 10% to 30-40% with
increasing salinity. The Co in the LMW fraction with typically 20% varied
insignificantly with salinity. Previous study suggested a co-precipitation with Fe
colloids (Sharp et al., 1982; Fu et al., 2013), which could be confirmed by the good
correlation between Fe and Co in the HMW fraction (Fig. 8). Furthermore, previous
study suggested that lateral inputs from aquaculture ponds could influence the
behavior of dissolved Co and the elevated DOC content in the WWR (Fu et al., 2013).
Cobalt is also well known to be able to form strong organic complexes in nature
waters (e.g., Zhang et al., 1990; Qian et al., 1998; Saito and Moffett, 2001). The
correlation between organic matter and Co in the UD fraction (Fig. 8) might indicate
such an association of Co in the UD fraction with low molecular organic matter.
Kapitel 5: Entwurf zur Publikation 2
- 104 -
0
1
2
3
0 2 4 6 8Fe (μmol/L)
Co
(nm
ol/L
)
WR HMWWWR HMW
0
1
2
3
0 1 2 3 4DOC (ppm)
Co
(nm
ol/L
)
WR UDWWR UD
Fig. 8. Co vs. Fe concentrations in the HMW fraction and Co vs. DOC in the UD fraction.
5.3.6. Nickel
The UD fraction of Ni was the dominant size fraction in the both estuaries, (Fig. 5)
which was also observed e.g., in the Venice Lagoon (Martin et al., 1995), in the San
Francisco Bay (Sañudo-Wilhelmy et al., 1996) and in the Ochlockonee estuary
(Powell et al., 1996). At the very beginning of the estuarine mixing, the LMW-Ni and
the HMW-Ni behaved conservatively, which indicates stable Ni complexes in both
size fractions, while UD-Ni shows fluctuations compared to the theoretical dilution line.
Biological regeneration processes may affect the colloidal distribution of Ni and may
cause such an excess (Dai and Martin, 1995). But no relationship between Ni and
DOC was observed in this study. It is more likely that riverine particle-bound Ni was
replaced by Ca2+ and Mg2+ during the estuarine mixing (e.g., Powell et al., 1996;
Mandal et al., 2002). However, Ni is largely unreactive in estuaries. Both small
suspended particular matter fluxes in the estuary and the relatively low particle
affinity of Ni which is due to its strong association with dissolved organic matter can
inhibit the interactions between sediment, suspended particular matter, the colloid
and the UD fractions (Turner et al., 1998). All size fractions behaved therefore
conservatively at higher salinities in both study areas. However, the more significant
HMW fraction of dissolved Ni in the WWR than in the WR probably indicated stronger
involvement of organic matter in the colloidal distribution of dissolved Ni in the WWR
due to its higher DOC content.
5.3.7. Spatial variability of colloidal trace metals in East-Hainan
As mentioned before, for all studied dissolved trace metals the ratio of their HMW
fraction decreased with increasing salinity, whereas the ratio of their UD fraction
became more important, especially at low salinity (Fig. 5). Such variability was widely
observed e.g., in the Ochlockonee estuary (Powell et al., 1996), in the Narragansett
Kapitel 5: Entwurf zur Publikation 2
- 105 -
Bay (Wells et al., 1998) and during laboratory experiments (Stolpe and Hassellöv,
2007). Since a negative deviation from the theoretical dilution line was observed for
dissolved Fe in the WR and the WWR (Fu et al., 2013), the more importance of the
UD fraction with increasing salinity could be related to the loss of large size colloidal
Fe due to its flocculation during the estuarine mixing (e.g., Kraepiel et al., 1997).
Similarly, a removal of large size colloidal Pb and Co with freshly precipitated Fe-
oxyhydroxides (e.g., Sharp et al., 1982; Balls, 1990) was suspected to be responsible
for the increasing significance of Pb and Co in the UD fraction. Conversely, the
dissolved Cd showed a positive deviation from the theoretical dilution line in both
estuaries (Fu et al., 2013). The mobilization of Cd due to formation of highly stable
Cd-chloro-complexes (e.g., Comans and van Dijk, 1988) enhanced the Cd content in
the UD fraction. Despite the conservative behavior of dissolved Cu and Ni in both
estuaries (Fu et al., 2013), a conversion from the HMW to the UD fraction appeared
to occur with stronger tendency for Cu than for Ni. In accordance with the affinity of
Cu to organic material (e.g., Wang et al., 2003; Shank et al., 2004) such conversion
could be attributed to the degradation of macro molecular organic matter (Masson et
al., 2011). During the estuarine mixing Ni in the HMW fraction was replaced by Ca2+
and Mg2+ although such modification was not pronounced (e.g., Powell et al., 1996;
Mandal et al., 2002).
5.4. Conclusions
In this study, a cross flow ultrafiltration technique was used to separate various
colloidal fractions of trace metals in estuarine water. The colloidal distribution and its
modification along the salinity gradient were investigated in the Wanquan River and
Wenchang/Wenjiao River estuaries of East Hainan. Iron and lead were mainly
associated with high molecular weight colloids in the size range 10 kDa - 0.45 μm (>
90%), whereas cadmium (76%) and nickel (65%) appeared in the truly dissolved
fractions (< 5 kDa). Copper and cobalt favor in the high molecular weight colloids at
low salinities and are preferentially tied to the truly dissolved fractions at high
salinities. Some processes affecting trace metals during their transport from river to
ocean could be identified. The removal of Fe and Pb is controlled by flocculation of
Fe-oxyhydroxides. Also cobalt might be scavenged during co-precipitation with Fe
colloids at low salinity. The formation of cadmium chloride complexes controlled the
estuarine distribution of dissolved cadmium. The competition between nickel and
calcium/magnesium for the binding site at the surface of macro size material
influenced colloidal distribution of dissolved nickel during the estuarine mixing. The
Kapitel 5: Entwurf zur Publikation 2
- 106 -
degradation of organic copper was assumed to be responsible for the modification of
dissolved copper distribution. For all of the investigated metals the truly dissolved
fraction becomes more important with increasing salinity whereas the ratio of high
molecular weight fraction decreased. For iron, cadmium, copper and nickel the high
molecular weight fraction was more present in the WWR than in the WR.
Acknowledgements
The authors are thankful to Dr. Dao-Ru Wang and Dr. Larissa Dsikowitzky for their
coordinating and managing efforts during the study. Appreciation is extended to the
colleagues from the Centre for Tropical Marine Ecology (ZMT) of Germany, the
Hainan Provincial Marine Development and Design Institute (HNMDDI) of China, the
East China Normal University (ECNU) and the Ocean University of China (OUC) for
their help in the field and during laboratory works. Our special thanks will be given to
Dr. Uwe Schüßler for his help with the HR-ICP-MS analyses. We gratefully
acknowledge the funding by the "Bundesministerium für Bildung und Forschung"
(BMBF, German Federal Ministry of Education and Research) at Berlin under the
contract number 03F0457 C and the financial support from the Ministry of Science
and Technology (MOST) through the contract No. 2007DFB20380 and under the
agreement between the State Oceanic Administration of China (SOA) and the BMBF.
Kapitel 6: Entwurf zur Publikation 3
- 107 -
6. Entwurf zur Publikation 3:
Cd, Co, Cu, Fe, Ni and Pb in solution, colloidal and particulate phases of Northeast Brazilian estuaries
Abstract
Dissolved and particulate cadmium, copper, iron, lead, cobalt and nickel were
analyzed in surface waters of the Mundaú-Manguaba lagoon system in Northeast
Brazil during the wet season (September/October 2007) and two dry seasons
(February 2008 and March 2009). During the wet season, dissolved Cadmium
showed precipitation in the Mundaú lagoon and conservative behavior in the
Manguaba lagoon, while enrichment in the solution phase seemingly occurred,
especially at high salinities, during both dry seasons. In general, the behavior of
dissolved copper and nickel could be identified as conservative. Dissolved iron
showed a negative deviation from the theoretical dilution line, whereas the opposite
was observed for dissolved cobalt. Dissolved Pb co-precipitated with Fe in the
Mundaú lagoon but was partially enriched in the Manguaba lagoon. Essentially,
particulate metals behaved nearly conservatively and showed no seasonal variability
except for particulate Fe and Pb, which are scavenged during estuarine mixing. The
distribution coefficients were calculated with an order Fe > Pb > Co > Ni > Cu > Cd.
The colloidal partition and distribution of trace metals was investigated during the wet
season of 2007 and the dry season of 2009. Most metals had a significant presence
in the colloidal fraction (5 kDa - 0.45 μm). The truly dissolved fraction (< 5 kDa) for all
the metals was more important during the dry season than during the wet season.
Keywords: Trace metals; Estuaries; Colloids; Northeast-Brazil
Kapitel 6: Entwurf zur Publikation 3
- 108 -
6.1. Introduction
A coastal lagoon is an inland water body which is separated from the ocean by a
barrier and connected to the ocean by one or more restricted inlets (Kjerfve, 1994).
The geomorphology often creates an estuarine system with low tidal influence.
Estuaries often suffer from strong anthropogenic impacts. Estuarine systems can
also be thought of as acting as filters of the river-transported chemical signals, which
often emerge from the mixing zone in a form that is considerably modified with
respect to that which entered the system (Chester, 1990). In view of the fact that
rivers are one of the most important pathways for the transport of anthropogenic
materials to the ocean, it is necessary to understand how the estuarine filter operates
and especially, which part is retained before reaching the ocean.
The estuarine filter is selective in the manner in which it acts on different elements,
and the effects of the filter can vary widely from one estuary to another (Chester,
1990). Especially trace metals are a particularly interesting aspect of estuarine
chemistry because their differing physical chemistries lead to a variety of
geochemical behaviors (Shiller and Boyle, 1991) which are strongly influenced by the
chemical form in which the elements occur (Sigleo and Helz, 1981). The behavior of
trace metals in estuaries differs from one to another. In a given region, estuaries may
have comparable environmental conditions. However, the trace metal distribution in
different estuaries may differ from each other (e.g., Benoit et al., 1994; Dai and
Martin, 1995; Fu et al., 2013). Even in a certain estuary, seasonal variability of trace
metal behavior is not infrequent (e.g., Zwolsman et al., 1997; Jiann et al., 2005;
Waeles et al., 2005). Many estuarine processes have been investigated in the last
decades: for instance, chloride-induced desorption of Cd is characteristic for most
estuaries (e.g., Elbaz-Poulichet et al., 1996; Waeles et al., 2004; Dabrin et al., 2009).
Copper behaves conservatively in many estuaries (e.g., Shiller and Boyle, 1991;
Guieu et al., 1998; Martino et al., 2002), but both positive (e.g., Santos-Echeandia et
al., 2008; Masson et al., 2011) and negative deviation from the theoretical dilution
line (e.g., Sholkovitz, 1978; Sharp et al., 1982) have also been noted. The rapid
flocculation of Fe and Pb are reported for most estuaries (e.g., Turner et al., 1992;
Wen et al., 1999; Wang and Liu, 2003), whereas Ni is relatively unreactive during the
estuarine mixing and behaves mostly conservatively (e.g., Shiller and Boyle, 1991;
Turner et al., 1998; Fernández et al., 2008). Manganese involved desorption of Co is
widely observed in estuaries (e.g., Chiffoleau et al., 1994; Wen et al., 1999; Takata et
al., 2010).
The trace metal chemistry has been studied in various Brazilian estuarine and lagoon
Kapitel 6: Entwurf zur Publikation 3
- 109 -
systems along the Atlantic coast, e.g., in the Amazon estuary (Boyle et al., 1982), in
the Guanabara Bay (e.g., De Luca Rebello et al., 1986), in the Sepetiba Bay (e.g., de
Lacerda et al., 1987), in the Patos lagoon (e.g., Windom et al., 1999) and in the
Paraíba do Sul estuary (e.g., Carvalho et al., 2002). But there are no publications
which investigate the behavior of trace metals in the Mundaú and Manguaba lagoons.
Thus, the first aim of this study is to determine the estuarine distribution of selected
metals along the salinity gradients in these two study areas.
Colloids are important ligands affecting the speciation, fate, transport,
biogeochemistry, bioavailability and toxicity of trace metals in aquatic systems
(Doucet et al., 2007). For many trace metals, the colloidal size fraction is
predominant in the estuarine waters. For instance, most of total dissolved Fe and Pb
was found to be associated with colloidal material, e.g., in the Venice Lagoon (Martin
et al., 1995), in the Penzé estuary (Waeles et al., 2008a) and in the Mississippi River
delta (Shim et al., 2012). Also Cd and Cu may be strongly associated with colloidal
materials (e.g., Martin et al., 1995; Waeles et al., 2005; Wen et al., 2011). It is
possible that colloidal trace metals also undergo estuarine modification and
consequently change their distribution and state. Hence, the second aim of this study
is to investigate the colloidal trace metal behavior in these two study areas. Because
of close links between dissolved trace metals and dissolved organic carbon (DOC),
especially in colloidal fractions, analyses of DOC (Balzer, unpublished) were
evaluated and used for interpretation of the trace metal results.
Colloidal material is defined to have at least one dimension between 1 nm and 1 μm
(e.g., Everett, 1988; Buffle and Leppard, 1995) with 1 nm ≈ 1000 Daltons (1 kDa; e.g.,
Wen et al. 1999). Colloids are mostly separated from larger fractions by using
ultrafiltration. However, the exact molecular weight or size cut-off of the ultra-filter is
not very sharp (Buesseler et al., 1996). For the sequential separation protocol of this
study, the size fraction between 10 kDa and 0.45 μm is operationally defined as high
molecular weight colloid fraction (HMW) and the size fraction between 5 kDa and 10
kDa as low molecular weight colloid fraction (LMW), while the permeate from the
ultrafiltration (< 5 kDa) is considered to be the truly dissolved fraction (UD) in the
present work.
Kapitel 6: Entwurf zur Publikation 3
- 110 -
6.2. Materials and Methods
6.2.1. Study area
Fig. 1. The Mundaú-Manguaba lagoon system in Northeast Brazil.
The Mundaú-Manguaba lagoon system is located in the state of Alagoas, Northeast
Brazil (Fig. 1). Its climate is tropical, semi-humid with well-defined dry (September-
March) and wet (April-August) seasons (Oliveira and Kjerfve, 1993; Brockmeyer and
Spitzy, 2011), with mean annual temperatures around 24°C (Diegues, 1994). The
Mundaú-Manguaba estuarine system is composed of two large lagoons (Melo-
Magalhães et al., 2009): the northern Mundaú lagoon (in the following: MD), with the
Rio Mundaú as its major tributary (catchment area 2135 km2), has a total surface
water area of 24 km2 and an average depth of 1.5 m; the southern Manguaba lagoon
(in the following: MG), which receives fresh water mainly from the Rio Paraíba do
Meio (catchment area 3299 km2) and the Rio Sumaúma (catchment area 372 km2),
has a total surface water area of 43 km2 and an average depth of 2 m (de Souza et
al., 2002; Costa et al., 2011). Both lagoons are interconnected by mainly mangrove-
lined narrow channels with a single inlet connecting them to the Atlantic. Such so-
called “choked lagoons” are characterized by one or more long and narrow entrance
channels, long residence times of water in the estuary and dominant wind forcing
Kapitel 6: Entwurf zur Publikation 3
- 111 -
(Kjerfve and Magill, 1989), and are particularly vulnerable to natural and
anthropogenic impacts (Knoppers et al., 1991). In both lagoons, the flux of salt water
from the coastal ocean into the lagoons is strongly dampened by the channel system,
where tidal cycles determine the direction and magnitude of currents (Brockmeyer
and Spitzy, 2011). During the wet season, the lagoons are strongly influenced by
freshwater and sediment transport in the rivers, while the river flow is low during the
dry season which means that the system is mainly controlled by tidal variation (Spörl,
2011). The average water residence time ranges from around 1-2 weeks for MD to 5-
7 weeks for MG (Oliveira and Kjerfve, 1993). In the river catchment areas, sugar
cane is the dominant land cover for both lagoons with a further impact by urban
effluents for the MD from the city of Maceío (population ca. 1 million; e.g.,
Brockmeyer and Spitzy, 2011).
The Mundaú-Manguaba lagoon system is strongly influenced by monocultural sugar
cane cultivation during the last decades, which led to raised nutrient inputs and may
result in spatial and seasonal differences of nutrient concentrations in the system
(Spörl, 2011). This may be the reason why dissolved organic carbon concentration in
the study area (4-11 mg/L) is considerably higher than in other estuaries in Brazil,
such as in the Amazon estuary (1-4 mg/L; Sholkovitz et al., 1978), in the Sepetiba
Bay (1.7-5.8 mg/L; de Lacerda et al., 2001) and in the Paraíba do Sul River estuary
(1.5-5.6 mg/L; Krüger et al., 2006).
6.2.2. Sampling and filtration
The sampling was carried out during September/October 2007 (nominal wet season)
as well as during February 2008 and March 2009 (nominal dry season) with the aid
of small boats. Surface waters from a depth of 0.3 m were filled by hand into pre-
cleaned low-density polyethylene (LDPE) bottles (Nalgene®, 1 L) upstream of the
ship. Within 24 hours the water samples were drawn through acid-cleaned 0.45 μm
polycarbonate filters (Nuclepore®; pre-weighed) in the laboratory. The filter cakes
were transferred to Petri dishes and stored at -10 °C. Subsequently, the filtrates were
separated into fractions of 30 kDa-0.45 μm, 10-30 kDa, 5-10 kDa and < 5 kDa using
cross flow filtration devices (CFF) with a flow rate of 500 mL/min (2 bar). The
fractions of 30 kDa-0.45 μm and 10-30 kDa have been put together, because these
two fractions contained no additional information. The filtrate, the retentate and
permeate were weighted before the next step began. After weighing, 10 mL of each
retentate/permeate was filled into an ampoule with addition of phosphoric acid for the
Kapitel 6: Entwurf zur Publikation 3
- 112 -
dissolved organic carbon measurement. The rest of the sample was filled into LDPE
bottles (Nalgene®) and acidified to pH 2 with sub-boiled concentrated nitric acid for
trace metal analysis. The sample bottles were wrapped in polyethylene bags and
stored at room temperature. The used cross-flow filtration devices (Vivaflow® 200,
Sartorius) constructed of polyethersulfone with an active membrane area of 200 cm2
were pre-washed with HCl (0.1%, 30 min) and EDTA (0.02 M, 30 min) and
subsequently rinsed with Milli-Q® water.
6.2.3. Pretreatment and measurement of dissolved and colloidal metals
Sample work-up was performed in a clean-air laboratory (class 1000) with additional
clean benches of class 100. The samples were first irradiated in a UV-digester (500W;
UV-digester 705, Metrohm) for 2 hours with addition of hydrogen peroxide in order to
destroy the organic complexes. Afterwards they were extracted by an
extraction/backextraction procedure based on the work of Danielsson et al. (1978): at
first, 40-100 g sample (depending on the metal concentration) was weighed into a
fluorinated ethylene propylene (FEP) separatory funnel (Nalgene®). After pH
adjustment to 4.0-4.5 with ammonium acetate buffer (pH ~ 9), the sample was well
shaken (15 min) after addition of 500 μL APDC/NaDETC solution (0.06 M) and 10 mL
Freon. The lower organic phase was drawn off into a polypropylene tube following
the phase separation. After addition of 100 μL concentrated nitric acid the tube was
shaken by hand for 3 min. Finally, 1900 μL Milli-Q® water was added and the tube
was again shaken by hand for 2 min for back-extraction. After the phase separation,
the upper aqueous extract was removed into an acid-cleaned sample tube using a
pipette. The extract was measured by high-resolution-inductively-coupled-plasma
mass-spectrometry (HR-ICP-MS; Element 2, Thermo Scientific).
6.2.4. Cross-flow filtration Blanks and Mass balance
To obtain CFF blanks by using a new pre-washed CFF device, a Milli-Q® water
sample was separated by ultrafiltration into 4 colloidal fractions (30 kDa - 0.45 μm;
10-30 kDa; 5-10 kDa and < 5 kDa). All acidified sub-samples were analyzed for trace
metals using HR-ICP-MS. All trace metal concentrations of the blanks were less than
1% of the minimum sample concentrations or not detectable.
Kapitel 6: Entwurf zur Publikation 3
- 113 -
Cd2007: y = 0.6777x
2009: y = 0.9382x0.00
0.04
0.08
0.12
0.00 0.04 0.08 0.12total dissolved (nmol/L)
HM
W+L
MW
+UD
(n
mol
/L)
Cu2007: y = 0.9862x
2009: y = 1.0632x0
10
20
30
0 10 20 30total dissolved (nmol/L)
HM
W+L
MW
+UD
(n
mol
/L)
Fe2007: y = 0.5416x
2009: y = 0.5724x0
2
4
6
0 2 4 6total dissolved (μmol/L)
HM
W+L
MW
+UD
(μ
mol
/L)
Pb2007: y = 0.6393x
2009: y = 1.0114x0.00
0.25
0.50
0.75
1.00
0.00 0.25 0.50 0.75 1.00total dissolved (nmol/L)
HM
W+L
MW
+UD
(n
mol
/L)
Co2007: y = 1.1526x
2009: y = 1.0386x0
1
2
3
4
0 1 2 3 4total dissolved (nmol/L)
HM
W+L
MW
+UD
(n
mol
/L)
Ni2007: y = 1.1725x
2009: y = 1.0003x0
5
10
15
20
0 5 10 15 20total dissolved (nmol/L)
HM
W+L
MW
+UD
(n
mol
/L)
Fig. 2. Calculated total dissolved concentration (HMW + LMW + UD) vs. separately measured total dissolved trace metal concentration : MD 2007; : MG 2007; : MD 2009; : MG 2009. Solid lines: Correlation between calculated and separately measured total dissolved concentration for Cd, Cu, Fe, Pb, Co and Ni during the wet season of 2007; Broken line: Correlation between calculated and separately measured total dissolved concentration for Cd, Cu, Fe, Pb, Co and Ni during the dry season of 2009).
Since there is no standard reference material to quantify the accuracy of the
ultrafiltration analysis, a concentration mass balance between the sum of all size
fractions (HMW + LMW + UD) and the 0.45 μm filtrated sample (total dissolved
fraction) was calculated. A mass balance < 100% indicates a loss of material
probably due to scavenging onto the filter membrane or sample reservoir, while a
mass balance > 100% may be a evidence for a contamination during the sample
pretreatment (e.g., Martin et al., 1995; Powell et al., 1996). Also the propagation of
sampling and analytical errors for the colloidal sub-samples and for the UD sample in
particular may cause a deviation of the mass balance from 100%. Especially for
Kapitel 6: Entwurf zur Publikation 3
- 114 -
lower concentrations, a small absolute error may create a remarkable aberrance in
the mass balance. Most calculations showed a mass balance around 100%. Only
those data sets were evaluated, which had a mass balance within a range of 90% ±
40%, 110% ± 30%, 80% ± 40%, 70% ± 40%, 100% ± 30% and 110% ± 40% during
the wet season of 2007, and 100% ± 40%, 100% ± 30%, 100% ± 50%, 100% ± 50%,
100% ± 20% and 100% ± 30% during the dry season of 2009 for Cd, Cu, Fe, Pb, Co
and Ni, respectively. The calculated coefficients of determination generally showed a
good correlation between the separately measured total dissolved trace metal
concentrations and the sum of the colloidal and the UD fractions (Fig. 2). The slopes
were mainly within a range of 0.9-1.1. However, the slopes for Cd, Fe and Pb during
the wet season of 2007 and the slope for Fe during the dry season of 2009 were
lower probably indicating a loss of material, but have not influenced the operative
mass balance. During the wet season of 2007, elevated slopes were found for Co
and Ni possibly evidencing a contamination. Errors during the sampling or the
calculation could also cause the fluctuation of the mass balance. From a total of 31
analyzed samples, 20 for Cd, 24 for Cu, 14 for Fe, 15 for Pb, 21 for Co and 20 for Ni
were selected for further evaluation, which had a mass balance in the respective
defined range.
6.2.5. Pretreatment and measurement of particulate metals
The pretreatment of particulate samples was also performed in a clean-air laboratory
(class 1000) with additional clean benches of class 100. The filter cakes were
weighed at least threefold with one day storage over conc. H2SO4 in between, before
the filter cakes have been dried over conc. H2SO4 for several days. After weighting,
the filter cakes were decomposed with the filter in sealed PTFE vessels using a
microwave-assisted pressure digestion technique (MWS-1200 mega microwave
digestion system with HPR 1000-6 rotor; MLS, Leutkirch, Germany) as Balzer et al.
(2013) did: The samples were destructed under a radiation power of up to 600 W
after adding a mixture of concentrated HNO3, HCl, HClO4 and HF (sub-boiled or
suprapure®), and subsequently under the same power after only adding HNO3 (for
further details, see Schüßler et al., 2005). The metal concentrations were also
measured by HR-ICP-MS. The certified reference material MESS-1 (marine sediment)
from the NRCC was analyzed in order to ensure the validity of the method. The
results were found to be in the certified ranges.
Kapitel 6: Entwurf zur Publikation 3
- 115 -
6.2.6. Other parameters
Environmental parameters such as salinity and water temperature were measured on
the spot using a multi-sensor-device (MultiLine P3, WTW). Dissolved organic carbon
was determined by high temperature catalytic oxidation (Total Carbon Analyzer,
Apollo 9000; Tecmar).
6.3. Results and Discussion
6.3.1. Salinity, suspended particulate matter and dissolved organic carbon
wet season 2007
0
20
40
60
80
0 10 20 30
SPM
(mg/
L)
dry season 2008
0
20
40
60
80
0 10 20 30
SPM
(mg/
L)
dry season 2009
0
20
40
60
80
0 10 20 30
SPM
(mg/
L)
(b)
wet season 2007
0
20
40
60
80
0 10 20 30
SPM
(mg/
L)
dry season 2008
0
20
40
60
80
0 10 20 30
SPM
(mg/
L)
dry season 2009
0
20
40
60
80
0 10 20 30
SPM
(mg/
L)
Fig. 3. The estuarine distribution of suspended particulate matter (SPM) vs. salinity (a) in the Mundaú Lagoon and (b) in the Manguaba Lagoon : MD; : MG; : Sumaúma; : coastal water; left side of dotted line: lagoon water; right side of dotted line: channel and coastal water).
The concentration of suspended particulate matter (SPM) decreased with increasing
salinity from typically 20 mg/L in the MD and 30 mg/L in the MG to less than 5 mg/L
during both the wet season and the dry seasons (Fig. 3). The extremely high SPM
concentration (76 mg/L) at salinity of 3 psu in the MD during the wet season of 2007
(Fig. 3a) probably originated from higher resuspension loadings. The SPM level in
the MG was generally higher than in the MD. The SPM concentration decreased with
increasing salinity from typically 30-40 mg/L to less than 5 mg/L during the wet
season of 2007 and the dry season of 2009 in the MG despite some scatter (Fig. 3b).
In the MG during the dry season of 2008, the SPM concentration in the riverine water
(S = 0) was relatively low (10 mg/L), and a mid turbidity maximum (20 mg/L) was
observed at a salinity between 20 and 25 psu. The freshwater end-member SPM in
the Rio Sumaúma was approximately at the same level as in the Rio Paraíba do
Meio, and could not influence the SPM distribution in the MG significantly, because of
Kapitel 6: Entwurf zur Publikation 3
- 116 -
its low discharge (Oliveira and Kjerfve, 1993). In comparison with estuaries in Europe
and North America (e.g., Uncles et al., 2002), the SPM concentration in the MD and
MG estuaries was at the mid level, and lower than most estuaries in South America
such as the Paraíba do Sul estuary (5-193 mg/L; Carvalho et al., 2002), the Río de la
Plata estuary (31-70 mg/L; Calliari et al., 2005) and the Bahía Blanca Estuary (78
mg/L; Guinder et al., 2009).
(a) wet season 2007
0
4
8
12
0 10 20 30
DOC
(mg/
L)
dry season 2008
0
4
8
12
0 10 20 30
DOC
(mg/
L)
dry season 2009
0
4
8
12
0 10 20 30
DOC
(mg/
L)
(b)
wet season 2007
0
4
8
12
0 10 20 30
DOC
(mg/
L)
dry season 2008
0
4
8
12
0 10 20 30
DO
C (m
g/L)
dry season 2009
0
4
8
12
0 10 20 30
DOC
(mg/
L)
Fig. 4. The estuarine distribution of DOC vs. salinity (a) in the Mundaú Lagoon and (b) in the Manguaba Lagoon : MD; : MG; : Sumaúma; : coastal water).
The distribution of DOC as determined by Balzer (unpublished) is depicted in Fig. 4.
During the wet season of 2007 and the dry season of 2008, the DOC concentration
generally decreased with increasing salinity from about 4 mg/L in the riverine water to
1 mg/L in coastal waters of both study areas, except for the MG during the dry
season of 2008 and the dry season of 2009. During the latter season, the DOC
concentration decreased from 9 mg/L in the MD and 12 mg/L in the MG to 4 mg/L
with increasing salinity. Cyanobacteria blooms were frequently observed in both
study areas (Porfirio et al., 1999; Spörl, 2011) which may promote the increase of
DOC (e.g., Norrman et al., 1995). The mid salinity maximum of DOC in the MG
during the dry season of 2008 and the DOC peak at a salinity of 15 psu in the MD
during the dry season of 2009 may therefore be a consequence of local
cyanobacteria blooms. The observed DOC data was comparable with earlier studies
in this study area (Wolf et al., 2010; Brockmeyer and Spitzy, 2011), but was higher
than in other Brazilian estuaries e.g., Sepetiba Bay (Luciani et al., 2007). It was
obviously at the higher limit when compared with the range of the European estuaries
(Abril et al., 2002).
Kapitel 6: Entwurf zur Publikation 3
- 117 -
6.3.2. Dissolved trace metals
(a) wet season 2007
0.00
0.04
0.08
0.12
0 10 20 30
Cd
(nm
ol/L
) dry season 2008
0.00
0.04
0.08
0.12
0 10 20 30
Cd
(nm
ol/L
) dry season 2009
0.00
0.04
0.08
0.12
0 10 20 30
Cd
(nm
ol/L
)
(b)
wet season 2007
0
1
2
3
4
0 10 20 30
Cd
(nm
ol/g
) dry season 2008
0
1
2
3
4
0 10 20 30
Cd
(nm
ol/g
) dry season 2009
0
1
2
3
4
0 10 20 30
Cd
(nm
ol/g
)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 3 5 21 36Salinity
Cd
ratio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 3 6 10 12 14 21 25 28Salinity
Cd
ratio
HMWLMWUD
Fig. 5. Estuarine distribution of Cd in the Mundaú Lagoon : MD; : coastal water): (a) total dissolved Cd, (b) particulate Cd and (c) Cd percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Kapitel 6: Entwurf zur Publikation 3
- 118 -
(a) wet season 2007
0.00
0.04
0.08
0.12
0 10 20 30
Cd
(nm
ol/L
) dry season 2008
0.00
0.04
0.08
0.12
0 10 20 30
Cd
(nm
ol/L
) dry season 2009
0.00
0.04
0.08
0.12
0 10 20 30
Cd
(nm
ol/L
)
(b)
wet season 2007
0
1
2
3
4
0 10 20 30
Cd
(nm
ol/g
) dry season 2008
0
1
2
3
4
0 10 20 30
Cd
(nm
ol/g
) dry season 2009
0
1
2
3
4
0 10 20 30
Cd
(nm
ol/g
)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 0 1 3 5 9 10 19Salinity
Cd
ratio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 0 2 4 8 17 34Salinity
Cd
ratio
HMWLMWUD
Fig. 6. Estuarine distribution of Cd in the Manguaba Lagoon : MG; : Sumaúma; : coastal water): (a) total dissolved Cd, (b) particulate Cd and (c) Cd percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Cadmium: Numerous estuarine investigations and laboratory experiments suggest
that an enrichment of dissolved Cd due to chloride-induced desorption from particles
is a typical feature in estuaries (e.g., Chiffoleau et al., 2001; Waeles et al., 2004;
Dabrin et al., 2009). But this effect was not clear in this study. In the MD during the
wet season of 2007, dissolved Cd decreased rapidly from about 0.08 nmol/L in the
freshwater to 0.03 nmol/L at high salinities (Fig. 5a). Such elimination (at low salinity)
was also observed e.g., in the Delaware estuary (Sharp et al., 1982), in the Göta
River estuary (Danielsson et al., 1983) and in the Danshuei River estuary (Jiann et
al., 2005) and is often attributed to adsorption onto Fe/Mn oxides (e.g., Turner et al.,
2008). In the MG, however, the freshwater dissolved Cd concentration was quite low
(< 0.02 nmol/L) and varied insignificantly with salinity (Fig. 6a). During both dry
seasons (2008 and 2009), dissolved Cd increased from typically 0.02 nmol/L in the
riverine water to about 0.1 nmol/L at high salinities. Such a more or less conservative
behavior was also observed e.g., in the Rhône estuary (Elbaz-Poulichet et al., 1996)
and in estuaries of East Hainan (Fu et al., 2013). However, during the dry season of
Kapitel 6: Entwurf zur Publikation 3
- 119 -
2009, the dissolved Cd concentration in the coastal water outside the lagoon system
was quite low (0.03 nmol/L) probably representing waters from remote oceanic sites.
Dissolved Cd seemingly shows a positive deviation from the theoretical dilution line.
A previous study reported a dissolved Cd concentration of 0.02 nmol/L in Brazilian
coastal water (Kremling, 1985). The calculation of mass balance showed also a
positive Cd net flux into the ocean (particulate + dissolved concentration in nmol/L;
not shown). Thus, in addition to possible anthropogenic contributions, desorption
from the sediment and/or suspended particles inside the lagoon system is assumed.
While at lower salinities (0-20 psu), such desorption was not pronounced, the
increase was significant at higher salinities as also observed in the Huanghe estuary
(Elbaz-Poulichet et al., 1987). The elevated Cd concentrations occurred mostly in the
channel (Fig. 3). This may be due to higher flow rate in the channel section and
subsequently stronger water-sediment exchange. Dissolved oxygen was neither
measured nor published for the dry seasons. But during the wet season of 2008,
dissolved oxygen in the lagoons was quite low in the surface water (about 5 mg/L;
Spörl, 2011), and presumably much lower at the water-sediment-interface during the
dry season. Such suboxic conditions may have inhibited the release of Cd in the
lagoons (e.g., Jiann et al., 2005).
Kapitel 6: Entwurf zur Publikation 3
- 120 -
(a) wet season 2007
0
5
10
15
20
0 10 20 30
Cu
(nm
ol/L
) dry season 2008
0
5
10
15
20
0 10 20 30
Cu
(nm
ol/L
) dry season 2009
0
5
10
15
20
0 10 20 30
Cu
(nm
ol/L
)
(b)
wet season 2007
0.00
0.15
0.30
0.45
0 10 20 30
Cu (μ
mol
/g) dry season 2008
0.00
0.15
0.30
0.45
0 10 20 30
Cu (μ
mol
/g) dry season 2009
0.00
0.15
0.30
0.45
0 10 20 30
Cu (μ
mol
/g)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 3 5 21 36Salinity
Cu
ratio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 3 6 10 12 14 21 25 28Salinity
Cu
ratio
HMWLMWUD
Fig. 7. Estuarine distribution of Cu in the Mundaú Lagoon ( : MD; : coastal water): (a) total dissolved Cu, (b) particulate Cu and (c) Cu percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Kapitel 6: Entwurf zur Publikation 3
- 121 -
(a) wet season 2007
0
5
10
15
20
0 10 20 30
Cu
(nm
ol/L
) dry season 2008
0
5
10
15
20
0 10 20 30
Cu
(nm
ol/L
) dry season 2009
0
5
10
15
20
0 10 20 30
Cu
(nm
ol/L
)
(b)
wet season 2007
0.00
0.15
0.30
0.45
0 10 20 30
Cu
(μm
ol/g
) dry season 2008
0.00
0.15
0.30
0.45
0 10 20 30
Cu
(μm
ol/g
) dry season 2009
0.00
0.15
0.30
0.45
0 10 20 30
Cu
(μm
ol/g
)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 0 1 3 5 9 10 19Salinity
Cu
ratio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 0 2 4 8 17 34Salinity
Cu
ratio
HMWLMWUD
Fig. 8. Estuarine distribution of Cu in the Manguaba Lagoon : MG; : Sumaúma; : coastal water): (a) total dissolved Cu, (b) particulate Cu and (c) Cu percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Copper: A supply of Cu to the solution phase during the estuarine mixing may be a
result of external sources such as desorption from particulate Cu or anthropogenic
activities (e.g., Zwolsman et al., 1997; Guo, et al., 2000b; Santos-Echeandia et al.,
2008). On the other side, a removal of Cu from the dissolved phase may occur due to
the formation of copper sulfide (Jiann and Wen, 2009). In this study, dissolved Cu
concentration decreased from typically 10 nmol/L to 1-2 nmol/L with increasing
salinity during all seasons in both study areas (Fig. 7a and Fig. 8a). Dissolved Cu
appeared to behave conservatively during both wet and dry seasons, which was
widely observed e.g., in the Mississippi River estuary (Wen et al., 2011) and in the
Nerbioi-Ibaizabal River estuary (Fernández et al., 2008). The fluctuations at low
salinities, especially in the MG, may be attributed to irregular anthropogenic inputs
into the riverine water.
Kapitel 6: Entwurf zur Publikation 3
- 122 -
(a) wet season 2007
0
2
4
6
0 10 20 30
Fe (μ
mol
/L)
dry season 2008
0
2
4
6
0 10 20 30
Fe (μ
mol
/L)
dry season 2009
0
2
4
6
0 10 20 30
Fe (μ
mol
/L)
(b)
wet season 2007
0.0
0.5
1.0
1.5
2.0
0 10 20 30
Fe (m
mol
/g) dry season 2008
0.0
0.5
1.0
1.5
2.0
0 10 20 30
Fe (m
mol
/g) dry season 2009
0.0
0.5
1.0
1.5
2.0
0 10 20 30
Fe (m
mol
/g)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 3 5 21 36Salinity
Fe ra
tio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 3 6 10 12 14 21 25 28Salinity
Fe ra
tio
HMWLMWUD
Fig. 9. Estuarine distribution of Fe in the Mundaú Lagoon ( : MD; : coastal water): (a) total dissolved Fe, (b) particulate Fe and (c) Fe percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Kapitel 6: Entwurf zur Publikation 3
- 123 -
(a) wet season 2007
0
2
4
6
0 10 20 30
Fe (μ
mol
/L)
dry season 2008
0
2
4
6
0 10 20 30
Fe (μ
mol
/L)
dry season 2009
0
2
4
6
0 10 20 30
Fe (μ
mol
/L)
(b)
wet season 2007
0.0
0.5
1.0
1.5
2.0
0 10 20 30
Fe (m
mol
/g) dry season 2008
0.0
0.5
1.0
1.5
2.0
0 10 20 30
Fe (m
mol
/g) dry season 2009
0.0
0.5
1.0
1.5
2.0
0 10 20 30
Fe (m
mol
/g)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 0 1 3 5 9 10 19Salinity
Fe ra
tio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 0 2 4 8 17 34Salinity
Fe ra
tio
HMWLMWUD
Fig. 10. Estuarine distribution of Fe in the Manguaba Lagoon : MG; : Sumaúma;
: coastal water): (a) total dissolved Fe, (b) particulate Fe and (c) Fe percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Iron: In the MD (Fig. 9a), dissolved Fe decreased rapidly from 2-6 μmol/L during the
wet season and 4 μmol/L during both dry seasons to typically 0.2 μmol/L at low
salinities (S < 10) and varied insignificantly at higher salinities (S = 10-33). In the MG,
dissolved Fe also decreased with increasing salinity from typically 1-6 μmol/L to 0.2
μmol/L at S > 31 (Fig. 10a). The considerable fluctuations in the freshwater and some
elevated outliers at low salinities are probably due to irregular input from sewage and
agricultural activities (especially in the MD during the dry season of 2009 and in the
Rio Sumaúma during the wet season of 2007). During various field works and
laboratory experiments, the estuarine removal of dissolved Fe (e.g., Aston and
Chester, 1973; Sholkovitz et al., 1978) was estimated to result in a loss of more than
90% (e.g., Boyle et al., 1977; Sholkovitz, 1978) as a consequence of flocculation of
colloidal Fe and subsequent sedimentation (e.g., Mayer, 1982; Fox and Wofsy, 1983).
In both lagoons, dissolved Fe behaved non-conservatively and also underwent
estuarine flocculation. The riverine dissolved Fe in the MD was mostly higher than in
Kapitel 6: Entwurf zur Publikation 3
- 124 -
the MG suggesting geochemical differences of the catchment areas between the Rio
Mundaú and the Rio Paraíba do Meio.
(a) wet season 2007
0.0
0.5
1.0
1.5
2.0
0 10 20 30
Pb (n
mol
/L) dry season 2008
0.0
0.5
1.0
1.5
2.0
0 10 20 30Pb
(nm
ol/L
) dry season 2009
0.0
0.5
1.0
1.5
2.0
0 10 20 30
Pb (n
mol
/L)
(b)
wet season 2007
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Pb (μ
mol
/g) dry season 2008
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Pb (μ
mol
/g) dry season 2009
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Pb (μ
mol
/g)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 3 5 21 36Salinity
Pb ra
tio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 3 6 10 12 14 21 25 28Salinity
Pb
ratio
HMWLMWUD
Fig. 11. Estuarine distribution of Pb in the Mundaú Lagoon ( : MD; : coastal water): (a) total dissolved Pb, (b) particulate Pb and (c) Pb percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Kapitel 6: Entwurf zur Publikation 3
- 125 -
(a) wet season 2007
0.0
0.5
1.0
1.5
2.0
0 10 20 30
Pb (n
mol
/L) dry season 2008
0.0
0.5
1.0
1.5
2.0
0 10 20 30
Pb (n
mol
/L) dry season 2009
0.0
0.5
1.0
1.5
2.0
0 10 20 30
Pb (n
mol
/L)
(b)
wet season 2007
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Pb (μ
mol
/g) dry season 2008
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Pb (μ
mol
/g) dry season 2009
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Pb (μ
mol
/g)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 0 1 3 5 9 10 19Salinity
Pb ra
tio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 0 2 4 8 17 34Salinity
Pb
ratio
HMWLMWUD
Fig. 12. Estuarine distribution of Pb in the Manguaba Lagoon : MG; : Sumaúma;
: coastal water): (a) total dissolved Pb, (b) particulate Pb and (c) Pb percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Lead: Dissolved Pb decreased with increasing salinity from typically 0.4 nmol/L to
0.1-0.2 nmol/L in the MD during both the wet season and two dry seasons (Fig. 11a).
The estuarine removal of dissolved Pb was established by various estuarine
investigations (e.g., Dai and Martin, 1995; Wen et al., 1999; Monbet, 2006), which
was attributed to co-precipitation with Fe and Mn oxides (e.g., Wen et al., 2008). In
the MG, however, such a decrease was not pronounced (Fig. 12a). During the dry
season of 2008, dissolved Pb was constant at about 0.2-0.3 nmol/L, such as the
conservative behavior e.g., in the Rhône Estuary (Elbaz-Poulichet et al., 1996) and in
the Gironde Estuary (Kraepiel et al., 1997). During the wet season of 2007 and the
dry season of 2009, dissolved Pb concentration in the river and the sea end-
members differed from the dry season of 2008 not significantly, but the estuarine
distribution showed a positive deviation from the theoretical dilution line. A similar
distribution was also noted e.g., in the Mersey estuary due to anthropogenic activities
(Comber et al., 1995) and in the Penzé estuary due to resuspension of historically
Kapitel 6: Entwurf zur Publikation 3
- 126 -
contaminated sediment (Waeles et al., 2007). Since the SPM load was higher during
both surveys (Fig. 3), a mobilization from labile particulates might have taken place.
The elevation of dissolved Pb in the MD at s ≈ 10 during the dry season of 2009 and
in the Rio Sumaúma during the wet season of 2007 is quite similar to dissolved Fe
and Ni. Thus, a contamination during the sample pretreatment might be excluded. It
is more likely, as discussed before, that irregular municipal input was responsible for
the excesses of dissolved Fe and Pb.
(a) wet season 2007
0
1
2
3
4
0 10 20 30
Co (n
mol
/L) dry season 2008
0
1
2
3
4
0 10 20 30
Co (n
mol
/L) dry season 2009
0
1
2
3
4
0 10 20 30
Co (n
mol
/L)
(b)
wet season 2007
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Co
(μm
ol/g
) dry season 2008
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Co
(μm
ol/g
) dry season 2009
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Co
(μm
ol/g
)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 3 5 21 36Salinity
Co
ratio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 3 6 10 12 14 21 25 28Salinity
Co
ratio
HMWLMWUD
Fig. 13. Estuarine distribution of Co in the Mundaú Lagoon ( : MD; : coastal water): (a) total dissolved Co, (b) particulate Co and (c) Co percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Kapitel 6: Entwurf zur Publikation 3
- 127 -
(a) wet season 2007
0
1
2
3
4
0 10 20 30
Co
(nm
ol/L
) dry season 2008
0
1
2
3
4
0 10 20 30
Co
(nm
ol/L
) dry season 2009
0
1
2
3
4
0 10 20 30
Co
(nm
ol/L
)
(b)
wet season 2007
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Co
(μm
ol/g
) dry season 2008
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Co
(μm
ol/g
) dry season 2009
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Co
(μm
ol/g
)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 0 1 3 5 9 10 19Salinity
Co
ratio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 0 2 4 8 17 34Salinity
Co
ratio
HMWLMWUD
Fig. 14. Estuarine distribution of Co in the Manguaba Lagoon : MG; : Sumaúma;
: coastal water): (a) total dissolved Co, (b) particulate Co and (c) Co percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Cobalt: Dissolved Co in the MD and MG was plotted as a function of salinity in Fig.
13a and Fig. 14a. During the wet season in both study areas, dissolved Co
decreased from about 0.6 nmol/L in the riverine water to 0.4 nmol/L in coastal waters.
During both dry seasons, dissolved Co also decreased from typically 2 nmol/L to less
than 0.4 nmol/L. The estuarine distribution behaved non-conservatively, and was
characterized by a positive deviation from the theoretical dilution line with a maximum
of about 2 nmol/L at low salinity during the wet season and maxima of about 3 nmol/L
during both dry seasons. The enrichment of dissolved Co appeared to take place in
particular at low salinities. Although removal of dissolved Co was observed e.g., in
the Changjiang estuary (Wang and Liu, 2003) and in East Hainan (Fu et al., 2013),
most field works and labor experiments indicated the inverse situation (e.g., Li et al.,
1984; Martino et al., 2002; Tovar-Sánchez et al., 2004; Takata et al., 2010). Release
of Co from the resuspended sediment may have contributed to the enrichment of
dissolved Co (e.g., Chiffoleau et al., 1994; Chaudry and Zwolsman, 2008).
Kapitel 6: Entwurf zur Publikation 3
- 128 -
(a) wet season 2007
0
5
10
15
0 10 20 30
Ni (
nmol
/L)
dry season 2008
0
5
10
15
0 10 20 30
Ni (
nmol
/L)
dry season 2009
0
5
10
15
0 10 20 30
Ni (
nmol
/L)
(b)
wet season 2007
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Ni (μm
ol/g
)
dry season 2008
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Ni (μm
ol/g
)
dry season 2009
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Ni (μm
ol/g
)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 3 5 21 36Salinity
Ni r
atio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 3 6 10 12 14 21 25 28Salinity
Ni r
atio
HMWLMWUD
Fig. 15. Estuarine distribution of Ni in the Mundaú Lagoon ( : MD; : coastal water): (a) total dissolved Ni, (b) particulate Ni and (c) Ni percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Kapitel 6: Entwurf zur Publikation 3
- 129 -
(a) wet season 2007
0
5
10
15
0 10 20 30
Ni (
nmol
/L)
dry season 2008
0
5
10
15
0 10 20 30
Ni (
nmol
/L)
dry season 2009
0
5
10
15
0 10 20 30
Ni (
nmol
/L)
(b)
wet season 2007
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Ni (μm
ol/g
)
dry season 2008
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Ni (μm
ol/g
)
dry season 2009
0.0
0.1
0.2
0.3
0.4
0 10 20 30
Ni (μm
ol/g
)
(c)
wet season 2007
0%
25%
50%
75%
100%
0 0 0 1 3 5 9 10 19Salinity
Ni r
atio
HMWLMWUD
dry season 2009
0%
25%
50%
75%
100%
0 0 2 4 8 17 34Salinity
Ni r
atio
HMWLMWUD
Fig. 16. Estuarine distribution of Ni in the Manguaba Lagoon : MG; : Sumaúma;
: coastal water): (a) total dissolved Ni, (b) particulate Ni and (c) Ni percentage of colloidal size fractions; dotted: high molecular weight fraction; hatched: low molecular weight fraction; solid: truly dissolved fraction.
Nickel: Despite some variabilities at low salinities, which may be explained by
fluctuations in the incoming river water (Chaudry and Zwolsman, 2008), dissolved Ni
decreased generally with increasing salinity from typically 10 nmol/L to less than 3
nmol/L in the coastal ocean. This was observed during the wet season and both dry
seasons and in both study areas except in the MD during the dry season of 2008,
during which dissolved Ni concentration was essentially constant at about 2 nmol/L
(Fig. 15a and Fig. 16a). When the high dissolved Ni concentration (15 nmol/L) at the
salinity of 34 psu in the MG during the dry season of 2009 could be identified as
irregular discharge or contamination during the sample pretreatment, dissolved Ni
essentially has a conservative behavior in both study areas and during all three
sample periods. Although also non-conservative estuarine behavior of dissolved Ni
has been reported (e.g., Sholkovitz 1978; Campbell et al., 1988; Wang and Liu,
2003), Ni is largely unreactive and mainly behaves conservatively in estuaries (e.g.,
Danielsson et al., 1983; Dai and Martin, 1995; Turner and Martino, 2006).
Kapitel 6: Entwurf zur Publikation 3
- 130 -
6.3.3. Particulate trace metals
Only during the wet season in the MD (Fig. 5b), the distribution pattern of particulate
Cd resembled the dissolved one. Its concentration decreased rapidly from about 5
nmol/g in the freshwater to about 0.5 nmol/g at high salinities. Such a decrease of
particulate Cd has been observed before, but always combined with an enrichment of
dissolved Cd (e.g., Paucot and Wollast, 1997). Such a “Fe-like” estuarine distribution
of both particulate and dissolved Cd was not found in any other estuary. Rapid co-
precipitation with Fe/Mn oxides (e.g., Turner et al., 2008) might be an explanation;
alternatively, riverine Cd might be removed by formation of Cd sulfides under anoxic
conditions as discussed before (e.g., Wen et al., 2011). During both dry seasons in
the MD, particulate Cd varied between 0.5 and 1.2 nmol/g and showed no
considerable change with salinity, which was widely observed e.g., in the Huanghe
estuary (Elbaz-Poulichet et al., 1987), in several other Chinese estuaries (Zhang and
Liu, 2002) and in the Galveston Bay (Wen et al., 2008). Similar levels for particulate
Cd (about 0.6 nmol/g) were also found in the MG during all three seasons (Fig. 6b)
and appear to be the predominant feature for both lagoons. Although seasonal
variability occurred in both study areas with respect to dissolved Cd, particulate Cd
does not seem to be influenced by seasonality. Simple dilution of particulate Cd is not
only the dominant feature in the MD and MG, but also in other estuaries in the world
(e.g., Paalman and van der Weijden, 1992; Wen et al., 2008).
Similar to the dissolved phase, particulate Cu generally followed the theoretical
dilution line in both study areas and during both wet and dry seasons despite some
scatter. Its value decreased from typically 0.2-0.3 μmol/g in the riverine water to
about 0.1-0.15 μmol/g at high salinities (Fig. 7b and Fig. 8b). Such a conservative
distribution of particulate Cu was often observed, e.g., in some Texas estuaries
(Benoit et al., 1994) and in the Gironde estuary (Kraepiel et al., 1997).
Particulate Fe rapidly decreased from typically 1.5 mmol/g to about 0.5 mmol/g at low
salinities and varied only slightly at higher salinities in both study areas and during
both wet and dry seasons (Fig. 9b and Fig. 10b). A similar distribution was observed
for particulate Pb, which decreased from typically 0.2 μmol/g to about 0.1 μmol/g and
stabilized at this level (Fig. 11b and Fig. 12b). Both particulate Fe and particulate Pb
were removed during the estuarine mixing as it occurs worldwide (e.g., Duinker and
Nolting, 1977; Benoit et al., 1994, Wen et al., 2008) as a consequence of flocculation
or co-precipitation as discussed before.
Particulate Co slightly decreased from typically 0.2 μmol/g at low salinities to about
Kapitel 6: Entwurf zur Publikation 3
- 131 -
0.1 μmol/g at high salinities in both study areas and during all seasons (Fig. 13b and
Fig. 14b). Particulate Ni varied between 0.2 and 0.3 μmol/g in both estuaries and
during all seasons (Fig. 15b and Fig. 16b), when a few low values are neglected.
Despite some fluctuations in the incoming river water, the estuarine distribution of
particulate Co and Ni may be considered as conservative mixing, which was widely
observed e.g., in the Rhine/Meuse estuary (Paalman and van der Weijden, 1992), in
the Hudson River estuary (Tovar-Sánchez et al., 2004) and in the estuaries of East
Hainan (Fu et al., 2013). This may be taken as a confirmation for the assumptions
made before, that the excess of dissolved Co has a sediment origin, and that Ni is
unreactive in the estuaries.
6.3.4. Distribution coefficient
The relative affinity of trace metals for dissolved and particulate phases can be
evaluated from the distribution coefficient KD as the ratio of the particulate (in mol/kg)
to the dissolved (in mol/L) concentration (e.g., Chiffoleau et al., 1994). It is a crucial
parameter for the interpretation of chemical transport in estuaries (Turner et al.,
1992). The logarithms of KD for the individual metals in both the MD and the MG were
plotted versus salinity in Fig. 17.
Among the six metals under investigation, Cd was the only metal, for which the
logarithm of KD clearly decreased from about 5.1 to 4.1 L/kg with increasing salinity
probably due to formation of highly stable Cd-chloro-complexes during the estuarine
mixing (e.g., Chiffoleau et al., 2001). The KD´s of Ni (~ 4.8 L/kg) was constant over
the whole salinity range. The logarithm of KD showed a slight increase with increasing
salinity (from 4.4 to 4.8 L/kg) for Cu, in contrast to the most estuaries worldwide (e.g.,
Balls et al., 1994; Fu et al., 2013), where no significant changes of KD´s with salinity
could be identified. Although both dissolved and particulate Cu appears to behave
conservatively, Cu seemed to favor the particulate phase with increasing salinity in
both study areas. Fe and Pb have the greatest affinity for the particulate phase (e.g.,
Fu et al., 2013). The logarithm of KD increased with increasing salinity from 5.5 to 6.5
L/kg for Fe. Despite the fluctuation at low salinities, the KD´s of Pb varied between 4.9
and 6.5 L/kg, and seem not to be dependent on salinity, which contrasts to the
positive correlation of KD vs. salinity in the estuaries of East Hainan (Fu et al., 2013).
The partial increase of dissolved Pb with salinity (Fig. 11a and Fig. 12a) probably
inhibited the increase of the KD´s. The enrichment of Co in solution and the slight
conservative decrease of particulate Co in both study areas resulted in a slight
Kapitel 6: Entwurf zur Publikation 3
- 132 -
increase of the KD with salinity from 4.9 to 5.2 L/kg. The higher KD´s for Co during the
wet season is a consequence of the lower dissolved concentration during that
season.
Cd
3
4
5
6
7
8
0 10 20 30Salinity
Log 1
0KD (L
/kg)
200720082009
Cu
3
4
5
6
7
8
0 10 20 30Salinity
Log 1
0KD (L
/kg)
Fe
3
4
5
6
7
8
0 10 20 30Salinity
Log 1
0KD (L
/kg)
Pb
3
4
5
6
7
8
0 10 20 30Salinity
Log 1
0KD (L
/kg)
Co
3
4
5
6
7
8
0 10 20 30Salinity
Log 1
0KD (L
/kg)
Ni
3
4
5
6
7
8
0 10 20 30Salinity
Log 1
0KD (L
/kg)
Fig. 17. Log10KD (L/kg) vs. salinity : Wet season of 2007; : Dry season of 2008; : Dry season of 2009).
A general order of log10KD for Fe (6.1 L/kg) > Pb (5.8 L/kg) > Co (5.0 L/kg) > Ni (4.8
L/kg) > Cd (4.6 L/kg) > Cu (4.5 L/kg) was calculated on the basis of the average
values over the whole estuary. This order generally agrees with the results presented
e.g., for the Weser estuary (Turner et al., 1992), for the Seine estuary (Chiffoleau et
al., 1994), for North Australian estuaries (Munksgaard and Parry, 2001) and for the
estuaries in East-Hainan (Fu et al., 2013). The relative magnitudes of the KD´s reflect
the affinity of Fe and Pb for the particulate phase and the favor of Cu and Cd for the
Kapitel 6: Entwurf zur Publikation 3
- 133 -
dissolved phase. Despite the differences in the source and the composition of the
suspended particles, the KD´s obtained in this study differ from the data in other
investigations (e.g., Balls et al., 1994; Kraepiel et al., 1997; Munksgaard and Parry,
2001; Fu et al., 2013) by less than one order of magnitude.
6.3.5. Colloidal trace metals
The colloidal size fraction analysis indicates for all dissolved metals under
consideration an increased importance of the truly dissolved (UD) fraction with
increasing salinity during both investigated seasons and in both study areas. Similar
features were also observed in the Ochlockonee estuary for Cd, Cu and Ni (Powell et
al., 1996) and in the San Francisco Bay for Cu and Fe (Hurst and Bruland, 2008).
During the wet season of 2007, the percentage of the UD-Cd in the total dissolved
fraction decreased from about 50% to about 25% at low salinity (S < 5), and
increased to more than 50% at middle and high salinities in both study areas, while
the HMW-Cd showed a converse trend (Fig. 5c and Fig. 6c). The percentage of the
LMW-Cd seems also to decrease at low salinity, but was constant at middle and high
salinities (20% in the MD and 15% in the MG). During the dry season of 2009, the
UD-Cd increased from 30% in the MD and from 50% in the MG to over 90% in the
salinity range S = 0 to S = 10 and remained constant at higher salinities. The HMW-
Cd decreased from 60% in the MD and 40% in the MG to less than 10% above S ≈
10, while the occurrence of the LMW with 5% on average was not significant. The
general increase of the UD fraction percentage with increasing salinity in both study
areas probably indicated the formation of highly stable Cd-chloro-complexes (e.g.,
Dabrin et al., 2009). During the dry season of 2009, the increase of the UD fraction
percentage could also be related to release of organo-bound-Cd from resuspended
sediment subsequent to its degradation (Waeles et al., 2005). During the wet season
of 2007, the decrease of the UD fraction percentage at low salinity in the MD was
probably attributed to adsorption onto Fe/Mn oxides as discussed. But in the MG, the
percentage of the UD fraction also decreased at low salinity, while total dissolved Cd
was essentially constant (Fig. 6a). An internal conversion between the UD and the
HMW fractions may have taken place. Furthermore, the lower percentage of the UD
in the incoming river water in both study areas during the wet season of 2007 is
combined with lower percentage in the lagoons (50% on average in both lagoons,
respectively), in contrast to the higher UD percentage in the freshwater and in the
lagoons (71% in the MD and 78% in the MG on average, respectively) during the dry
Kapitel 6: Entwurf zur Publikation 3
- 134 -
season of 2009. The presence of colloidal Cd appears to be associated with the
riverine input as observed e.g., in the Hudson River estuary (Yang and Sañudo-
Wilhelmy, 1998). Also the level of colloidal Cd involved in the estuarine chemistry
differs from one estuary to another. For instance, more than 90% of dissolved Cd was
found to be associated with molecules smaller than 10 kDa in the Rhône delta (Dai et
al., 1995) and in the Seine estuary (Chiffoleau et al., 2001), but only 60-70% in the
Venice Lagoon (Martin et al., 1995) and in the Danube River estuary (Guieu et al.,
1998). The partitioning of Cd is influenced by many variables such as pH (e.g., Hatje
et al., 2003b), temperature (e.g., Warren and Zimmerman, 1994), Ca2+ (e.g., Waeles
et al., 2005), Cl- (e.g., Dabrin et al., 2009) and DOC (e.g., Wen et al., 1999). It is
therefore difficult to forecast the colloidal composition of Cd in the estuary. As
discussed in the methodological section, a major contribution to these differences is
probably due to the different operational conditions.
In the MD and MG, the percentage of the UD fraction in total dissolved Cu increased
with increasing salinity from typically 20% to more than 50% during both wet and dry
seasons, while the percentage of the HMW-Cu decreased from typically 40% to less
than 10% (Fig. 7c and Fig. 8c). The LMW fraction with a percentage of 37% on
average during the wet season and 38% during the dry season varied insignificantly
with salinity. The increases of the UD-Cu content at the expense of the HMW-Cu
content were widely observed e.g., in the Rhône delta (Dai et al., 1995), in the San
Francisco Bay (Sañudo-Wilhelmy et al., 1996) and in the Galveston Bay (Wen et al.,
1999). With regard to the conservative distribution of dissolved and particulate Cu
(Fig. 7a, Fig. 7b, Fig. 8a and Fig. 8b), a conversion from the HMW to the UD fraction
might be suggested. Copper is well known to have a strong affinity to organic
functional groups (e.g., Shank et al., 2004). Dissolved and colloidal Cu is controlled
by dominant organic ligands in estuaries (e.g., Wen et al., 1999). Thus, such
conversion could be a consequence of the degradation of high molecular weight
dissolved organic materials (e.g., Masson et al., 2011) and/or the processing by
phytoplankton and bacteria (e.g., Wen et al., 2011) because of the high bio-
productivity in both study areas (Melo-Magalhães et al., 2009). It is also possible that
the removal of dissolved Cu (Fig. 7a and Fig. 8a) originates from a partial flocculation
of large organic colloids, while particulate Cu may associate with strong low
molecular weight ligands (Waeles et al., 2008a). The level of colloidal Cu in estuaries
worldwide varied significantly from 46% of the total dissolved fraction in the Venice
Lagoon (10 kDa - 0.4 μm; Martin et al., 1995) to 94% in the Penzè estuary (5 kDa -
0.45 μm; Waeles et al., 2008a). In both study areas, 57% of total dissolved Cu during
Kapitel 6: Entwurf zur Publikation 3
- 135 -
the wet season and 53% during the dry season could be identified in the colloidal
fraction (5 kDa - 0.45 μm). The highly varying results may be related - at least partly -
to the nature of organic matter within the system of concern (Waeles et al., 2008a).
The HMW of Fe with over 80% of the total dissolved fraction was the most important
fraction in both study areas and during both wet and dry seasons, while the HMW
and the LMW fractions ranged generally less than 10%, respectively (Fig. 9c and Fig.
10c). Despite the small extent of available data, a generally decrease of HMW
fraction and an associated increase of UD fraction tends to occur also for Fe. Both
bulk colloidal Fe-hydroxides and small size colloidal organo-bound-Fe is destabilized
by sea water ions such as Mg2+ and Ca2+ and consequently flocculate (e.g., Boyle et
al., 1977; Nowostawska et al., 2008; Shim et al., 2012). It is remarkable that the UD
and the LMW fraction of Fe during the dry season of 2009 (12% and 11% on average,
respectively) were more important than during the wet season of 2007 (1% and 4%
on average, respectively). With respect to the higher DOC during the dry season of
2009, Fe in the UD and the LMW fractions might be stabilized by humic material
against coagulation (e.g., Batchelli et al., 2010).
Similar to Fe, a complete estimation of the colloidal Pb distribution was difficult
because of the scarcity of usable data. But a decreasing tendency of the HMW
fraction and an increasing tendency of the UD fraction is observable (Fig. 11c and Fig.
12c). Dissolved Pb was mostly found in the HMW fraction (65% in the MD and 88%
in the MG) during the wet season, while the UD and the LMW fraction together
comprised less than one third of total dissolved Pb. Such Fe-like distribution was also
observed e.g., in the Galveston Bay (Benoit et al., 1994) and in the estuaries of East
Hainan (Fu, unpublished; see chapter 5), and was mainly attributed to the removal of
dissolved Pb with freshly precipitated Fe-hydroxides (e.g., Waeles et al., 2008a; Wen
et al., 2008). During the dry season, around 58% of total dissolved Pb was
associated with the UD fraction and only 37% with the HMW fraction in both study
areas. This may be explained by the absence of low salinity samples in the MD. But
the low salinity samples in the MG did contain a significant portion of UD-Pb. As
mentioned before, dissolved Pb was enriched due to release from SPM. It is possible
that the introduced Pb could not be effective scavenged because of the saturation of
the binding sites on suspended particles, or dissolved Pb was bound by strong
organic ligands (Kozelka et al., 1997).
The colloidal partition of Co in estuaries is highly variable. For instance, 81% of total
dissolved Co was reported to reside in the UD fraction (< 1 kDa) throughout the
Galveston Bay (Wen et al., 1999). On the other hand, over 50% of dissolved Co was
Kapitel 6: Entwurf zur Publikation 3
- 136 -
found in colloidal forms (1-100 kDa) in the estuaries in NW Russia (Pokrovsky and
Schott, 2002). Around 40% of UD-Co and 40% of HMW-Co dominated the estuaries
in East Hainan (Fu, unpublished; see chapter 5). In this study, total dissolved Co
contained on average about 37% UD fraction, 36% LMW fraction and 28% HMW
fraction during the wet season, and about 59% UD fraction, 31% LMW fraction and
10% HMW fraction during the dry season (Fig. 13c and Fig. 14c). The percentage of
the UD-Co clearly increased with increasing salinity from typically 25% to 60% during
the wet season and to more than 70% during the dry season, whereas the
percentages of LMW and HMW decreased correspondingly. A formation of stable low
molecular weight (chloro-, sulphato-, carbonato-and organo-) complexes might occur
(e.g., Saito and Moffett, 2001; Takara et al., 2010).
The UD fraction of Ni with a percentage of about 60% on average of total dissolved
fraction during both wet and dry seasons was the dominant size fraction in both study
areas, and clearly increased from typically 30% in the riverine water to more than
70% in the seawater except during the wet season of 2007 because of the scarcity of
usable data (Fig. 15c and Fig. 16c). The LMW-Ni, which represented 30% of total
dissolved Ni during both wet and dry seasons, decreased with increasing salinity, but
less than the HMW-Ni. The dominance of low molecular Ni (UD + LMW; < 10 kDa) in
this study is consistent with the observations e.g., in the Venice Lagoon (Martin et al.,
1995) and in the San Francisco Bay (Sañudo-Wilhelmy et al., 1996) and in the
Mississippi River estuary (Wen et al., 2011). Also an increasing significance of the
UD-Ni fraction with salinity was reported (Powell et al., 1996). Since dissolved and
particulate Ni was considered to behave conservatively as discussed before, the
HMW and the LMW fraction appear to be converted directly into the UD fraction. Ni is
able to complex with organic materials (e.g., Martin et al., 1995; Turner et al., 1998;
Wen et al., 2011). Thus, bulk organic-bound-Ni may also dissociate due to the
replacement by Ca2+ and Mg2+ during the estuarine mixing (e.g., Powell et al., 1996).
Also biological regeneration processes may cause the high proportion of dissolved Ni
in the UD fraction (Dai and Martin, 1995).
Summarizing the seasonal differences in the colloidal distribution, the UD fraction of
all metals was generally more important during the dry season of 2009 (74%, 47%,
12%, 55%, 59% and 58% on average in the whole study area for Cd, Cu, Fe, Pb, Co
and Ni, respectively) than during the wet season of 2007 (50%, 43%, 1%, 12%, 37%
and 59% on average in the whole study areas for Cd, Cu, Fe, Pb, Co and Ni,
respectively).
The dry season of 2009 was characterized by higher water temperatures, higher
Kapitel 6: Entwurf zur Publikation 3
- 137 -
DOC contents and lower water discharges in comparison with the wet season of
2007. The higher water temperature probably enhanced the bioactivity. For some
metals such as Cu (e.g., Wen et al., 2011), Fe (e.g., Hurst and Bruland, 2008) and Ni
(e.g., Dai and Martin, 1995) this may induce their colloidal redistribution. The higher
DOC content provides more binding sites for the metals and may stabilize the metals
in the dissolved phase. For instance, the enrichment of dissolved Cd was found not
only to be due to the formation of Cd-chloro-complexes but also to be due to the
binding of Cd with low-molecular-weight organic ligands (Waeles et al., 2008a).
Particulate Cu is known to be able to associate with low molecular weight organic
ligands and to form strong complexes (Waeles et al., 2008a). The lower water
discharge during the dry seasons led to a longer water residence time in the lagoons.
Some metals (Cd, Co, Fe and Pb; e.g., Takara et al., 2010; Braungardt et al., 2011)
can be mobilized from the particulate phase as either free ions or low molecular
weight organic complexes due to e.g., the more complete particle-water-exchange,
although some metals are rapidly re-scavenged (Caetano et al., 2003). However,
other parameters such as pH and the nature of the incoming colloids can also
influence the seasonal variability of the colloidal partition (e.g., Braungardt et al.,
2011). Since only a few comparable studies have been carried out and because of
the diversity of the environmental variables, it is difficult to estimate the reactions,
their mechanisms and the fate for the estuarine trace metal colloids with regard to
their seasonal variability.
6.3.6. Spatial and seasonal differences
In the concentration and estuarine distribution of the six trace metals, some
differences were found between the MD and the MG. The SPM loading in the MG
was a factor of 2 higher than in the MD. This fact did not cause a remarkable
difference of the trace metal distribution in the two estuaries. Although a release of
Cd during the dry seasons, of Pb during the surveys of 2007 and 2009 and of Co
during all seasons occurred, particulate trace metals behaved conservatively or even
remained constant. The transfer of the studied metals to the solution phase seems to
have a sediment-origin rather than release from the SPM. It might be partly attributed
to the different nature of the river inputs. The higher dissolved Cu and Ni in the MG at
S = 0 could also originals from the incoming river water. The same holds for the
different distribution of Cd in the two study areas during the wet season.
Seasonal differences of the concentration and the estuarine distribution of most
Kapitel 6: Entwurf zur Publikation 3
- 138 -
investigated metals were observed in this study. The dissolved Cd was enriched
more strongly in the channel system during both dry seasons. The lower residence
time and/or the stronger dilution of saline water as a consequence of more
freshwater input during the wet season may inhibit the resolution of Cd from the
resuspended sediment. The dissolved Co level during both dry seasons was a factor
of 2 higher than during the wet season. The unexplainably higher concentration in the
freshwater input may originate such occurrence.
For all the investigated metals, a peak in the diagram of DOC and dissolved metals
vs. salinity was observed at s ≈ 10 in the MD during the dry season of 2009 with
more intensity for Cd, Fe and Pb than for the other metals. This may reflect the
influence of urban outputs from the city of Maceío for the MD (Brockmeyer and Spitzy,
2011). This addition of dissolved trace metals, however, was efficiently leveled out
during estuarine processing.
In the Mundaú-Manguaba Lagoon system, this work estimated an averaged
concentration of 0.03 nmol/L, 7.0 nmol/L, 1.5 μmol/L, 0.3 nmol/L, 1.6 nmol/L and 5.0
nmol/L for dissolved Cd, Cu, Fe, Pb, Co and Ni, respectively, and 0.4 nmol/g, 1.6
nmol/g, 5.7 μmol/g, 1.1 nmol/g, 1.1 nmol/g and 1.8 nmol/g for particulate Cd, Cu, Fe,
Pb, Co and Ni, respectively. Compared to other estuarine systems in South America
such as the Sepetiba Bay (e.g., de Lacerda et al., 1987), the Patos Lagoon (e.g.,
Windom et al., 1999) and the Paraíba do Sul Estuary (e.g., Carvalho et al., 2002) and
to estuaries worldwide such as the Mississippi River Estuary (Shiller and Boyle,
1991), the Gironde Estuary (Kraepiel et al., 1997) and the Changjiang Estuary (Wang
and Liu, 2003), both dissolved and particulate metal concentrations are on the same
level or even lower except Fe probably indicating the different geographic feature in
both study areas. The dissolved and particulate concentrations of investigated trace
metals suggest no significant contamination in the Mundaú-Manguaba Lagoon
system.
6.4. Conclusions
This study provides data along salinity gradients for dissolved and particulate
cadmium, copper, iron, lead, cobalt and nickel concentrations in the surface water of
the Mundaú-Manguaba lagoon system in Northeast Brazil during the wet season
(September/October 2007) and two dry seasons (February 2008 and March 2009).
Some universal estuarine behaviors for metals such as the formation of chloro-
cadmium-complexes, the co-precipitation of iron and lead, the conservative
Kapitel 6: Entwurf zur Publikation 3
- 139 -
distribution of copper due to the strong association with organic materials, the
remobilization of Cobalt and the weak activity of Nickel were also observed in this
investigation. Cadmium underwent the formation of highly stable soluble chloro-
complexes, particularly at high salinities during both dry seasons. The suboxic
conditions in both lagoons possibly inhibited the release of dissolved cadmium from
the sediment. In the channel system, where dissolved oxygen was abundant, the
enrichment of dissolved cadmium occurred as in the most estuaries worldwide. But
during the wet season, both a precipitation in the Mundaú lagoon and a conservative
mixing in the Manguaba lagoon happened. In the MG, lead was likely released from
suspended particulate matter during the surveys of 2007 and 2009 due to higher
particulate matter loads despite its affinity to the particulate phase. Dissolved cobalt
was enriched at low salinity suggesting a release of cobalt from resuspended
sediment.
Distribution coefficients were calculated with the following order of the average
logarithms of KD´s: Fe (6.1 L/kg) > Pb (5.9 L/kg) > Co (4.9 L/kg) > Ni (4.8 L/kg) > Cu
(4.6 L/kg) > Cd (4.5 L/kg). Both the order and the values of the distribution
coefficients essentially agree with other estuarine studies. The distribution coefficient
of cadmium decreased with increasing salinity, while for lead and nickel it showed no
variations with salinity. For copper, iron and cobalt, the distribution coefficient was
positively correlated with salinity.
In addition, the colloidal partitioning of the trace metals and their relation to salinity
was investigated during the wet season of 2007 and the dry season of 2009. Most
metals had a significant presence in the colloidal fraction (5 kDa - 0.45 μm). The
significance of the truly dissolved fraction for almost all the metals increased with
salinity and was more important during the dry season than during the wet season. In
the tropical Mundaú-Manguaba lagoon system, the higher bioactivity and the longer
water residence time during the dry season may favor the degradation of high
molecular weight organo-bound metals and the release of metals from the sediment.
The released metals are probably associated with small organic ligands caused by
higher DOC contents during the dry season, thus enhancing their appearance in the
true dissolved fraction.
Acknowledgments
The authors wish to thank the “Bundesministerium für Bildung und Forschung (BMBF,
German Federal Ministry of Education and Research)” at Berlin and “the Brazilian
Kapitel 6: Entwurf zur Publikation 3
- 140 -
Research Council (CNPq)” who funded our sub-project from the Project POLCAMAR
(POllution from sugar CAne in MARine systems) within the frame of the Brazilian-
German cooperation in science and technology [Marine Sciences Program,
BMBF/CNPq research project number.: 03F0455C (BMBF) and project nr.
590002/2005-8 (CNPq)]. We would like to thank all colleagues from the POLCAMAR
team, in special Professor Dr. Bastiaan Knoppers and Olaf Wilhelm for their
coordinating and managing efforts during the study in Brazil. We thank Dr. Uwe
Schüßler for his help with the ICP-MS analyses.
Literaturverzeichnis
- 141 -
Literaturverzeichnis
Abril, G., Nogueira, B., Etcheber, H., Cabeçadas, G., Lemaire, E., Brogueira, M.J., 2002. Behaviour of organic carbon in nine contrasting European estuaries. Estuarine, Coastal and Shelf Science 54, 241-262.
Achterberg, E.P. and Van den Berg, C.M.G., 1997. Chemical speciation of chromium and nickel in the western Mediterranean. Deep Sea Research Part II: Topical Studies in Oceanography 44, 693-720.
Achterberg, E.P., Braungardt, C.B., Sandford, R.C., Worsfold, P.J., 2001. UV digestion of seawater samples prior to the determination of copper using flow injection with chemiluminescence detection. Analytica Chimica Acta 440, 27-36.
Aston S.R. and Chester, R., 1973. The influence of suspended particles on the precipitation of iron in natural waters. Estuarine and Coastal Marine Science 1, 225-231.
Audry, S., Blanc, G., Schäfer, J., Chaillou, G., Robert, S., 2006. Early diagenesis of trace metals (Cd, Cu, Co, Ni, U, Mo, and V) in the freshwater reaches of a macrotidal estuary. Geochimica et Cosmochimica Acta 70, 2264-2282.
Bach, B.A., 2012. Konzentration und Speziation von verschmutzungsrelevanten Spurenmetallen in tropischen, huminstoffreichen Fluss- und Küstensystemen Ost-Sumatras. Dissertation, Universität Bremen.
Baker, R.W. and Strathmann, H., 1970. Ultrafiltration of macromolecular solutions with high-flux membranes. Journal of Applied Polymer Science 14, 1197-1214.
Balls, P.W., 1989. The partition of trace metals between dissolved and particulate phases in European coastal waters: A compilation of field data and comparison with laboratory studies. Netherlands Journal of Sea Research 23, 7-14.
Balls, P.W., 1990. Distribution and composition of suspended particulate material in the Clyde Estuary and associated sea lochs. Estuarine, Coastal and Shelf Science 30, 475-487.
Balls, P.W., Laslett, R.E., Price, N.B., 1994. Nutrient and trace metal distributions over a complete semi-diurnal tidal cycle in the Forth Estuary, Scotland. Netherlands Journal of Sea Research 33, 1-17.
Balzer, W., 1982. On the distribution of iron and manganese at the sediment/water interface: Thermodynamic versus kinetic control. Geochimica et Cosmochimica Acta 46, 1153-1161.
Balzer, W., Boehler, E., Tang, X.-L., Ren, J.-L., Zhang, J., Wang, D.-R., 2013. Arsenic in solution, colloidal and particulate phases of East-Hainan estuaries. Continental Shelf Research 57, 73-81.
Barceloux, D.G., 1999. Cobalt. Clinical Toxicology 37, 201-216.
Batchelli, S., Muller, F.L.L., Baalousha, M., Lead, J.R., 2009. Size fractionation and optical properties of colloids in an organic-rich estuary (Thurso, UK). Marine Chemistry 113, 227-237.
Batchelli, S., Muller, F.L.L., Chang, K.-C., and Lee, C.-L., 2010. Evidence for strong but dynamic iron−humic colloidal associations in humic-rich coastal waters. Environmental Science and Technology 44, 8485-8490.
Beck, A.J. and Sañudo-Wilhelmy, S.A., 2007. Impact of water temperature and dissolved oxygen on copper cycling in an urban estuary. Environmental Science and Technology 41, 6103-6108.
Bedsworth, W.W. and Sedlak, D.L., 1999. Sources and environmental fate of strongly complexed nickel in estuarine waters: The role of ethylenediaminetetraacetate. Environmental Science and Technology 33, 926-931.
Benoit, G., Oktay-Marshall, S.D., Cantu, A., II, Hood, E.M., Colemann, C.H., Corapcioglu, M.O., Santschi, P.H., 1994. Partitioning of Cu, Pb, Ag, Zn, Fe, Al, and Mn between filter-
Literaturverzeichnis
- 142 -
retained particles, colloids, and solution in six Texas estuaries. Marine Chemistry 45, 307-336.
Benoit, G. and Rozan, T.F., 1999. The influence of size distribution on the particle concentration effect and trace metal partitioning in rivers. Geochimica et Cosmochimica Acta 63, 113-127.
Berner, E.K. and Berner, R.A., 2012. Global Environment: Water, air and geochemical cycles. Princeton University Press, New Jersey.
Blake, A.C., Chadwick, D.B., Zirino, A., Rivera-Duarte, I., 2004. Spatial and temporal variations in copper speciation in San Diego Bay. Estuaries and Coasts 27, 437-447.
Boulègue, J. 1983. Trace metals (Fe, Cu, Zn, Cd) in anoxic environments. In: Wong, C. S., Boyle, E., Bruland, K.W., Burton, J.D., Goldberg, E.D. (Eds.), Trace Metals in Seawater. Plenum Press, New York, pp. 563-577.
Boyle, E.A., Edmond, J.M., Sholkovitz, E.R., 1977. The mechanism of iron removal in estuaries. Geochimica et Cosmochimica Acta 41, 1313-1324.
Boyle, E.A., Huested, S.S., Grant, B., 1982. The chemical mass balance of the Amazon plume—II. Copper, nickel, and cadmium. Deep Sea Research Part A. Oceanographic Research Papers 29, 1355-1364.
Braungardt, C.B., Howell, K.A., Tappin, A.D., Achterberg, E.P., 2011. Temporal variability in dynamic and colloidal metal fractions determined by high resolution in situ measurements in a UK estuary. Chemosphere 84, 423-431.
Brockmeyer, B. and Spitzy, A., 2011. Effects of sugar cane monocultures on origin and characteristics of dissolved organic matter in the Manguaba Lagoon in northeast Brazil. Organic Geochemistry 42, 74-82.
Bruland, K.W., 1983. Trace elements in sea water. In: Riley, J.P., and Chester, R., (Eds.), Chemical Oceanography, Vol. 8. Academic Press, London, pp. 157-220.
Bruland, K.W., Coale, K.H., Mart, L., 1985. Analysis of seawater for dissolved cadmium, copper and lead: An intercomparison of voltammetric and atomic absorption methods. Marine Chemistry 17, 285-300.
Bruland, K.W., Donat, J.R., Hutchins, D.A., 1991. Interactive influences of bioactive trace metals on biological production in oceanic waters. Limnology and Oceanography 36, 1555-1577.
Buckley, P.J.M. and Van den Berg, C.M.G., 1986. Copper complexation profiles in the Atlantic Ocean: A comparative study using electrochemical and ion exchange techniques. Marine Chemistry 19, 281-296.
Buesseler, K.O., Bauer, J.E., Chen, R.-F., Eglinton, T.I., Gustafsson, O., Landing, W., Mopper, K., Moran, S.B., Santschi, P.H., VernonClark, R., Wells, M.L., 1996. An intercomparison of cross-flow filtration techniques used for sampling marine colloids: Overview and organic carbon results. Marine Chemistry 55, 1-31.
Buffle, J. and Leppard, G.G., 1995. Characterization of aquatic colloids and macromolecules. 1. Structure and behavior of colloidal material. Environmental Science and Technology 29, 2169-2175.
Bundesregierung, 1972. Umweltschutz: Das Umweltprogramm der Bundesregierung. Verlag W. Kohlhammer, Stuttgart.
Byrne, R.H., Kump, L.R., Cantrell, K.J., 1988. The influence of temperature and pH on trace metal speciation in seawater. Marine Chemistry 25, 163-181.
Caetano, M., Madureira, M.-J., Vale, V., 2003. Metal remobilisation during resuspension of anoxic contaminated sediment: short-term laboratory study. Water, Air and Soil Pollution 143, 23-40.
Calliari, D., Gómez, M., Gómez, N., 2005. Biomass and composition of the phytoplankton in the Río de la Plata: large-scale distribution and relationship with environmental variables during a spring cruise. Continental Shelf Research 25, 197-210.
Literaturverzeichnis
- 143 -
Campbell, J.A., Whitelaw, K., Riley, J.P., Head, P.C., Jones, P.D., 1988. Contrasting behaviour of dissolved and particulate nickel and zinc in a polluted estuary. The Science of the Total Environment 71, 141-155.
Capodaglio, G., Coale, K.H., Bruland, K.W., 1990. Lead speciation in surface waters of the eastern North Pacific. Marine Chemistry 29, 221-233.
Carvalho, C.E.V., Salomão, M.S.M.B., Molisani, M.M., Rezende, C.E., Lacerda, L.D., 2002. Contribution of a medium-sized tropical river to the particulate heavy-metal load for the South Atlantic Ocean. Science of the Total Environment 284, 85-93.
Chaudry, M.A. and Zwolsman, J.J.G., 2008. Seasonal dynamics of dissolved trace metals in the Scheldt Estuary: Relationship with redox conditions and phytoplankton activity. Estuaries and Coasts 31, 430-443.
Chester, R., 1990. Marine Geochemistry. Unwin Hyman, London.
Chiffoleau, J.-F., Cossa, D., Auger, D., Truquet, I., 1994. Trace metal distribution, partition and fluxes in the Seine Estuary (France) in low discharge regime. Marine Chemistry 47, 145-158.
Chiffoleau, J.-F., Auger, D., Chartier, E., Michel, P., Truquet, I., Ficht, A., Gonzalez, J.-L., Romana, L.-A., 2001. Spatiotemporal changes in cadmium contamination in the Seine Estuary (France). Estuaries and Coasts 24, 1029-1040.
Collins, R.N. and Kinsela, A.S., 2010. The aqueous phase speciation and chemistry of cobalt in terrestrial environments. Chemosphere 79, 763-771.
Comans, R.N.J. and van Dijk, C.P.J., 1988. Role of complexation processes in cadmium mobilization during estuarine mixing. Nature, 336, 151-154.
Comber, S.D.W., Gunn, A.M., Whalley, C., 1995. Comparison of the partitioning of trace metals in the Humber and Mersey Estuaries. Marine Pollution Bulletin 30, 851-860.
Ćosović, B., Degobbis, D., Bilinski, H., Branica, M., 1982. Inorganic cobalt species in seawater. Geochimica et Cosmochimica Acta 46, 151-158.
Costa, T.L.F., Araújo, M.P., Knoppers, B.A., Carreira, R.S., 2011. Sources and distribution of particulate organic matter of a tropical estuarine-lagoon system from NE Brazil as indicated by lipid biomarkers. Aquatic Geochemistry 17, 1-19.
Crump, B.C., Hopkinson, C.S., Sogin, M.L., Hobbie, J.E., 2004. Microbial biogeography along an estuarine salinity gradient: Combined influences of bacterial growth and residence time. Applied and Environmental Microbiology 78, 1494-1505.
Dabrin, A., Schäfer, J., Blanc, G., Strady, E., Masson, M., Bossy, C., Castelle, S., Girardot, N., Coynel, A., 2009. Improving estuarine net flux estimates for dissolved cadmium export at the annual timescale: Application to the Gironde Estuary. Estuarine, Coastal and Shelf Science 84, 429-439.
Dai, M.-H. and Martin, J.-M., 1995. First data on trace metal level and behaviour in two major Arctic river-estuarine systems (Ob and Yenisey) and in the adjacent Kara Sea, Russia. Earth and Planetary Science Letters 131, 127-141.
Dai, M.-H., Martin, J.-M., Cauwet, G., 1995. The significant role of colloids in the transport and transformation of organic carbon and associated trace metals (Cd, Cu and Ni) in the Rhône Delta (France). Marine Chemistry 51, 159-175.
Dai, M.-H., Buesseler, K.O., Ripple, P., Andrews, J., Belastock, R.A., Gustafsson, Ö., Moran, S.B., 1998. Evaluation of two cross-flow ultrafiltration membranes for isolating marine organic colloids. Marine Chemistry 62, 117-136.
Danielsson, L.–G., Magnusson, B., Westerlund, S., 1978. An improved metal extraction procedure for the determination of trace metals in sea water by atomic absorption spectrometry with electrothermal atomization. Analytica Chimica Acta 98, 47-57.
Danielsson, L.-G., Magnusson, B., Westerlund, S., Zhang, K. -R., 1983. Trace metals in the Göta River Estuary. Estuarine, Coastal and Shelf Science 17, 73-85.
Literaturverzeichnis
- 144 -
Davies, J.L., 1964. A morphogenetic approach to world shoreline. Zeitschrift für Geomorphologie 8, 127-142.
Davison, W., Phillips, N., Tabner, B.J., 1999. Soluble iron sulfide species in natural waters: Reappraisal of their stoichiometry and stability constants. Aquatic Science 61, 23-43.
De Luca Rebello, A., Haekel, W., Moreira, J., Santilli, R., Schroeder, F., 1986. The fate of heavy metals in an estuarine tropical system. Marine Chemistry 18, 215-225.
De Souza, L., Machado, E.C., Knoppers, B.A., 2002. Mundaú/Manguaba Coastal Lagoon System. In: Camacho-Ibar, V., Dupra, V., Wulff, F., Smith, S.V., Marshall Crossland, J.I., Crossland, C.J., (Eds.), 2002. Estuarine systems of the Latin American region (regional workshop V) and estuarine systems of the Arctic region: Carbon, nitrogen and phosphorus fluxes, LOICZ Reports and Studies 23. LOICZ, Texel, Netherlands, 34-37.
Diegues, A.C.S. (Eds.), 1994. An inventory of Brazilian wetlands. IUCN, Gland, Switzerland.
Donat, J.R., Lao, K.A., Bruland, K.W., 1994. Speciation of dissolved copper and nickel in South San Francisco Bay: A multi-method approach. Analytica Chimica Acta 284, 547-571.
Doucet, F.J., Lead, J.R., Santschi, P.H., 2007. Colloid - Trace element interactions in aquatic systems. In: Wilkinson, K.J. and Lead, J.R. (Eds.), Environmental Colloids and Particles. IUPAC Series on Analytical and Physical Chemistry of Environmental Systems, Vol. 10. John Wiley & Sons, pp. 95–158.
Duce, R.A. and Tindale, N.W., 1991. Atmospheric transport of iron and its deposition in the ocean. Limnology and Oceanography 36, 1715-1726.
Duinker, J.C. and Nolting, R.F., 1977. Dissolved and particulate trace metals in the Rhine Estuary and the Southern Bight. Marine Pollution Bulletin 8, 65-71.
Duinker J.C., 1986. Formation and transformation of element species in estuaries. In: Bernhard, M., Brinckman, F.E., Sadler P.J. (Eds.), The Importance of Chemical Speciation in Environmental Processes. Springer-Verlag, Berlin, pp. 365–384.
Edmond, J.M., Spivack, A., Grant, B.C., Hu, M.-H., Chen, Z.-X., Chen, S., Zeng, X.-S., 1985. Chemical dynamics of the Changjiang Estuary. Continental Shelf Research 4, 17-36.
Elbaz-Poulichet, F., Holliger, P., Huang, W.-W., Martin, J.-M., 1984. Lead cycling in estuaries, illustrated by the Gironde Estuary, France. Nature 208, 409-414.
Elbaz-Poulichet, F., Martin, J.M., Huang, W.-W., Zhu, J.-X., 1987. Dissolved Cd behaviour in some selected French and Chinese estuaries. Consequences on Cd supply to the ocean. Marine Chemistry 22, 125-136.
Elbaz-Poulichet, F., Garnier, J.-M., Guan, D.-M., Martin, J.-M., Thomas, A.J., 1996. The conservative behaviour of trace metals (Cd, Cu, Ni and Pb) and As in the surface plume of stratified estuaries: Example of the Rhône River (France). Estuarine, Coastal and Shelf Science 42, 289-310.
Ellwood, M.J. and Van den Berg, C.M.G., 2001. Determination of organic complexation of cobalt in seawater by cathodic stripping voltammetry. Marine Chemistry 75, 33-47.
Ellwood, M.J., 2004. Zinc and cadmium speciation in subantarctic waters east of New Zealand. Marine Chemistry 87, 37-58.
Everett, D.H., 1988. Basic Principles of Colloid Science. Royal Society of Chemistry, London.
Fang, T.-H. and Lin, C.-L., 2002. Dissolved and particulate trace metals and their partitioning in a hypoxic estuary: The Tanshui Estuary in Northern Taiwan. Estuaries and Coasts 25, 598-607.
Farag, A.M., Nimick, D.A., Kimball, B.A., Church, S.E., Harper, D.D., Brumbaugh, M.G., 2007. Concentrations of metals in water, sediment, biofilm, benthic macroinvertebrates, and fish in the Boulder River watershed, Montana, and the role of colloids in metal uptake. Archives of Environmental Contamination and Toxycology 52, 397-409.
Fernández, S., Villanueva, U., de Diego, A., Arana, G., Madariaga, J.M., 2008. Monitoring
Literaturverzeichnis
- 145 -
trace elements (Al, As, Cr, Cu, Fe, Mn, Ni and Zn) in deep and surface waters of the estuary of the Nerbioi-Ibaizabal River (Bay of Biscay, Basque Country). Journal of Marine Systems 72, 332-341.
FGG Weser, 2012. Wasserqualität im Wesereinzugsgebiet, die Flussgebietsgemeinschaft Weser. www.fgg-weser.de.
Filella, M., 2007. Colloidal properties of submicron particles in natural waters. In: Wilkinson, K.J. and Lead, J.R. (Eds.), Environmental Colloids and Particles. IUPAC Series on Analytical and Physical Chemistry of Environmental Systems, Vol. 10. John Wiley & Sons, pp. 17–93.
Fox, L.E. and Wofsy, S.C., 1983. Kinetics of removal of iron colloids from estuaries. Geochimica et Cosmochimica Acta 47, 211-216.
Fu, J., Tang, X.-L., Zhang, J., Balzer, W., 2013. Estuarine modification of dissolved and particulate trace metals in major rivers of East-Hainan. Continental Shelf Research 57, 59-72.
Gaffney, J.W., White, K.N., Boult, S., 2008. Oxidation state and size of Fe controlled by organic matter in natural waters. Environmental Science and Technology 42, 3575-3581.
Galler, J., 1999. Lehrbuch Umweltschutz. Ecomed Verlagsgesellschaft, Landsberg.
Gao, J.-H., Chen, G.-Q., Ou, W.-X., Zhu, D.-K., 2004. The coast evolution and regulation in Wanquan River Estuary, Hainan Island. Journal of Geographical Science 14, 375-381.
Ge, C.-D., Slaymaker, O., Pedersen, T.F., 2003. Change in the sedimentary environment of Wanquan River Estuary, Hainan Island, China. Chinese Science Bulletin 48, 2357-2361.
Gérente, C., Couespel du Mesnil, P., Andrès, Y., Thibault, J.-F., Le Cloirec, P., 2000. Removal of metal ions from aqueous solution on low cost natural polysaccharides: Sorption mechanism approach. Reactive and Functional Polymers 46, 135-144.
Gledhill, M., Nimmo, M., Hill, S.J., Brown, M.T., 1997. The toxicity of copper (II) species to marine algae, with particular reference to macroalgae. Journal of Phycology 33, 2-11.
Golimowski, J. and Golimowska, K., 1996. UV-photooxidation as pretreatment step in inorganic analysis of environmental samples. Analytica Chimica Acta 325, 111-133.
Guieu, C., Martin, J.-M., Tankéré, S.P.C., Mousty, F., Trincherini, P., Bazot, M., Dai, M.-H., 1998. On trace metal geochemistry in the Danube River and Western Black Sea. Estuarine, Coastal and Shelf Science 47, 471-485.
Guinder, V.A., Popovich, C.A., Perillo, G.M.E., 2009. Particulate suspended matter concentrations in the Bahía Blanca Estuary, Argentina: implication for the development of phytoplankton blooms. Estuarine, Coastal and Shelf Science 85, 157-165.
Guo, H.-M., Zhang, B., Zhang, Y., 2011. Control of organic and iron colloids on arsenic partition and transport in high arsenic groundwaters in the Hetao Basin, Inner Mongolia. Applied Geochemistry 26, 360-370.
Guo L.-D. and Santschi, P.H., 1996. A critical evaluation of the cross-flow ultrafiltration technique for sampling colloidal organic carbon in seawater. Marine Chemistry 55, 113-127.
Guo, L.-D., Wen, L.-S., Tang, D.-G., Santschi, P.H., 2000a. Re-examination of cross-flow ultrafiltration for sampling aquatic colloids: Evidence from molecular probes. Marine Chemistry, 69, 75-90.
Guo, L.-D., Santschi, P.H., Warnken, K.W., 2000b. Trace metal composition of colloidal organic material in marine environments. Marine Chemistry 70, 257-275.
Guo, L.-D. and Santschi, P.H., 2007. Ultrafiltration and its applications to sampling and characterization of aquatic colloids. In: Wilkinson, K.J. and Lead, J.R. (Eds.), Environmental Colloids and Particles. IUPAC Series on Analytical and Physical Chemistry of Environmental Systems, Vol. 10. John Wiley & Sons, pp. 159–221.
Literaturverzeichnis
- 146 -
Gustafsson, Ö., Buesseler, K.O., Gschwend, P.M., 1996. On the integrity of cross-flow filtration for collecting marine organic colloids. Marine Chemistry 55, 93-111.
Gustafsson, Ö. and Gschwend, P.M., 1997. Aquatic colloids: Concepts, definitions, and current challenges. Limnology and Oceanography 42, 519-528.
Hamilton, E.I., 1994. The geobiochemistry of cobalt. Science of the Total Environment 150, 7-39.
Harper, D.J., 1991. The distribution of dissolved cadmium, lead and copper in the Bristol Channel and the outer Severn Estuary. Marine Chemistry 33, 131-143.
Hatje, V., Apte, S.C., Hales, L.T., Birch, G.F., 2003a. Dissolved trace metal distributions in Port Jackson Estuary (Sydney Harbour), Australia. Marine Pollution Bulletin 46, 719-730.
Hatje, V., Payne, T.E., Hill, D.M., McOrist, G., Szymczak, R., 2003b. Kinetics of trace element uptake and release by particles in estuarine waters: Effects of pH, salinity, and particle loading. Environment international 29, 619-629.
Helmers, 1994. Speciation of cadmium in seawater — A direct voltammetric approach. Fresenius' Journal of Analytical Chemistry 350, 62-67.
Hem, J.D., 1970. Study and Interpretation of the Chemical Characteristics of Natural Water, Second Edition. United States Government Printing Office, Washington.
Herbeck, L.S., Unger, D., Wu, Y., Jennerjahn, T.C., 2013. Effluent, nutrient and organic matter export from shrimp and fish ponds causing eutrophication in coastal and back-reef waters of NE Hainan, tropical China. Continental Shelf Research 57, 92-104.
Hong, H.-S., Wang, M.-H., Huang, X.-G., Wang, D.-Z., 2009. Effects of macronutrient additions on nickel uptake and distribution in the dinoflagellate Prorocentrum donghaiense Lu. Environmental Pollution 157, 1933-1938.
Hurst, M.P. and Bruland, K.W., 2008. The effects of the San Francisco Bay plume on trace metal and nutrient distributions in the Gulf of the Farallones. Geochimica et Cosmochimica Acta 72, 395-411.
Järup, L. and Åkesson, A., 2009. Current status of cadmium as an environmental health problem. Toxicology and Applied Pharmacology 238, 201-208.
Jenner, G.A, Longerich, H.P., Jackson, S.E., Fryer, B.J., 1990. ICP-MS — A powerful tool for high-precision trace-element analysis in earth sciences: Evidence from analysis of selected U.S.G.S. reference samples. Chemical Geology 83, 133-148.
Jiann, K.-T., Wen, L.-S., Santschi, P.H., 2005. Trace metal (Cd, Cu, Ni and Pb) partitioning, affinities and removal in the Danshuei River Estuary, a macro-tidal, temporally anoxic estuary in Taiwan. Marine Chemistry 96, 293-313.
Jiann K.-T. and Wen, L.-S., 2009. Intra-annual variability of distribution patterns and fluxes of dissolved trace metals in a subtropical estuary (Danshuei River, Taiwan). Journal of Marine Systems 75, 87-99.
Judson, R., Richard, A., Dix,D.J., Houck, K., Martin, M., Kavlock, R., Dellarco, V., Henry, T., Holderman, T., Sayre, P., Tan, S., Carpenter, T., Smith, E., 2009. The toxicity data landscape for environmental chemicals. Environmental Health Perspectives 117, 685–695.
Kim, J.H., Gibb, H.J., Howe, P.D., 2006. Cobalt and Inorganic Cobalt Compounds (Concise International Chemical Assessment Document 69). Wissenschaftliche Verlagsgesellschaft mbH, Stuttgart.
Kjerfve, B. and Magill, K.E., 1989. Geographic and hydrodynamic characteristics of shallow coastal lagoons. Marine Geology 88, 187-199.
Kjerfve, B., 1994. Coastal lagoons. In: Kjerfve, B., (Eds.), 1994. Coastal Lagoon Processes. Elsevier Science B.V., Amsterdam, Netherlands, 1-8.
Knoppers, B., Kjerfve, B., Carmouze, J.-P., 1991. Trophic state and water turn-over time in six choked coastal lagoons in Brazil. Earth and Environmental Science 14, 149-166.
Literaturverzeichnis
- 147 -
Koch, R., 1995. Umweltchemikalien: Physikalisch-Chemische Daten, Toxizitäten, Grenz-und Richtwerte, Umweltverhalten, dritte Auflage. VCH, Weinheim.
Kogut, M.B. and Voelker, B.M., 2003. Kinetically inert Cu in coastal waters. Environmental Science and Technology 37, 509-518.
Kola, H. and Wilkinson, K.J., 2005. Cadmium uptake by a green alga can be predicted by equilibrium modelling. Environmental Science and Technology 39, 3040-3047.
Komjarova, I. and Blust, R., 2006. Comparison of liquid–liquid extraction, solid-phase extraction and co-precipitation preconcentration methods for the determination of cadmium, copper, nickel, lead and zinc in seawater. Analytica Chimica Acta 576, 221-228.
Kosta, L., 1982. Contamination as a limiting parameter in trace analysis. Talanta 29, 985-992.
Kottelat, R., Vignati, D.A.L., Chanudat, V., Dominik, J., 2008. Comparison of small- and large-scale ultrafiltration systems for organic carbon and metals in freshwater at low concentration factor. Water, Air and Soil Pollution 187, 343-351.
Kozar, S., Bilinski, H., Branica, M., Schwuger, M.J., 1992. Adsorption of Cd(II) and Pb(II) on bentonite under estuarine and seawater conditions. Science of the Total Environment 121, 203-216.
Kozelka, P.B., Sañudo-Wilhelmy, S., Flegal, A.R., Bruland, K.W., 1997. Physico-chemical speciation of lead in South San Francisco Bay. Estuarine, Coastal and Shelf Science 44, 649-658.
Kraepiel, A.M.L., Chiffoleau, J.-F., Martin, J.-M., Morel, F.M.M., 1997. Geochemistry of trace metals in the Gironde Estuary. Geochimica et Cosmochimica Acta 61, 1421-1436.
Kremling, K., 1985. The distribution of cadmium, copper, nickel, manganese, and aluminium in surface waters of the open Atlantic and European shelf area. Deep Sea Research Part A 32, 531-555.
Krüger, G.C.T., de Carvalho, C.E.V., Suzuki, M.S., 2006. Dissolved nutrient, chlorophyll-a and DOC dynamic under distinct riverine discharges and tidal cycles regimes at the Paraíba do Sul River Estuary, R.J., Brazil. Journal of Coastal Research 39, 724-730.
Lacerda, L.D, Pfeiffer, W.C., Fiszman, M., 1987. Heavy metal distribution, availability and fate in Sepetiba Bay, S.E. Brazil. The Science of the Total Environment 65, 163-173.
Lacerda, L.D., Marins, R.V., Paraquetti, H.H.M., Mounier, S., Benaim, J., Fevrier, D., 2001. Mercury distribution and reactivity in waters of a subtropical coastal lagoon, Sepetiba Bay, SE Brazil. Journal of the Brazilian Chemical Society 12, 93-98.
Laglera, L.M. and van den Berg, C.M.G., 2009. Evidence for geochemical control of iron by humic substances in seawater. Limnology and Oceanography 54, 610-619.
Lane, T.W. and Morel, F.M.M., 2000. A biological function for cadmium in marine diatoms. Proceedings of the National Academy of Sciences of the United States of America, 97, 4627–4631.
Lapp, B. and Balzer, W.,1994. Early diagenesis of trace metals used as an indicator of past productivity changes in coastal sediments. Geochimica et Cosmochimica Acta 57, 4639-4652.
Laxen, D.P.H. and Chandler, I.M., 1983. Size distribution of iron and manganese species in freshwaters. Geochimica et Cosmochimica Acta 47, 731-741.
Lead, J.R. and Wilkinson, K.J., 2007. Environmental colloids and particles: Current knowledge and future developments. In: Wilkinson, K.J. and Lead, J.R. (Eds.), Environmental Colloids and Particles. IUPAC Series on Analytical and Physical Chemistry of Environmental Systems, Vol. 10. John Wiley & Sons, pp. 1–16.
Li, R.-H., Liu, S.-M., Zhang, G.-L., Ren, J.-L., Zhang, J., 2013. Biogeochemistry of nutrients in an estuary affected by human activities: The Wanquan River Estuary, eastern Hainan Island, China. Continental Shelf Research 57, 18-31.
Literaturverzeichnis
- 148 -
Li, Y.-H., Burkhardt, L., Teraoka, H., 1984. Desorption and coagulation of trace elements during estuarine mixing. Geochimica et Cosmochimica Acta 48, 1879-1884.
Liang, Q., Jing, H., Gregoire, D.C., 2000. Determination of trace elements in granites by inductively coupled plasma mass spectrometry. Talanta 51, 507-513.
Lion, L.W., Altmann, R.S., Leckie, J.O., 1982. Trace-metal adsorption characteristics of estuarine particulate matter: evaluation of contributions of iron/manganese oxide and organic surface coatings. Environmental Science and Technology 16, 660-666.
Liu, S.-M, Li, R.-H., Zhang, G.-L., Wang, D.-R., Herbeck, L.S., Zhang, J., Ren, J.-L., 2011. The inpact of anthropogenic activities on nutrient dynamics in the tropical Wenchanghe and Wenjiaohe Estuary and Lagoon System in East Hainan, China. Marine Chemistry 125, 49-68.
Liu, X.-W. and Millero, F.J., 1999. The solubility of iron hydroxide in sodium chloride solutions. Geochimica et Cosmochimica Acta 63, 3487-3497.
Luciani, X., Mounier, S., Paraquetti, H.H.M., Redon, R., Lucas, Y., Bois, A., Lacerda, L.D., Raynaud, M., Ripert, M., 2007. Tracing of dissolved organic matter from the Sepetiba Bay (Brazil) by PARAFAC analysis of total luminescence matrices. Marine Environmental Research 65, 148-157.
Luoma, S.N., Van Geen, A., Lee, B.-G., Cloern, J.E, 1998. Metal uptake by phytoplankton during a bloom in South San Francisco Bay: Implications for metal cycling in estuaries. Limnology and Oceanography 43, 1007-1016.
Lyvén, B., Hassellöv, M., Turner, D.R., Haraldsson, C., Andersson, K., 2003. Competition between iron- and carbon-based colloidal carriers for trace metals in a freshwater assessed using flow field-flow fractionation coupled to ICPMS. Geochimica et Cosmochimica Acta 67, 3791-3802.
Ma, J.-L., Wei, G.-J., Xu, Y.-G., Long, W.-G., Sun, W.-D., 2007. Mobilization and re-distribution of major and trace elements during extreme weathering of basalt in Hainan Island, South China. Geochimica et Cosmochimica Acta 71, 3223-3237.
Mandal, R., Hassan, N.M., Murimboh, J., Chakrabarti, C.L., Back, M.H., Rahayu, U., Lean, D.R.S., 2002. Chemical speciation and toxicity of nickel species in natural waters from the Sudbury Area (Canada). Environmental Science and Technology 36, 1477-1484.
Martin, J.-M., Dai, M.-H., Cauwet, G., 1995. Significance of colloids in the biogeochemical cycling of organic carbon and trace metals in the Venice Lagoon (Italy). Limnology and Oceanography 40, 119-131.
Martino, M., Turner, A., Millward, G.E., 2002. Resuspension, reactivity and recycling of trace metals in the Mersey Estuary, UK. Marine Chemistry 77, 171-186.
Martino, M., Turner, A., Millward, G.E., 2003. Influence of organic complexation on the adsorption kinetics of nickel in river waters. Environmental Science and Technology 37, 2383-2388.
Martino, M., Turner, A., Nimmo, M., 2004. Distribution, speciation and particle-water interactions of nickel in the Mersey Estuary, UK. Marine Chemistry 88, 161-177.
Masson, M., Blanc, G., Schäfer, J., Parlanti, E., Le Coustumer, P., 2011. Copper addition by organic matter degradation in the freshwater reaches of a turbid estuary. Science of the Total Environment 409, 1539-1549.
Mayer, L.M., 1982. Aggregation of colloidal iron during estuarine mixing: kinetics, mechanism, and seasonality. Geochimica et Cosmochimica Acta 46, 2527-2535.
Melo-Magalhães, E.M., Medeiros, P.R.P., Lira, M.C.A., Koening, M.L., Moura, A.N., 2009. Determination of eutrophic areas in Mundaú/Manguaba Lagoons, Alagoas-Brazil, through studies of the phytoplanktonic community. Brazilian Journal of Biology, 69, 271-280.
Mildenberger, F., 2009. Die Geburt der Umwelt. Werk und Wirkung Jakob v. Uexkülls (1864-1944). In: Herrmann, B. (Hg.), Beiträge zum Göttinger Umwelthistorischen Kolloquium
Literaturverzeichnis
- 149 -
(2009-2010). Universitätsverlag Göttingen, Göttingen, pp. 1-26.
Mill, A.J.B., 1980. Colloidal and macromolecular forms of iron in natural waters 1: A review. Environmental Technology Letters 1, 97-108.
Millero, F.J., 1995. Solubility of Fe(III) in seawater. Earth and Planetary Science Letters 154, 323-329.
Millero, F.J., Woosley, R., Ditrolio, B., Waters, J., 2009. Effect of ocean acidification on the speciation of metals in seawater. Oceanography 22, 72-85.
Moffett, J.W. and Ho, J., 1996. Oxidation of cobalt and manganese in seawater via a common microbially catalyzed pathway. Geochimica et Cosmochimica Acta 60, 3415-3424.
Monbet, P., 2006. Mass balance of lead through a small macrotidal estuary: the Morlaix River Estuary (Brittany, France). Marine Chemistry 98, 59-80.
Mota, A.M. and Correia Dos Santos, M.M., 1995. Trace metal speciation of labile chemical species in natural waters: Electrochemical methods. In: Tessier, A. and Turner, D.R. (Eds.), Metal Speciation and Bioavailability in Aquatic Systems. John Wiley & Sons, Chichester, pp. 205-257.
Mouvet and Bourg, 1983. Speciation (including adsorbed species) of copper, lead, nickel and zinc in the Meuse River: Observed results compared to values calculated with a chemical equilibrium computer program. Water Research 17, 641-649.
Munksgaard, N.C. and Parry, D.L., 2001. Trace metals, arsenic and lead isotopes in dissolved and particulate phases of North Australian coastal and estuarine seawater. Marine Chemistry 75, 165-184.
Mylon, S.E., Chen, K.-L., Elimelech, M., 2004. Influence of natural organic matter and ionic composition on the kinetics and structure of hematite colloid aggregation: Implications to iron depletion in estuaries. Langmuir 20, 9000-9006.
Nelson, Y.M., Lion, L.W., Shuler, M.L., Ghiorse, W.C., 1999. Lead binding to metal oxide and organic phases of natural aquatic biofilms. Limnology and Oceanography 44, 1715-1729.
Newell, A.D. and Sanders, J.G., 1986. Relative copper binding capacities of dissolved organic compounds in a coastal-plain estuary. Environmental Science and Technology 20, 817-821.
Niencheski, L.F. and Baumgarten, M.G.Z., 2000. Distribution of particulate trace metal in the southern part of the Patos Lagoon Estuary. Aquatic Ecosystem Health and Management 3, 515-520.
Nolan, C.V., Fowler, S.W., Teyssie, J.-L., 1992. Cobalt speciation and bioavailability in marine organisms. Marine Ecology Progress Series 88, 105-116.
Norisuye, K., Ezoe, M., Nakatsuka, S., Umetani, S., Sohrin, Y., 2007. Distribution of bioactive trace metals (Fe, Co, Ni, Cu, Zn and Cd) in the Sulu Sea and its adjacent seas. Deep Sea Research Part II: Topical Studies in Oceanography 54, 14-37.
Norrman, B., Zweifel, U.L., Hopkinson, Jr., C.S., Fry, B., 1995. Production and utilization of dissolved organic carbon during an experimental diatom bloom. Limnology and Oceanography 40, 898-907.
Nowostawska, U., Kim, J.P., Hunter, K.A., 2008. Aggregation of riverine colloidal iron in estuaries: A new kinetic study using stopped-flow mixing. Marine Chemistry 110, 205-210.
Oliveira, A.M. and Kjerfve, B., 1993. Environmental responses of a tropical coastal lagoon system to hydrological variability: Mundaú–Manguaba, Brazil. Estuarine, Coastal and Shelf Science 37, 575-591.
Paalman, M.M.A. and van der Weijden, C.H., 1992. Trace metals in suspended matter from the Rhine/Meuse Estuary. Netherlands Journal of Sea Research 29, 311-321.
Pakhomova, S.V., Hall, P.O.J., Kononets, M.Y., Rozanov, A.G., Tengberg, A., Vershinin, A.V.,
Literaturverzeichnis
- 150 -
2007. Fluxes of iron and manganese across the sediment–water interface under various redox conditions. Marine Chemistry 107, 319-331.
Paucot, H. and Wollast, R., 1997. Transport and transformation of trace metals in the Scheldt Estuary. Marine Chemistry 58, 229-244.
Pham, M.K. and Garnier, J.-M., 1998. Distribution of trace elements associated with dissolved compounds (< 0.45 μm - 1 nm) in freshwater using coupled (frontal cascade) ultrafiltration and chromatographic separations. Environmental Science and Technology 32, 440-449.
Pigeon, C., Ilyin, G., Courselaud, B., Leroyer, P., Turlin, B., Brissot, P., Loréal, O., 2001. A new mouse liver-specific gene, encoding a protein homologous to human antimicrobial peptide hepcidin, is overexpressed during iron overload. The Journal of Biological Chemistry 276, 7811-7819.
Pokrovsky, O.S. and Schott, J., 2002. Iron colloids/organic matter associated transport of major and trace elements in small boreal rivers and their estuaries (NW Russia). Chemical Geology 190, 141-179.
Porfirio, Z., Ribeiro, M.P., Estevam, C.S., Houly, R.L.S., Sant'Ana, A.E.G, 1999. Hepatosplenomegaly caused by an extract of cyanobacterium Microcystis Aeruginosa bloom collected in the Manguaba Lagoon, Alagoas – Brazil. Revista de Microbiologia 30, 278-285.
Powell, R.T., Landing, W.M., Bauer, J.E., 1996. Colloidal trace metals, organic carbon and nitrogen in a southeastern U.S. estuary. Marine Chemistry 55, 165-176.
Pritchard, D.W., 1967. What is an estuary: Physical viewpoint. In: Lauff, G.H. (Eds.), Estuaries. American Association for the Advancement of Science Publication, Washington D.C., pp. 3-5.
Qian, J., Xue, H.-B., Sigg, L., Albrecht, A., 1998. Complexation of cobalt by natural ligands in freshwater. Environmental Science and Technology 32, 2043-2050.
Ren, H.-M., Liu, H.-J., Qu, J.-H., Berg, M., Qi, W.-X., Xu, W., 2010. The influence of colloids on the geochemical behavior of metals in polluted water using as an example Yongdingxin River, Tianjin, China. Chemosphere 78, 360-367.
Rodushkin, I. and Ruth, T., 1997. Determination of trace metals in estuarine and sea-water reference materials by high resolution inductively coupled plasma mass spectrometry. Journal of Analytical Atomic Spectrometry 12, 1181-1185.
Rose, A.L. and Waite, T.D., 2005. Reduction of organically complexed ferric iron by superoxide in a simulated natural water. Environmental Science and Technology 39, 2645-2650.
Rue, E.L. and Bruland, K.W., 1995. Complexation of iron(III) by natural organic ligands in the Central North Pacific as determined by a new competitive ligand equilibration/adsorptive cathodic stripping voltammetric method. Marine Chemistry 50,117-138.
Sandroni, V., Smith, C.M.M., Donovan, A., 2003. Microwave digestion of sediment, soils and urban particulate matter for trace metal analysis. Talenta 60, 715-723.
Santschi, P.H., Lenhart, J.J., Honeyman, B.D., 1997. Heterogeneous processes affecting trace contaminant distribution in estuaries: The role of natural organic matter. Marine Chemistry 58, 99-125.
Saito, M.A. and Moffett, J.W., 2001. Complexation of cobalt by natural organic ligands in the Sargasso Sea as determined by a new high-sensitivity electrochemical cobalt speciation method suitable for open ocean work. Marine Chemistry 75, 49-68.
Salomons, W., 1980. Adsorption processes and hydrodynamic conditions in estuaries. Environmental Technology Letters 1, 356-365.
Santos-Echeandia, J., Laglera, L.M., Prego, R., van den Berg, C.M.G., 2008. Dissolved copper speciation behaviour during estuarine mixing in the San Simon Inlet (wet
Literaturverzeichnis
- 151 -
season, Galicia). Influence of particulate matter. Estuarine, Coastal and Shelf Science 76, 447-453.
Sañudo-Wilhelmy, S.A., Rivera-Duarte, I., Flegal, A.R., 1996. Distribution of colloidal trace metals in the San Francisco Bay Estuary. Geochimica et Cosmochimica Acta 60, 4933-4944.
Satarug, S., Garrett, S.H., Sens, M.A., Sens, D.A., 2010. Cadmium, environmental exposure, and health outcomes. Environmental Health Perspectives 118, 182-190.
Schlosser, C. and Croot, P.L., 2008. Application of cross-flow filtration for determining the solubility of iron species in open ocean seawater. Limnology and Oceanography: Methods 6, 630-642.
Schüßler, U., Balzer, W., Deeken, A., 2005. Dissolved Al distribution, particulate Al fluxes and coupling to atmospheric Al and dust deposition in the Arabian Sea. Deep Sea Research Part II: Topical Studies in Oceanography 52, 1862-1878.
Schwedt, G., 2008. Analytische Chemie: Grundlagen, Methoden und Praxis. Wiley-VCH Verlag, Weinheim.
Shank, G.C., Skrabal, S.A., Whitehead, R.F., Kieber, R.J., 2004. Strong copper complexation in an organic-rich estuary: the importance of allochthonous dissolved organic matter. Marine Chemistry 88, 21-39.
Sharp, J.H., Culberson, C.H., Church, T.M., 1982. The chemistry of the Delaware Estuary. General considerations. Limnology and Oceanography 27, 1015-1028.
Shiller, A.M. and Boyle, E.A., 1991. Trace elements in the Mississippi River Delta outflow region: Behavior at high discharge. Geochimica et Cosmochimica Acta 55, 3241-3251.
Shim, M.-J., Swarzenski, P.W., Shiller, A.M., 2012. Dissolved and colloidal trace elements in the Mississippi River Delta outflow after Hurricanes Katrina and Rita. Continental Shelf Research 42, 1-9.
Sholkovitz, E.R., 1978. The flocculation of dissolved Fe, Mn, Al, Cu, Ni, Co and Cd during estuarine mixing. Earth and Planetary Science Letters 41, 77-86.
Sholkovitz, E.R., Boyle, E.A., Price, N.B., 1978. The removal of dissolved humic acids and iron during estuarine mixing. Earth and Planetary Science Letters 40, 130-136.
Sigleo, A.C. and Helz, G.R., 1981. Composition of estuarine colloidal material: major and trace elements. Geochimica et Cosmochimica Acta 45, 2501-2509.
Spörl, G., 2011. Impact of sugar cane cultivation on biogeochemistry and phytoplankton dynamics in a tropical lagoon and estuary in Brazil. Dissertation, Universität Bremen.
Stoeppler, M., 1994. Probennahme und Aufschluss. Springer-Verlag, Berlin/Heidelberg.
Stolpe, B. and Hassellöv, M., 2007. Changes in size distribution of fresh water nanoscale colloidal matter and associated elements on mixing with seawater. Geochimica et Cosmochimica Acta 71, 3292-3301.
Stolpe, B., Guo, L.-D., Shiller, A.M., Hassellöv, M., 2010. Size and composition of colloidal organic matter and trace elements in the Mississippi River, Pearl River and the northern Gulf of Mexico, as characterized by flow field-flow fractionation. Marine Chemistry 118, 119-128.
Sulzberger, B., Suter, D., Siffert, C., Banwart, S., Stumm, W., 1989. Dissolution of Fe(III)(hydr)oxides in natural waters; Laboratory assessment on the kinetics controlled by surface coordination. Marine Chemistry 28, 127-144.
Sunda, W.G. and Huntsman, S.A., 1998. Processes regulating cellular metal accumulation and physiological effects: Phytoplankton as model systems. Science of the Total Environment 219, 165-181.
Tang, A.-K., Liu, R.-H., Ling, M., Xu, L.-G., Wang, J.-Y., 2010. Distribution characteristics and controlling factors of soluble heavy metals in the Yellow River Estuary and adjacent sea. Procedia Environmental Sciences 2, 1193-1198.
Literaturverzeichnis
- 152 -
Tang, D.-G., Warnken, K.W., Santschi, P.H., 2001. Organic complexation of copper in surface waters of Galveston Bay. Limnology and Oceanography 46, 321-330.
Tang, D.-G., Warnken K.W., Santschi, P.H., 2002. Distribution and partitioning of trace metals (Cd, Cu, Ni, Pb, Zn) in Galveston Bay waters. Marine Chemistry 78, 29-45.
Takara, H., Aono, T., Tagami, K., Uchida, S., 2010. Processes controlling cobalt distribution in two temperate estuaries, Sagami Bay and Wakasa Bay, Japan. Estuarine, Coastal and Shelf Science 89, 294-305.
Tessier, A., Fortin, D., Belzile, N., De Vitre, R.R., Leppard, G.G., 1996. Metal sorption to diagenetic iron and manganese oxyhydroxides and associated organic matter: Narrowing the gap between field and laboratory measurements. Geochimica et Cosmochimica Acta 60, 387-404.
Tovar-Sánchez, A., Sañudo-Wilhelmy, S.A., Flegal, A.R., 2004. Temporal and spatial variations in the biogeochemical cycling of cobalt in two urban estuaries: Hudson River Estuary and San Francisco Bay. Estuarine, Coastal and Shelf Science 60, 717-728.
Town, R.M. and Filella, M., 2002. Implications of natural organic matter binding heterogeneity on understanding lead(II) complexation in aquatic systems. Science of the Total Environment 300, 143-154.
Turner, A., Millward, G.E., Morris, A.W., 1991. Particulate metals in five major North Sea estuaries. Estuarine, Coastal and Shelf Science32, 325-346.
Turner, A., Millward, G.E., Schuchardt, B., Schirmer, M., Prange, A., 1992. Trace metal distribution coefficients in the Weser Estuary (Germany). Continental Shelf Research 12, 1277-1292.
Turner, A., 1996. Trace-metal partitioning in estuaries: Importance of salinity and particle concentration. Marine Chemistry 54, 27-39.
Turner, A., Nimmo, M., Thuresson, K.A., 1998. Speciation and sorptive behaviour of nickel in an organic-rich estuary (Beaulieu, UK). Marine Chemistry 63, 105-118.
Turner, A., Milward, G.E., Le Roux, S.M., 2004. Significance of oxides and particulate organic matter in controlling trace metal partitioning in a contaminated estuary. Marine Chemistry 88, 179-192.
Turner, A. and Martino, M., 2006. Modelling the equilibrium speciation of nickel in the Tweed Estuary, UK: Voltammetric determinations and simulations using WHAM. Marine Chemistry 102, 198-207.
Turner, A., Le Roux, S.M., Millward, G.E., 2008. Adsorption of cadmium to iron and manganese oxides during estuarine mixing. Marine Chemistry 108, 77-84.
Turner, D.R., Whitfield, M., Dickson, A.G., 1981. The equilibrium speciation of dissolved components in freshwater and sea water at 25°C and 1 atm pressure. Geochimica et Cosmochimica Acta 45, 855-881.
Umweltbericht Niedersachsen, 2013. Schwermetalle im Harzvorland, Umweltbericht Niedersachsen. www.umwelt.niedersachsen.de/umweltbericht.
Uncles, R.J., Stephens, J.A., Smith, R.E., 2002. The dependence of estuarine turbidity on tidal intrusion length, tidal range and residence time. Continental Shelf Research 22, 1835-1856.
Unger, D., Herbeck, L.S., Li, M., Bao, H., Wu, Y., Zhang J., 2013. Sources, transformation and fate of particulate amino acids and hexosamines under varying hydrological regimes in the tropical Wenchang Wenjiao Rivers and Estuary, Hainan, China. Continental Shelf Research 57, 44-58.
Vallee, B.L. and Ulmer, D.D., 1972. Biochemical effects of mercury, cadmium, and lead. Annual Review of Biochemistry 41, 91-128.
Van den Berg, C.M.G. and De Luca Rebello, A., 1986. Organic-copper interactions in Guanabara Bay, Brazil. An electrochemical study of copper complexation by dissolved organic material in a tropical bay. Science of The Total Environment 58, 37-45.
Literaturverzeichnis
- 153 -
Van der Weijden, C.H., Arnoldus, M. J. H. L., Meurs, C.J., 1977. Desorption of metals from suspended material in the Rhine Estuary. Netherlands Journal of Sea Research 11, 130-145.
Von Uexküll, J.J., 1909. Umwelt und Innenwelt der Tiere. Julius Springer, Berlin.
Waeles, M., Riso, R.D., Maguer, J.-F., Le Corre, P., 2004. Distribution and chemical speciation of dissolved cadmium and copper in the Loire Estuary and North Biscay continental shelf, France. Estuarine, Coastal and Shelf Science 59, 49-57.
Waeles, M., Riso, R.D., Le Corre, P., 2005. Seasonal variations of cadmium speciation in the Penzé Estuary, NW France. Estuarine, Coastal and Shelf Science 65, 143-152.
Waeles, M., Riso, R.D., Le Corre, P., 2007. Distribution and seasonal changes of lead in an estuarine system affected by agricultural practices: the Penzé Estuary, NW France. Estuarine, Coastal and Shelf Science 74, 570-578.
Waeles, M., Tanguy, V., Lespes, G., Riso, R.D., 2008a. Behaviour of colloidal trace metals (Cu, Pb and Cd) in estuarine waters: An approach using frontal ultrafiltration (UF) and stripping chronopotentiometric methods (SCP). Estuarine, Coastal and Shelf Science 80, 538-544.
Waeles, M., Riso, R.D., Maguer, J.-F., Guillaud, J.-F., Le Corre, P., 2008b. On the distribution of dissolved lead in the Loire Estuary and the North Biscay continental shelf, France. Journal of Marine Systems 72, 358-365.
Wang, C.-Y., Wang, X.-L., Wang, B.-D., Zhang, X.-Y., Zhu, C.-J., 2009. Level and fate of heavy metals in the Changjiang Estuary and its adjacent waters. Oceanology 49, 64-72.
Wang, W.-S., Wen, B., Zhang, S.-Z., Shan, X.-Q., 2003. Distribution of heavy metals in water and soil solutions based on colloid-size fractionation. International Journal of Environmental Analytical Chemistry 83, 357-365.
Wang, Z.-L. and Liu, C.-Q., 2003. Distribution and partition behavior of heavy metals between dissolved and acid-soluble fractions along a salinity gradient in the Changjiang Estuary, eastern China. Chemical Geology 202, 383-396.
Walters, R.A., Cheng, R.-T., Conomos, T.J., 1985. Time scales of circulation and mixing processes of San Francisco Bay waters. Hydrobiologia 129, 13-36.
Warren, L.A. and Zimmerman, A.P., 2003. The influence of temperature and NaCl on cadmium, copper and zinc partitioning among suspended particulate and dissolved phases in an urban river. Water Research 28, 1921-1931.
Wells, M.L., Kozelka, P.B., Bruland, K.W., 1998. The complexation of ‚dissolved’ Cu, Zn, Cd and Pb by soluble and colloidal organic matter in Narragansett Bay, RI. Marine Chemistry 62, 203-217.
Wen, L.-S., Stordal, M.C., Tang, D.-G., Gill, G.A., Santschi, P.H., 1996. An ultraclean cross-flow ultrafiltration technique for the study of trace metal phase speciation in seawater. Marine Chemistry 55, 129-152.
Wen, L.-S., Santschi, P., Gill, G., Paternostro, C., 1999. Estuarine trace metal distributions in Galveston Bay: importance of colloidal forms in the speciation of the dissolved phase. Marine Chemistry 63, 185-212.
Wen, L.-S., Warnken, K.W., Santschi, P.H., 2008. The role of organic carbon, iron, and aluminium oxyhydroxides as trace metal carriers: comparison between the Trinity River and the Trinity River Estuary (Galveston Bay, Texas). Marine Chemistry 112, 20-37.
Wen, L.-S., Santschi, P., Gill, G., Warnken, K.W:, Davison, W., Zhang, H., Li, H.-P., Jiann, K.-T., 2011. Molecular weight and chemical reactivity of dissolved trace metals (Cd, Cu, Ni) in surface waters from the Mississippi River to Gulf of Mexico. Estuarine, Coastal and Shelf Science 92, 649-658.
Whitehouse, B.G., Yeats, P.A., Strain, P.M., 1990. Cross-flow filtration of colloids from aquatic environments. Limnology and Oceanography 35, 1368-1375.
Wiberg, N., 2007. Lehrbuch der Anorganischen Chemie, 102. Auflage. Walter de Gruyter,
Literaturverzeichnis
- 154 -
Berlin.
Willard, H.H., Merritt Jr., L.L., Dean, J.A., Settle Jr., F.A., 1988. Instrumental Methods of Analysis, Seventh Edition. Wadsworth, California.
Windom, H.L., Niencheski, L.F., Smith Jr., R.G., 1999. Biogeochemistry of nutrients and trace metals in the estuarine region of the Patos Lagoon (Brazil). Estuarine, Coastal and Shelf Science 48, 113-123.
Wolf, L., Schwalger, B., Knoppers, B.A., Da Silva, L.A.F., Medeiros, P.R.P., Pollehne, F., 2010. Distribution of prokaryotic organisms in a tropical estuary influenced by sugar cane agriculture in northeast Brazil. Brazilian Journal of Microbiology, 41, 890-898.
Wurtz, M. and Maeder, M., 2002. Einführung in die Umwelttoxikologie. Verlag Natur & Wissenschaft, Solingen.
Yang, M. and Sañudo-Wilhelmy, S.A., 1998. Cadmium and manganese distributions in the Hudson River Estuary: Interannual and seasonal variability. Earth and Planetary Science Letters 160, 403-418.
Zamuda, C.D. and Sunda, W.G., 1982. Bioavailability of dissolved copper to the American oyster Crassostrea virginica. I. Importance of chemical speciation. Marine Biology 66, 77-82.
Zeng, Z.-X. and Zeng, X.-Z., 1989. Physicogeography of the Hainan Island. Science Press, Beijing. (In Chinese)
Zhang, H., Van den Berg, C.M.G., Wollast, R., 1990. The determination of interactions of cobalt (II) with organic compounds in seawater using cathodic stripping voltammetry. Marine Chemistry 28, 285-300.
Zhang, J., 1999. Heavy metal compositions of suspended sediments in the Changjiang (Yangtze River) Estuary: significance of riverine transport to the ocean. Continental Shelf Research 19, 1521-1543.
Zhang, J. and Liu, C.-L., 2002. Riverine composition and estuarine geochemistry of particulate metals in China — weathering features, anthropogenic impact and chemical fluxes. Estuarine, Coastal and Shelf Science 54, 1051-1070.
Zhou, J.-L., Liu, Y.-P., Abrahams, P.W., 2003. Trace metal behaviour in the Conwy Estuary, North Wales. Chemosphere 51, 429-440.
Zhu, D.-K., Yin, Y., Martini, I.P., 2005. Geomorphology of the Boao coastal system and potential effects of human activities — Hainan Island, South China. Journal of Geographical Sciences 15, 187-198.
Zlotorzynski, A., 1995. The application of microwave radiation to analytical and environmental chemistry. Critical Reviews in Analytical Chemistry 25, 43-76.
Zwolsman, J.J.G., van Eck, B.T.M., van der Weijden, C.H., 1997. Geochemistry of dissolved trace metals (cadmium, copper, zinc) in the Scheldt Estuary, southwestern Netherlands: impact of seasonal variability. Geochimica et Cosmochimica Acta 61, 1635–1652.
Anhang
- 155 -
Anha
ng
Anha
ng 1
. U
mw
eltp
aram
eter
, di
e ge
löst
e un
d di
e pa
rtik
ulär
e M
etal
lkon
zent
ratio
nen
sow
ie d
ie L
ogK
D W
erte
in
der
Troc
kens
aiso
n 20
06 i
n O
st-
Hai
nan
(n.d
.: ni
cht d
etek
tierb
ar).
Um
wel
tpar
amet
er
gelö
ste
Met
alle
pa
rtik
ulär
e M
etal
le
LogK
D
Stat
ion
Salin
ität
SPM
D
OC
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
psu
mg/
L m
g/L
μmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
mm
ol/g
μm
ol/g
μm
ol/g
μm
ol/g
nm
ol/g
μm
ol/g
L/
kg
L/kg
L/
kg
L/kg
L/
kg
L/kg
WR
A 7.
0 7.
8 1.
09
0.05
3.
99
16.4
3 7.
11
0.13
0.
16
1.03
0.
52
0.41
0.
30
3.32
0.
25
7.3
5.1
4.4
4.6
4.5
6.2
B 2.
1 10
.9
2.16
0.
11
7.78
1.
30
8.45
0.
09
0.41
1.
00
0.41
0.
23
0.27
4.
02
0.19
7.
0 4.
7 5.
3 4.
5 4.
7 5.
7 C
0.
0 12
.2
1.37
0.
17
2.39
0.
23
8.99
0.
01
0.13
1.
13
0.53
0.
28
0.32
5.
29
0.25
6.
8 5.
3 6.
1 4.
6 5.
9 6.
3 D
0.
0 10
.7
1.51
0.
19
2.91
0.
22
5.64
0.
01
0.06
1.
22
0.60
0.
29
0.35
5.
39
0.26
6.
8 5.
3 6.
1 4.
8 5.
7 6.
6 E
0.0
7.9
1.51
0.
34
4.31
0.
36
35.1
8 0.
03
0.54
1.
15
0.59
0.
33
0.33
5.
20
0.26
6.
5 5.
1 6.
0 4.
0 5.
3 5.
7 F
16.1
15
.8
0.97
0.
01
7.63
0.
55
40.1
6 0.
10
0.34
0.
61
0.44
0.
19
0.26
1.
83
0.14
7.
7 4.
8 5.
5 3.
8 4.
3 5.
6
WW
R
G
31
.9
9.5
-- 0.
04
6.68
0.
67
10.9
3 0.
14
0.11
0.
56
0.57
0.
21
0.38
1.
46
0.16
7.
2 4.
9 5.
5 4.
5 4.
0 6.
1 H
30
.2
8.8
-- 0.
08
6.01
1.
32
6.04
0.
14
0.05
0.
51
0.49
0.
18
0.28
1.
57
0.10
6.
8 4.
9 5.
1 4.
7 4.
1 6.
3 J
29.8
11
.5
0.96
0.
01
5.71
0.
71
10.9
6 0.
14
0.11
0.
68
0.66
0.
28
0.35
1.
74
0.13
8.
0 5.
1 5.
6 4.
5 4.
1 6.
1 I
29.7
11
.1
0.95
0.
02
5.09
1.
22
5.68
0.
12
0.04
0.
69
0.66
0.
25
0.37
2.
13
0.12
7.
6 5.
1 5.
3 4.
8 4.
3 6.
5 K
18.8
9.
9 1.
93
0.06
9.
97
4.92
6.
35
0.11
0.
02
0.74
0.
57
0.27
0.
32
2.89
0.
09
7.1
4.8
4.7
4.7
4.5
6.7
L 10
.7
4.8
2.41
1.
50
15.0
1 7.
23
10.4
4 0.
10
0.26
1.
28
1.04
0.
41
0.63
4.
46
0.11
5.
9 4.
8 4.
8 4.
8 4.
7 5.
6 N
2.
3 7.
9 2.
21
0.22
11
.25
8.36
9.
22
0.08
0.
08
1.99
1.
55
0.33
0.
68
2.63
0.
13
7.0
5.1
4.6
4.9
4.6
6.2
O
0.0
7.2
-- 0.
16
10.5
7 4.
69
8.57
0.
13
0.02
2.
22
1.81
0.
49
0.80
9.
28
0.18
7.
1 5.
2 5.
0 5.
0 4.
9 6.
9 P
0.0
11.2
3.
00
0.71
5.
61
0.43
7.
83
0.01
0.
10
1.44
0.
79
0.20
0.
35
2.85
0.
19
6.3
5.2
5.7
4.6
5.8
6.3
Anhang
- 156 -
Anha
ng 2
. Um
wel
tpar
amet
er, d
ie g
elös
te u
nd d
ie p
artik
ulär
e M
etal
lkon
zent
ratio
nen
sow
ie d
ie L
og K
D W
erte
in d
er R
egen
sais
on 2
007
in O
st-H
aina
n (n
.d.:
nich
t det
ektie
rbar
).
U
mw
eltp
aram
eter
ge
löst
e M
etal
le
part
ikul
äre
Met
alle
Lo
gKD
Stat
ion
Salin
ität
SPM
D
OC
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
psu
mg/
L m
g/L
μmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
mm
ol/g
μm
ol/g
μm
ol/g
μm
ol/g
nm
ol/g
μm
ol/g
L/
kg
L/kg
L/
kg
L/kg
L/
kg
L/kg
WR
BB1
32.0
2.
2 0.
52
0.28
3.
81
0.41
2.
27
0.06
0.
15
0.57
0.
45
0.22
0.
25
3.18
0.
17
6.3
5.1
5.7
5.0
4.8
6.1
BB2
2.4
1.6
1.06
2.
59
4.65
1.
01
7.74
0.
05
1.09
1.
18
0.65
0.
28
0.38
2.
41
0.27
5.
7 5.
1 5.
4 4.
7 4.
7 5.
4 BB
3 0.
0 2.
0 2.
02
4.03
7.
17
1.51
10
.20
0.05
1.
50
1.16
0.
85
0.43
0.
44
4.15
0.
26
5.5
5.1
5.4
4.6
4.9
5.2
BB4
0.0
1.8
1.53
4.
31
6.02
1.
76
9.15
0.
08
1.21
1.
13
0.76
0.
45
0.47
5.
12
0.25
5.
4 5.
1 5.
4 4.
7 4.
9 5.
3 BB
5 0.
6 2.
3 2.
00
3.99
6.
73
1.86
7.
78
0.06
1.
16
0.95
0.
46
0.27
0.
38
3.02
0.
20
5.4
4.8
5.2
4.7
4.8
5.2
BB6
7.8
2.0
2.99
2.
46
5.64
1.
26
8.39
0.
06
1.09
0.
90
0.51
0.
31
0.32
2.
19
0.22
5.
6 5.
0 5.
4 4.
6 4.
6 5.
3 BB
7 4.
6 2.
4 2.
23
2.53
4.
85
1.20
6.
52
0.05
0.
86
1.08
0.
72
0.35
0.
41
2.29
0.
25
5.6
5.2
5.5
4.8
4.7
5.5
BB8
14.6
1.
8 1.
19
2.07
4.
84
0.96
6.
68
0.07
0.
91
0.76
0.
49
0.24
1.
04
1.46
0.
21
5.6
5.0
5.4
5.2
4.4
5.4
BB9
26.5
2.
7 1.
07
1.20
4.
64
0.55
9.
33
0.07
0.
64
0.56
0.
41
0.17
0.
69
3.14
0.
16
5.7
4.9
5.5
4.9
4.7
5.4
BB10
13
.1
2.0
1.55
2.
48
5.70
1.
14
7.50
0.
05
0.78
0.
86
0.54
0.
25
0.29
1.
74
0.17
5.
5 5.
0 5.
3 4.
6 4.
6 5.
3 BB
11
6.2
1.0
2.28
2.
69
7.91
1.
34
6.95
0.
04
0.76
0.
90
0.54
0.
25
0.38
1.
69
0.19
5.
5 4.
8 5.
3 4.
7 4.
7 5.
4
WW
R
W
W1
0.0
1.2
4.22
5.
49
6.08
2.
54
7.52
0.
02
0.88
1.
64
1.03
0.
43
0.31
2.
02
0.27
5.
5 5.
2 5.
2 4.
6 5.
1 5.
5 W
W2
3.2
2.3
3.67
3.
53
5.79
1.
72
6.31
0.
03
0.53
1.
60
1.04
0.
65
0.55
3.
21
0.22
5.
7 5.
3 5.
6 4.
9 5.
0 5.
6 W
W3
18.2
2.
3 1.
40
0.80
6.
75
2.18
9.
21
0.07
0.
24
0.51
0.
53
0.19
0.
33
5.05
0.
17
5.8
4.9
4.9
4.6
4.9
5.9
WW
4 23
.7
1.4
-- 0.
84
5.36
1.
87
6.89
0.
07
0.30
0.
71
0.70
0.
26
0.37
5.
17
0.30
5.
9 5.
1 5.
1 4.
7 4.
9 6.
0 W
W5
22.6
2.
2 --
0.34
4.
93
2.13
6.
65
0.07
0.
19
0.40
0.
45
0.17
0.
22
3.26
0.
25
6.1
5.0
4.9
4.5
4.7
6.1
WW
6 0.
0 4.
6 2.
69
1.82
17
.77
2.24
9.
03
0.03
0.
15
2.11
1.
47
1.40
0.
73
4.44
0.
19
6.1
4.9
5.8
4.9
5.2
6.1
WW
7 9.
5 2.
9 2.
37
2.77
12
.31
6.09
7.
41
0.05
0.
38
1.45
3.
93
0.44
0.
67
2.01
0.
18
5.7
5.5
4.9
5.0
4.6
5.7
WW
8 22
.2
2.6
1.41
0.
25
6.50
2.
20
6.13
0.
08
0.08
0.
34
0.33
0.
13
0.16
2.
34
0.09
6.
1 4.
7 4.
8 4.
4 4.
5 6.
1 W
W9
22.3
2.
8 1.
01
0.29
7.
46
2.48
7.
42
0.12
0.
08
0.23
0.
25
0.08
0.
13
3.02
0.
07
5.9
4.5
4.5
4.2
4.4
5.9
WW
10
20.5
4.
9 1.
38
0.23
7.
06
2.04
6.
95
0.10
0.
04
0.20
0.
25
0.09
0.
13
3.84
0.
06
5.9
4.5
4.7
4.3
4.6
6.2
WW
11
28.5
2.
2 1.
12
0.16
4.
60
0.91
6.
21
0.09
0.
06
0.26
0.
32
0.11
0.
26
4.16
0.
09
6.2
4.8
5.1
4.6
4.7
6.2
Anhang
- 157 -
Anha
ng 3
. Um
wel
tpar
amet
er, d
ie g
elös
te u
nd d
ie p
artik
ulär
e M
etal
lkon
zent
ratio
nen
sow
ie d
ie L
og K
D W
erte
in d
er R
egen
sais
on 2
008
in O
st-H
aina
n (n
.d.:
nich
t det
ektie
rbar
).
U
mw
eltp
aram
eter
ge
löst
e M
etal
le
part
ikul
äre
Met
alle
Lo
gKD
Stat
ion
Salin
ität
SPM
D
OC
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
psu
mg/
L m
g/L
μmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
mm
ol/g
μm
ol/g
μm
ol/g
μm
ol/g
nm
ol/g
μm
ol/g
L/
kg
L/kg
L/
kg
L/kg
L/
kg
L/kg
WR
A1
0.0
3.2
3.58
3.
47
1.06
1.
35
4.37
0.
03
0.72
0.
75
0.36
0.
31
0.51
6.
63
0.22
5.
3 5.
5 5.
4 5.
1 5.
3 5.
5 A4
0.
0 14
.1
2.15
3.
03
3.78
1.
30
6.68
0.
03
1.07
1.
54
1.04
0.
62
1.20
9.
00
0.55
5.
7 5.
4 5.
7 5.
3 5.
6 5.
7 B1
4.
6 1.
4 2.
74
6.26
9.
21
1.82
11
.95
0.08
1.
58
0.92
0.
88
0.30
0.
83
3.11
0.
28
5.2
5.0
5.2
4.8
4.6
5.3
B2
33.6
2.
7 1.
68
0.16
4.
18
0.09
1.
71
0.10
0.
01
0.42
0.
37
0.15
0.
50
2.81
0.
16
6.4
4.9
6.2
5.5
4.5
7.2
B3
16.5
0.
6 1.
64
0.85
5.
81
0.49
5.
25
0.09
0.
46
0.98
0.
76
0.19
0.
66
2.29
0.
26
6.1
5.1
5.6
5.1
4.4
5.8
B4
7.7
0.8
1.71
1.
10
3.86
0.
77
7.39
0.
06
0.25
1.
07
0.86
0.
33
0.72
3.
28
0.30
6.
0 5.
3 5.
6 5.
0 4.
8 6.
1 B5
22
.5
1.5
0.79
0.
19
2.40
0.
33
3.99
0.
10
0.08
0.
75
0.54
0.
18
0.54
2.
91
0.24
6.
6 5.
4 5.
7 5.
1 4.
5 6.
5 B6
11
.0
0.8
2.85
2.
46
5.08
1.
10
6.52
0.
08
0.62
0.
98
0.77
0.
23
0.65
2.
76
0.25
5.
6 5.
2 5.
3 5.
0 4.
6 5.
6 B8
0.
0 3.
5 1.
88
4.67
3.
63
2.19
8.
65
0.03
1.
32
0.86
0.
79
0.37
0.
80
5.43
0.
25
5.3
5.3
5.2
5.0
5.3
5.3
B9
0.0
1.0
2.47
3.
74
2.63
1.
48
6.90
0.
03
0.97
0.
90
0.73
0.
30
0.72
5.
90
0.27
5.
4 5.
4 5.
3 5.
0 5.
3 5.
4
WW
R
W
1 28
.3
0.4
3.37
0.
35
2.76
0.
67
7.56
0.
09
0.25
0.
58
1.34
0.
21
0.89
3.
82
0.22
6.
2 5.
7 5.
5 5.
1 4.
7 5.
9 W
2s
20.0
0.
9 6.
85
0.71
9.
30
1.28
12
.04
0.11
0.
36
0.62
0.
82
0.27
0.
99
4.29
0.
22
5.9
4.9
5.3
4.9
4.6
5.8
W3s
18
.3
1.8
4.09
0.
95
13.0
7 1.
44
12.2
1 0.
07
0.31
0.
43
0.63
0.
22
0.79
3.
83
0.12
5.
7 4.
7 5.
2 4.
8 4.
8 5.
6 W
4s
15.5
1.
4 4.
18
0.80
6.
71
1.43
12
.65
0.06
0.
28
0.46
0.
72
0.26
0.
89
4.72
0.
15
5.8
5.0
5.3
4.8
4.9
5.7
W5
0.0
1.2
2.30
7.
26
10.9
6 4.
08
12.9
9 0.
02
0.72
1.
24
1.65
0.
61
1.28
4.
95
0.14
5.
2 5.
2 5.
2 5.
0 5.
4 5.
3 W
6 3.
4 1.
3 5.
56
5.56
15
.32
4.59
11
.48
0.04
0.
94
1.04
1.
38
0.50
1.
11
3.33
0.
18
5.3
5.0
5.0
5.0
4.9
5.3
W7
5.9
1.4
6.68
4.
89
13.1
6 3.
75
13.9
4 0.
05
0.95
0.
92
1.37
0.
61
1.18
3.
58
0.19
5.
3 5.
0 5.
2 4.
9 4.
9 5.
3 W
8 9.
7 0.
9 5.
34
2.31
9.
47
2.36
11
.51
0.11
1.
08
0.67
0.
88
0.31
0.
74
2.49
0.
15
5.5
5.0
5.1
4.8
4.4
5.1
W9
10.5
0.
9 6.
37
1.80
9.
48
1.81
12
.25
0.07
0.
49
0.41
0.
60
0.17
0.
53
2.87
0.
13
5.4
4.8
5.0
4.6
4.7
5.4
W10
12
.5
0.9
4.91
1.
81
7.48
1.
99
11.2
6 0.
08
0.46
0.
56
0.80
0.
38
0.84
3.
98
0.16
5.
5 5.
0 5.
3 4.
9 4.
8 5.
5 W
11
0.0
4.2
7.85
9.
71
5.58
5.
04
11.2
0 0.
04
1.53
0.
64
0.85
0.
34
0.69
3.
20
0.17
4.
8 5.
2 4.
8 4.
8 4.
9 5.
0 W
12
0.0
1.7
7.79
10
.81
5.18
3.
81
10.5
1 0.
02
2.04
0.
67
0.68
0.
28
0.55
2.
06
0.17
4.
8 5.
1 4.
9 4.
7 5.
1 4.
9 W
13
0.0
1.9
6.17
10
.64
5.43
3.
82
10.7
2 0.
02
1.92
0.
74
0.80
0.
32
0.73
2.
52
0.19
4.
8 5.
2 4.
9 4.
8 5.
3 5.
0 W
14
0.2
1.4
6.59
13
.72
6.56
4.
62
11.6
8 0.
03
2.24
0.
98
0.74
0.
29
0.63
2.
50
0.20
4.
9 5.
1 4.
8 4.
7 5.
0 4.
9 W
15
7.6
0.8
5.89
7.
25
11.3
2 3.
79
10.6
4 0.
08
1.35
0.
94
0.85
0.
24
1.07
3.
90
0.22
5.
1 4.
9 4.
8 5.
0 4.
7 5.
2
Anhang
- 158 -
Anha
ng 4
. Um
wel
tpar
amet
er, d
ie g
elös
te u
nd d
ie p
artik
ulär
e M
etal
lkon
zent
ratio
nen
sow
ie d
ie L
og K
D W
erte
in d
er R
egen
sais
on 2
007
in N
ordw
est-
Bra
silie
n (n
.d.:
nich
t det
ektie
rbar
).
U
mw
eltp
aram
eter
ge
löst
e M
etal
le
part
ikul
äre
Met
alle
Lo
gKD
Stat
ion
Salin
ität
SPM
D
OC
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
psu
mg/
L m
g/L
μmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
mm
ol/g
μm
ol/g
μm
ol/g
μm
ol/g
nm
ol/g
μm
ol/g
L/
kg
L/kg
L/
kg
L/kg
L/
kg
L/kg
MD
285
0.0
5.3
0.73
5.
73
3.29
0.
79
3.72
0.
09
0.37
1.
58
0.23
0.
22
0.23
3.
17
0.34
5.
4 4.
8 5.
4 4.
8 4.
5 6.
0 28
7 0.
0 9.
1 2.
55
1.36
4.
85
0.34
7.
61
0.05
0.
16
1.36
0.
24
0.19
0.
22
7.27
0.
18
6.0
4.7
5.8
4.5
5.1
6.1
288
2.9
12.3
4.
20
0.11
9.
19
1.90
14
.00
0.06
0.
04
0.83
0.
34
0.21
0.
31
1.43
0.
15
6.9
4.6
5.0
4.4
4.4
6.6
289
4.9
75.5
3.
88
0.24
5.
89
1.36
10
.86
0.01
0.
06
0.62
0.
29
0.13
0.
19
1.27
0.
11
6.4
4.7
5.0
4.3
5.1
6.3
290
36.2
16
.2
3.77
0.
40
3.50
0.
24
2.12
0.
03
0.14
0.
41
0.23
0.
07
0.13
0.
41
0.08
6.
0 4.
8 5.
5 4.
8 4.
2 5.
8 29
1 20
.8
6.2
--
0.04
6.
99
1.23
6.
87
0.03
0.
18
0.57
0.
32
0.11
0.
22
0.43
0.
12
7.1
4.7
5.0
4.5
4.2
5.8
MG
264
19.4
22
.7
2.37
1.
22
4.77
0.
95
6.45
0.
02
0.47
0.
56
0.27
0.
12
0.19
0.
43
0.11
5.
7 4.
8 5.
1 4.
5 4.
3 5.
4 26
5 10
.0
11.3
3.
46
2.06
3.
55
1.03
5.
66
0.02
0.
53
0.68
0.
30
0.12
0.
20
0.54
0.
14
5.5
4.9
5.1
4.5
4.5
5.4
266
8.6
17.4
4.
55
1.41
6.
80
1.18
13
.12
0.01
0.
66
0.55
0.
26
0.11
0.
17
0.52
0.
11
5.6
4.6
5.0
4.1
4.6
5.2
267
4.8
26.8
4.
78
1.03
6.
34
0.93
11
.87
0.01
0.
40
0.60
0.
31
0.13
0.
20
0.67
0.
14
5.8
4.7
5.1
4.2
4.8
5.5
268
3.0
-- 2.
89
4.84
6.
93
3.09
10
.47
0.02
1.
04
0.62
0.
28
0.12
0.
23
0.63
0.
12
5.1
4.6
4.6
4.3
4.5
5.1
270
1.1
27.6
4.
41
2.34
4.
37
1.21
7.
37
0.01
0.
54
0.58
0.
26
0.13
0.
18
0.47
0.
11
5.4
4.8
5.0
4.4
4.5
5.3
271
0.0
32.7
--
0.
32
11.0
6 0.
68
20.6
1 0.
01
0.12
0.
57
0.31
0.
12
0.20
0.
52
0.13
6.
3 4.
4 5.
2 4.
0 4.
8 6.
0 27
2 0.
0 --
3.63
0.
46
8.93
1.
48
10.0
2 0.
01
0.17
1.
84
0.32
0.
29
0.26
0.
97
0.18
6.
6 4.
6 5.
3 4.
4 5.
1 6.
0 27
3 0.
0 30
.0
4.00
0.
85
8.07
0.
26
11.8
8 0.
01
0.18
1.
08
0.33
0.
23
0.26
0.
91
0.17
6.
1 4.
6 6.
0 4.
3 5.
2 6.
0
Sum
aúm
a
405
0.0
12.1
8.
22
21.5
0 9.
50
3.80
13
.86
0.02
2.
02
0.76
0.
20
0.11
0.
24
1.50
0.
32
4.5
4.3
4.4
4.2
4.9
5.2
Mar
ine
33
5 31
.7
2.9
2.88
0.
01
2.83
0.
41
3.25
0.
02
0.11
0.
54
0.29
0.
09
0.19
1.
00
0.17
7.
8 5.
0 5.
4 4.
8 4.
7 6.
2 33
6 29
.2
4.3
1.17
0.
06
3.50
0.
69
4.13
0.
02
0.02
0.
46
0.27
0.
09
0.16
0.
53
0.14
6.
9 4.
9 5.
1 4.
6 4.
4 6.
8
Anhang
- 159 -
Anha
ng 5
. U
mw
eltp
aram
eter
, di
e ge
löst
e un
d di
e pa
rtik
ulär
e M
etal
lkon
zent
ratio
nen
sow
ie d
ie L
og K
D W
erte
in
der
Troc
kens
aiso
n 20
08 i
n N
ordw
est-B
rasi
lien
(n.d
.: ni
cht d
etek
tierb
ar).
ge
löst
e M
etal
le
part
ikul
äre
Met
alle
Lo
gKD
Stat
ion
Salin
ität
SPM
D
OC
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
psu
mg/
L m
g/L
μmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
mm
ol/g
μm
ol/g
μm
ol/g
μm
ol/g
nm
ol/g
μm
ol/g
L/
kg
L/kg
L/
kg
L/kg
L/
kg
L/kg
MD
590
36.2
2.
8 3.
61
0.36
2.
54
0.37
1.
83
0.06
0.
12
0.13
0.
06
0.01
0.
05
0.53
0.
04
5.5
4.4
4.6
4.4
4.0
5.5
591
35.5
0.
8 2.
39
0.23
1.
36
0.48
1.
42
0.04
0.
08
0.45
0.
09
0.12
0.
09
1.12
0.
09
6.3
4.8
5.4
4.8
4.4
6.0
592
26.3
1.
0 4.
07
0.82
3.
53
2.06
3.
68
0.06
0.
39
0.56
0.
25
0.20
0.
21
1.06
0.
13
5.8
4.8
5.0
4.7
4.3
5.5
593
18.6
10
.8
2.63
0.
05
0.08
1.
58
6.30
0.
00
0.33
0.
53
0.33
0.
17
0.22
0.
55
0.11
7.
0 6.
6 5.
0 4.
5 5.
4 5.
5 59
8 16
.8
5.0
4.01
0.
08
n.d.
2.
27
5.07
0.
00
n.d.
0.
32
0.31
0.
17
0.15
0.
55
0.08
6.
6 --
4.9
4.5
5.6
-- 60
1 10
.6
11.2
3.
54
0.14
3.
26
2.73
5.
35
0.01
0.
17
0.91
0.
30
0.21
0.
24
0.65
0.
13
6.8
5.0
4.9
4.7
4.8
5.9
602
1.2
3.8
10.8
3 4.
08
2.77
3.
44
8.74
0.
00
0.27
1.
24
0.31
0.
18
0.30
0.
52
0.17
5.
5 5.
1 4.
7 4.
5 5.
2 5.
8 60
3 4.
3 5.
3 4.
45
2.81
2.
25
3.60
5.
51
0.01
0.
24
1.07
0.
27
0.17
0.
27
0.51
0.
14
5.6
5.1
4.7
4.7
5.0
5.8
680
0.3
18.1
2.
24
0.13
0.
73
1.83
7.
80
0.00
0.
01
1.28
0.
37
0.20
0.
36
0.94
0.
17
7.0
5.7
5.0
4.7
5.6
7.5
692
0.0
-- 4.
88
4.37
3.
85
2.09
9.
92
0.02
0.
67
-- --
-- --
-- --
-- --
-- --
-- --
MG
623
22.2
20
.1
6.19
0.
42
3.07
2.
27
5.06
0.
05
0.16
0.
60
0.27
0.
12
0.22
0.
50
0.13
6.
2 4.
9 4.
7 4.
6 4.
0 5.
9 62
5 24
.0
18.2
3.
44
0.03
2.
84
1.95
5.
06
0.05
0.
13
0.48
0.
27
0.12
0.
18
0.54
0.
14
7.2
5.0
4.8
4.6
4.1
6.0
626
31.0
1.
2 3.
24
0.96
3.
91
1.32
3.
87
0.07
0.
29
0.40
0.
20
0.09
0.
14
0.93
0.
10
5.6
4.7
4.8
4.6
4.1
5.5
627
35.7
--
1.67
0.
73
1.30
0.
49
2.65
0.
07
0.16
0.
32
0.18
0.
07
0.11
0.
48
0.08
5.
6 5.
1 5.
1 4.
6 3.
9 5.
7 62
8 31
.7
4.9
2.81
0.
06
1.28
0.
91
2.14
0.
04
0.06
0.
31
0.22
0.
06
0.11
0.
28
0.08
6.
7 5.
2 4.
8 4.
7 3.
9 6.
1 63
0 13
.8
4.1
6.63
0.
43
2.99
2.
33
4.96
0.
06
0.14
0.
42
0.22
0.
10
0.17
0.
92
0.11
6.
0 4.
9 4.
6 4.
5 4.
2 5.
9 63
3 0.
0 10
.1
2.47
6.
03
2.66
1.
60
3.99
n.
d.
0.21
1.
52
0.16
0.
08
0.20
0.
70
0.26
5.
4 4.
8 4.
7 4.
7 --
6.1
637
6.1
7.5
7.33
0.
99
1.17
2.
44
6.36
0.
01
0.31
0.
08
0.05
0.
04
0.05
0.
29
0.02
4.
9 4.
6 4.
2 3.
9 4.
4 4.
8 63
9 2.
6 5.
2 5.
01
1.82
7.
10
2.75
3.
88
0.04
0.
20
1.12
0.
20
0.20
0.
34
0.93
0.
08
5.8
4.5
4.9
4.9
4.3
5.6
641
4.5
10.0
5.
46
1.46
5.
47
3.27
11
.17
0.01
0.
30
0.48
0.
30
0.11
0.
23
0.49
0.
11
5.5
4.7
4.5
4.3
4.5
5.6
691
0.0
-- 4.
88
0.96
8.
39
2.38
20
.52
0.02
0.
32
-- --
-- --
-- --
-- --
-- --
-- --
Sum
aúm
a
542
0.0
11.3
4.
04
0.87
0.
64
0.96
6.
94
0.00
0.
03
0.88
0.
16
0.15
0.
23
1.46
0.
25
6.0
5.4
5.2
4.5
5.9
7.0
690
0.0
7.9
2.79
0.
12
4.24
1.
08
7.27
0.
01
0.04
--
0.36
0.
25
0.42
2.
43
0.25
--
4.9
5.4
4.8
5.4
6.8
Mar
ine
66
5 37
.1
2.0
1.16
0.
01
1.71
0.
18
2.10
0.
08
0.07
0.
26
0.18
0.
04
0.11
1.
32
0.07
7.
3 5.
0 5.
3 4.
7 4.
2 5.
9 66
6 35
.4
2.3
4.41
0.
02
2.37
0.
42
2.09
0.
09
0.07
0.
29
0.22
0.
05
0.12
0.
67
0.09
7.
3 5.
0 5.
0 4.
8 3.
9 6.
1 66
9 37
.3
2.0
1.58
0.
56
2.90
0.
29
4.07
0.
10
0.34
0.
37
0.23
0.
04
0.32
0.
98
0.13
5.
8 4.
9 5.
2 4.
9 4.
0 5.
6
Anhang
- 160 -
Anha
ng 6
. U
mw
eltp
aram
eter
, di
e ge
löst
e un
d di
e pa
rtik
ulär
e M
etal
lkon
zent
ratio
nen
sow
ie d
ie L
og K
D W
erte
in
der
Troc
kens
aiso
n 20
09 i
n N
ordw
est-B
rasi
lien
(n.d
.: ni
cht d
etek
tierb
ar).
ge
löst
e M
etal
le
part
ikul
äre
Met
alle
Lo
gKD
Stat
ion
Salin
ität
SPM
D
OC
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
psu
mg/
L m
g/L
μmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
mm
ol/g
μm
ol/g
μm
ol/g
μm
ol/g
nm
ol/g
μm
ol/g
L/
kg
L/kg
L/
kg
L/kg
L/
kg
L/kg
MD
940
27.5
7.
5 2.
69
0.11
5.
42
1.03
4.
60
0.08
0.
25
0.51
0.
25
0.07
0.
20
0.57
0.
12
6.6
4.7
4.8
4.6
3.9
5.7
941
24.5
9.
5 2.
84
0.16
5.
49
1.40
5.
65
0.04
0.
17
0.51
0.
28
0.09
0.
22
0.52
0.
13
6.5
4.7
4.8
4.6
4.1
5.9
942
20.6
9.
6 3.
30
0.15
6.
17
1.65
5.
98
0.03
0.
12
0.47
0.
30
0.09
0.
22
0.51
0.
14
6.5
4.7
4.7
4.6
4.3
6.1
943
14.2
6.
8 7.
72
0.05
5.
56
1.58
6.
61
0.02
0.
10
0.42
0.
27
0.12
0.
18
0.75
0.
12
6.9
4.7
4.9
4.4
4.7
6.1
944
11.8
3.
7 5.
97
1.54
9.
25
1.78
7.
16
0.04
0.
77
0.39
0.
31
0.15
0.
19
0.71
0.
11
5.4
4.5
4.9
4.4
4.3
5.2
946
10.0
3.
7 5.
62
4.18
7.
93
3.03
9.
05
0.04
1.
20
0.43
0.
28
0.19
0.
20
1.10
0.
12
5.0
4.5
4.8
4.3
4.4
5.0
948
5.6
7.1
5.29
0.
16
7.46
2.
60
10.2
5 0.
02
0.08
0.
58
0.31
0.
16
0.20
1.
16
0.13
6.
6 4.
6 4.
8 4.
3 4.
7 6.
2 95
0 3.
0 13
.9
5.82
0.
53
8.66
2.
51
10.1
9 0.
02
0.09
0.
96
0.22
0.
19
0.19
1.
30
0.11
6.
3 4.
4 4.
9 4.
3 4.
8 6.
1 95
1 0.
1 19
.9
9.02
3.
58
13.4
9 2.
36
9.45
0.
02
0.32
1.
78
0.32
0.
39
0.27
1.
06
0.19
5.
7 4.
4 5.
2 4.
5 4.
8 5.
8
MG
900
34.0
12
.8
6.11
0.
33
15.0
4 0.
49
2.61
0.
10
0.55
0.
35
0.24
0.
08
0.13
0.
67
0.07
6.
0 4.
2 5.
2 4.
7 3.
8 5.
1 90
2 17
.3
14.5
4.
57
0.10
5.
77
2.01
6.
95
0.04
1.
35
0.59
0.
24
0.09
0.
16
0.59
0.
13
6.8
4.6
4.6
4.4
4.1
5.0
903
8.1
13.2
5.
11
1.96
5.
06
2.21
6.
47
0.03
0.
47
0.59
0.
26
0.11
0.
19
0.78
0.
15
5.5
4.7
4.7
4.5
4.4
5.5
904
2.2
20.2
7 6.
16
0.95
4.
17
2.05
4.
02
0.02
0.
14
0.94
0.
20
0.17
0.
17
0.83
0.
18
6.0
4.7
4.9
4.6
4.7
6.1
905
0.0
19.0
12
.10
1.26
4.
52
1.65
3.
20
0.01
0.
13
1.61
0.
15
0.20
0.
17
0.93
0.
24
6.1
4.5
5.1
4.7
4.9
6.3
907
3.8
36.6
4.
69
0.66
6.
84
3.00
9.
68
0.02
0.
24
0.72
0.
17
0.05
0.
11
0.71
0.
05
6.0
4.4
4.2
4.1
4.6
5.4
908
0.0
26.7
7.
79
1.16
12
.66
1.65
13
.93
0.02
0.
31
1.29
0.
28
0.33
0.
27
0.77
0.
16
6.0
4.3
5.3
4.3
4.6
5.7
Sum
aúm
a
983
0.0
36.0
2.
85
0.93
1.
85
0.22
4.
87
n.d.
0.
08
-- 0.
16
0.13
0.
23
1.75
0.
27
-- 4.
9 5.
8 4.
7 --
6.5
Mar
ine
99
5 36
.8
4.1
3.69
n.
d.
n.d.
n.
d.
0.16
0.
03
0.05
0.
26
0.22
0.
06
0.09
0.
73
0.06
--
-- --
5.8
4.4
6.1
Anhang
- 161 -
Anha
ng 7
. Met
allk
onze
ntra
tione
n in
der
HM
W-,
der L
MW
- und
der
TD
-Fra
ktio
n so
wie
die
Mas
senb
ilanz
en in
der
Tro
cken
sais
on 2
006
in O
st-H
aina
n (n
.d.:
nich
t det
ektie
rbar
).
HM
W-F
rakt
ion
LMW
-Fra
ktio
n
TD-F
rakt
ion
Mas
senb
ilanz
St
atio
n Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
Fe
Ni
C
o C
u C
d Pb
μmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L
nmol
/L μm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
μm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
WR
A1
1.
67
-- --
1.50
0.
01
0.38
0.
00
-- --
1.63
0.
00
0.01
n.
d.
-- --
n.d.
0.
01
0.02
0.
48
-- --
0.72
0.
64
0.56
B1
--
-- --
-- --
-- --
-- --
-- --
-- --
-- --
-- --
-- --
-- --
-- --
-- B2
0.
09
-- --
-- 0.
00
-- 0.
00
-- --
-- 0.
00
-- 0.
00
-- --
-- 0.
05
-- 0.
57
-- --
-- 0.
55
-- B3
0.
62
0.31
0.
17
0.21
0.
00
0.28
0.
00
1.56
0.
11
2.19
0.
00
0.01
0.
00
2.29
0.
21
2.31
0.
04
0.01
0.
74
0.72
1.
00
0.90
0.
46
0.65
B4
1.
11
0.51
0.
29
0.62
0.
00
-- 0.
01
0.60
0.
18
1.85
0.
00
-- 0.
00
2.48
0.
35
3.30
0.
03
-- 1.
02
0.93
1.
05
0.78
0.
56
-- B5
--
-- 0.
10
0.23
0.
00
-- --
-- 0.
06
0.83
0.
00
-- --
-- 0.
20
1.98
0.
06
-- --
-- 1.
11
0.76
0.
63
-- B6
--
0.34
0.
27
0.47
0.
00
0.26
--
0.40
0.
13
1.49
0.
00
0.01
--
3.14
0.
39
3.33
0.
04
0.02
--
0.76
0.
71
0.81
0.
62
0.46
B9
3.
84
-- 1.
53
3.18
--
-- 0.
00
-- 0.
17
2.21
--
-- n.
d.
-- 0.
22
0.18
--
-- 1.
03
-- 1.
30
0.81
--
--
WW
R
W1
0.29
--
0.18
0.
70
0.00
0.
10
0.00
--
0.09
1.
46
0.00
0.
00
0.00
--
0.33
3.
92
0.06
0.
02
0.84
--
0.90
0.
80
0.78
0.
47
W2s
--
1.12
0.
54
2.01
0.
01
-- --
0.43
0.
20
1.86
0.
00
-- --
5.48
0.
45
5.42
0.
07
-- --
0.76
0.
93
0.77
0.
68
-- W
3s
-- --
0.38
1.
48
0.01
--
-- --
0.24
1.
69
0.00
--
-- --
0.42
5.
52
0.03
--
-- --
0.72
0.
71
0.64
--
W4s
0.
38
0.88
0.
37
2.11
0.
02
0.12
0.
00
1.78
0.
49
1.71
0.
00
0.00
0.
00
3.67
0.
48
5.82
0.
03
0.01
0.
48
0.94
0.
93
0.76
0.
73
0.45
W
5 3.
67
2.60
2.
30
4.40
--
0.45
0.
00
3.18
0.
42
3.51
--
0.00
n.
d.
4.82
0.
46
2.07
--
0.00
0.
51
0.97
0.
78
0.77
--
0.63
W
6 --
2.80
1.
64
1.82
0.
01
0.35
--
0.08
0.
00
2.46
0.
00
0.00
--
6.92
2.
08
4.06
0.
02
0.01
--
0.64
0.
81
0.73
0.
71
0.39
W
7 3.
12
3.30
2.
42
2.59
0.
02
0.55
0.
00
1.69
0.
71
1.76
0.
00
0.00
0.
00
6.25
0.
64
4.91
0.
02
0.01
0.
64
0.86
1.
00
0.66
0.
73
0.58
W
8 1.
73
1.88
1.
08
1.51
--
-- 0.
00
1.36
0.
54
2.68
--
-- 0.
00
5.32
0.
71
6.22
--
-- 0.
75
0.90
0.
99
0.90
--
-- W
9 1.
32
1.34
0.
62
1.23
0.
02
0.30
0.
00
1.12
0.
43
2.47
0.
00
0.00
0.
00
5.12
0.
73
7.23
0.
02
0.01
0.
73
0.80
0.
98
0.89
0.
58
0.64
W
10
-- 1.
25
0.53
2.
19
0.01
--
-- 1.
24
0.42
2.
45
0.00
--
-- 3.
44
0.65
5.
07
0.06
--
-- 0.
79
0.81
0.
86
0.95
--
W12
7.
16
-- 2.
72
4.32
--
1.44
0.
00
-- 0.
57
3.04
--
0.00
n.
d.
-- 0.
37
2.09
--
0.02
0.
66
-- 0.
96
0.90
--
0.72
W
15
-- 1.
91
0.64
1.
17
0.01
0.
48
-- 1.
93
0.63
2.
67
0.00
0.
00
-- 4.
33
1.78
4.
31
0.04
0.
05
-- 0.
72
0.80
0.
77
0.66
0.
39
Anhang
- 162 -
Anha
ng 8
. Met
allk
onze
ntra
tione
n in
der
HM
W-,
der
LMW
- un
d de
r TD
-Fra
ktio
n so
wie
die
Mas
senb
ilanz
en in
der
Reg
ensa
ison
200
7 in
Nor
dwes
t-B
rasi
lien
(n.d
.: ni
cht d
etek
tierb
ar).
HMW
-Fra
ktio
n LM
W-F
rakt
ion
TD
-Fra
ktio
n M
asse
nbila
nz
Stat
ion
Fe
Ni
Co
Cu
Cd
Pb
Fe
Ni
Co
Cu
Cd
Pb
Fe
Ni
Co
Cu
Cd
Pb
Fe
Ni
Co
Cu
Cd
Pb
μm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
μm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
μm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
MD
28
5 2.
68
-- 0.
72
-- --
0.23
0.
02
-- 0.
13
-- --
0.05
0.
01
-- 0.
06
-- --
0.05
0.
47
-- 1.
15
-- --
0.90
28
7 --
1.04
0.
09
2.10
0.
00
0.08
--
2.65
0.
26
3.87
0.
01
0.05
--
2.63
0.
10
2.97
0.
02
0.03
--
1.30
1.
30
1.18
0.
50
1.03
28
8 --
0.64
0.
39
1.81
--
-- --
2.25
0.
79
4.29
--
-- --
4.13
0.
71
5.19
--
-- --
0.76
1.
00
0.81
--
-- 28
9 --
0.44
0.
31
2.62
0.
01
-- --
2.19
0.
68
3.69
0.
00
-- --
4.45
0.
75
4.72
0.
00
-- --
1.20
1.
28
1.02
1.
05
-- 29
0 --
-- --
-- --
-- --
-- --
-- --
-- --
-- --
-- --
-- --
-- --
-- --
-- 29
1 --
0.00
0.
03
0.73
0.
01
0.04
--
1.51
0.
53
3.00
0.
01
0.00
--
8.93
1.
00
5.15
0.
02
0.01
--
1.49
1.
27
1.29
1.
05
0.33
MG
26
4 0.
47
-- 0.
20
1.00
0.
00
0.12
0.
02
-- 0.
25
2.28
0.
00
0.01
0.
00
-- 0.
68
5.42
0.
01
0.01
0.
40
-- 1.
19
1.35
0.
93
0.30
26
5 2.
31
-- 0.
86
-- --
0.39
0.
00
-- 0.
35
-- --
0.00
0.
01
-- 0.
11
-- --
0.16
1.
12
-- 1.
28
-- --
1.03
26
6 --
-- 0.
18
2.40
0.
00
-- --
-- 0.
37
2.66
0.
00
-- --
-- 0.
68
6.68
0.
00
-- --
-- 1.
05
0.89
0.
66
-- 26
7 0.
76
-- --
-- 0.
00
-- 0.
01
-- --
-- 0.
00
-- 0.
00
-- --
-- 0.
01
-- 0.
75
-- --
-- 0.
90
-- 26
8 2.
47
-- --
2.55
0.
01
0.57
0.
00
-- --
4.28
0.
00
0.01
0.
06
-- --
7.37
0.
00
0.01
0.
52
-- --
1.36
0.
67
0.57
27
0 --
-- 0.
19
-- --
-- --
-- 0.
79
-- --
-- --
-- 0.
49
-- --
-- --
-- 1.
21
-- --
-- 27
1 --
-- --
2.42
0.
00
0.11
--
-- --
10.5
7 0.
00
0.00
--
-- --
3.35
0.
00
n.d.
--
-- --
0.79
1.
26
1.00
27
2 0.
24
-- --
3.87
0.
00
-- 0.
05
-- --
5.08
0.
00
-- 0.
00
-- --
4.09
0.
00
-- 0.
64
-- --
1.30
0.
89
-- 27
3 0.
66
2.08
--
5.31
--
0.14
0.
03
3.54
--
4.63
--
0.02
0.
03
5.31
--
3.19
--
0.00
0.
84
1.36
--
1.10
--
0.84
Anhang
- 163 -
Anha
ng 9
. Met
allk
onze
ntra
tione
n in
der
HM
W-,
der L
MW
- und
der
TD
-Fra
ktio
n so
wie
die
Mas
senb
ilanz
en in
der
Tro
cken
sais
on 2
009
in N
ordw
est-
Bra
silie
n (n
.d.:
nich
t det
ektie
rbar
).
H
MW
Fra
ktio
n
LMW
Fra
ktio
n
TD F
rakt
ion
M
asse
nbila
nz
Stat
ion
Fe
Ni
Co
Cu
Cd
Pb
Fe
Ni
Co
Cu
Cd
Pb
Fe
Ni
Co
Cu
Cd
Pb
Fe
Ni
Co
Cu
Cd
Pb
μm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
μm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
μm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
nm
ol/L
M
D
940
-- 0.
44
0.04
0.
42
0.01
0.
18
-- 1.
01
0.22
1.
45
0.00
0.
03
-- 3.
72
0.88
3.
20
0.05
0.
12
-- 0.
95
1.10
1.
10
0.74
1.
33
941
0.08
0.
34
0.01
0.
25
0.00
0.
02
0.02
1.
22
0.31
1.
84
0.00
0.
02
0.02
3.
38
1.00
2.
52
0.03
0.
07
0.76
0.
90
0.94
0.
82
0.89
0.
60
942
0.11
0.
19
0.02
0.
34
-- 0.
05
0.05
1.
59
0.46
2.
15
-- 0.
02
0.01
4.
32
1.19
3.
62
-- 0.
06
1.12
0.
99
1.01
1.
02
-- 1.
13
943
0.05
0.
00
0.04
0.
28
-- 0.
04
0.01
1.
38
0.55
2.
47
-- 0.
01
0.05
5.
76
1.15
3.
95
-- 0.
07
1.49
1.
29
1.10
1.
01
-- 1.
08
944
-- 0.
40
0.25
0.
65
-- --
-- 1.
80
0.66
2.
79
-- --
-- 3.
94
0.90
3.
35
-- --
-- 0.
66
1.01
0.
95
-- --
946
2.33
1.
61
0.96
--
0.00
--
0.02
2.
21
0.77
--
0.00
--
0.01
3.
08
0.86
--
0.05
--
0.56
0.
87
0.86
--
1.19
--
948
-- 0.
65
0.16
0.
88
0.01
--
-- 2.
51
1.16
4.
35
0.00
--
-- 3.
77
1.10
4.
09
0.01
--
-- 0.
93
0.93
0.
91
0.84
--
950
-- 1.
71
1.03
1.
05
0.01
--
-- 3.
15
1.02
6.
94
0.00
--
-- 3.
88
0.96
4.
82
0.01
--
-- 1.
01
1.20
1.
26
1.14
--
951
-- 6.
24
-- 5.
16
0.01
--
-- 5.
15
-- 4.
79
0.00
--
-- 3.
57
-- 2.
79
0.01
--
-- 1.
11
-- 1.
35
1.40
--
MG
90
0 --
-- --
-- --
0.00
--
-- --
-- --
0.00
--
-- --
-- --
0.56
--
-- --
-- --
1.02
90
2 --
0.27
0.
06
0.79
0.
00
-- --
1.28
0.
47
2.10
0.
00
-- --
4.69
1.
54
3.53
0.
04
-- --
1.08
1.
04
0.92
1.
05
-- 90
3 1.
12
0.36
0.
28
0.52
0.
00
-- 0.
03
0.91
0.
56
1.92
0.
00
-- 0.
02
4.97
1.
62
4.35
0.
04
-- 0.
59
1.23
1.
11
1.05
1.
22
-- 90
4 0.
49
0.62
0.
08
0.43
0.
00
0.08
0.
09
1.65
0.
66
1.75
0.
00
0.01
0.
02
2.56
1.
68
2.14
0.
02
0.07
0.
64
1.16
1.
18
1.07
1.
08
1.15
90
5 --
1.13
--
1.34
0.
01
-- --
1.23
--
1.24
0.
00
-- --
3.35
--
1.47
0.
01
-- --
1.26
--
1.26
1.
29
-- 90
7 0.
32
1.18
0.
23
1.12
0.
01
0.12
0.
01
1.67
1.
26
3.53
0.
00
0.01
0.
01
2.92
1.
25
3.85
0.
01
0.06
0.
52
0.84
0.
92
0.88
0.
88
0.76
90
8 --
4.34
--
6.61
--
-- --
4.39
--
5.47
--
-- --
4.49
--
3.44
--
-- --
1.04
--
1.11
--
--
top related