150 Pages
English

Dynamics of phosphorus and sulphur in a mangrove forest in Bragança, North Brazil [Elektronische Ressource] / vorgelegt von Ursula Maria Neira Mendoza

-

Gain access to the library to view online
Learn more

Description

Dynamics of phosphorus and sulphur in a mangrove forest in Bragança, North Brazil DISSERTATIONzur Erlangung des Grades eines Doktors der Naturwissenschaften - Dr. rer.nat. - Angefertig am Zentrum für Marine Tropenökologie dem Fachbereich Biologie/Chemie der Universität Bremen vorgelegt von: Ursula Maria Neira Mendoza Bremen Juni, 2007 IIErster Gutachter : Dr. habil. Rubén Lara ZMT an der Universität Bremen Zweiter Gutachter : Prof. Dr. Venugopalan Ittekkot en Erster Prüfer : Dr. habil. Matthias Zabel MARUM an der Universität Bremen Zweiter Prüfer : Prof. Dr. Ulrich Saint-Paul ZMT an der Universität Bremen Erstes Mitglied : Antje Baum Cand.rer.nat. ZMT an der Universität Bremen Zweites Mitglied : Wolfger Strauss Student der Biologie an der Universität Bremen Tag des öffentlichen Kolloquiums: 26 Juni 2007 ???IIIAbstractThree forests along a 600-m transect within a tropical mangrove in North Brazil were studied in a series of 1-m cores at rainy season to determine the chemical and physicochemical change relative to hydrology and species-specific effects. In North Brazil, macrotides (spring range 4 m) are an important factor associated with the spatial variation of the mangrove species. The results indicated an inundation two times higher at low plain (LP) colonized by a monospecific forests Rhizophoramangle than in high plain (HP) at Avicennia germinans forests.

Subjects

Informations

Published by
Published 01 January 2007
Reads 61
Language English
Document size 3 MB

Dynamics of phosphorus and sulphur in a

mangrove forest in Bragança, North Brazil

DISSERTATION

zur Erlangung des Grades eines

issenschaften Doktors der Naturw

t. - - Dr. rer.na

Angefertig am

Zentrum für Marine Tropenökologie

gie/Chemie lodem Fachbereich Bio

der Universität Bremen

vorgelegt von:

Ursula Maria Neira Mendoza

Bremen

Juni, 2007

: Erster Gutachter

Zweiter Gutachter Erster Prüfer

Zweiter Prüfer

Erstes Mitg lied Zweites Mitglied:

:

:

:

:

Dr. habil. Rubén Lara en ZMT an der Universität Bremnugopalan Ittekkot Prof. Dr. Veen ZMT an der Universität Brem

Dr. habil. MatthMARUM an der Universität Bremias Zabel en
Prof. Dr. Ulrich Saint-Paul en ZMT an der Universität Brem

Antje BaumCand.rer.nat. ZMT an der Universität Bremen lfger Strauss oWen Student der Biologie an der Universität Brem

s: 26 Juni 2007 Tag des öffentlichen Kolloquium

II

Abstract

III

studied in a series of 1-mThree forests along a 600-m cores at ra transect within a tropical miny season to determaine the chemngrove in North Brazil were ical and
ical change relative to hydrology and species-specific effects. In North physicochemBrazil, mspatial variation of the macrotides (sparing range 4 mngrove sp)ecies are an important factor associated with th. The results indicated an inundation two e
timmanglees higher at low plain (LP) co than in high plain (HP) at lonized by a mAvicennia germinans forestsonospecific forests .The flooding gradientRhizophora
mV) levels in LP sedimresult in a slightly acidicents compared to HP (200 m to basic pH (6.3-7.5) and lower redox potentiaV). l (Eh) (0-100

submAt surface sedimergence, whereas during subments (10 cmergence the Eh dropped dras) (HP), the Eh and pH showed stable values during nontically ~ 4.5-fold -
fromsubm 78 to -253 mV, concomergence the relation ~50 mV and ~60 mV per pH unit was mitantly with pH increase to basic values.aintained, in During
submsedimergence was restored in a sinuents and water, respectively. The ~3-fold sulphidesoidal increase over time.drop in the early stage of Clearly, the effect of
submcan amplify the intensity of the measuremergence on these trends is irrefutable. Meanwhile, the presence of the electrodents. The speciation of phosphorous (P) e
sedimsuggested that iron/aluments (0.35 ± 0.09 to 0.56 ± 0.26 miniumg.g-bound P (P-Fe/Al) is th-1) come mpared to calciumain chemi-bound P (P-Ca) cal bound in
(0.03 ± 0.01 mg.g-1). In the ho(extr.-P) with inundation frequency (IF) was confirmrizontal layer (30 cm), the variationed by a significant positive of extractable P
(4-8), reflected the low Ca-phosphate mcorrelation (r=0.81, n=29). At HP, the higher Cianerals precipitation, probably associated to : P-Ca ratio (9-12) in relation to LP
inundation (2-3.5 h). The significant negative correlation (r= -0.74, ethe short-timinium phosphate n=23) between extr.-P and Al : P-Fe/Al ratio, suggested that the alumsedimfraction seements. s to be the most significant fraction in controlling the P concentration in

) profiles, two oxidation and two reduction horizons were In the vertical (50 cmrecognised. Stands of R. mangle andA. germinans differs in their responses to low
sediment Eh conditions. R. mangle showed a moderate but constant root-induced
oxidation capacity (ǻEh=66-125 mV), whereas A. germinans varied from low
(ǻEh=22 mV, MP) to greater (ǻEh=193 mV, LP) oxidation. Implications of these
findings in relation to P availability showed two features. In the first, the intense root-induced oxidation increased P-Fe/Al, total P (tot.-P) and acidification immobilized the
inution of the root-oxidation was accompanied by the extr.-P. In the second, the dime increase of inant pools of tot.-P and P-Fe/Al reflected in an extremdepletion of domextr.-P.

respect to tidal entary OM (4.0 ± 0.8 %) (LP), supported evidence withThe low sedimhydrology, but do not clearly behavsignificant negative correlation (r= -0.74, n=13) between C:N mes with respect to decomposition porlar ratio and Eh ocesses. The
ents, supported by ed the association of higher C:N ratios with reducing sedimconfirmrg.-C) related to total nitrogen (tot.-N). Differences the increase of organic carbon (o

IV

capacity greatly influence sulphuof sulphur speciation between both species, demonstrated that their root-oxidationr cycling in sediments. The SO42- reduction was a
less imhigh abundance of iron keeps accuportance diagenetic pamulating pore water suthway or the reoxidation of hydrogen sulphide due to the lphide concentrations low.
In sediments below (34-47 μM), whereas at LP higher sulphateA. germinans (50 m, sulphide was only detected at mM) occurred. The presence of twoiddle plain (MP )
ation via ic interfaces (Eh = 0 mV) were favourable for pyrite formoxic-anoxzation (DOP) values obtained here are polysulphide pathway. The degree of pyritimiindicative for sulphide-limnute dispersed grains, octahedral crystaiting suboxic cls, framboids and fraonditions. Pyrite morphologies occur as mboidal cluster. The
proxy to identify depositional environmwide variation in pyrite texture suggests thatents. Phosphorus, sulphur and iron dyna caution is needed in using pyrite as redox mics
suggesting that a comare closely coupled to the activity of mabined effect of flooding with the P dynamngroves trees and sulphate-reducing bacteria, ics and the plant-
masedimngroves. ent-microbial feedback contributes to the zonation of the investigated
.

Zusammenfassung

V

lder entlang eines 600 Meter langen Abschnittes innerhalb eines tropischen äDrei W den Mangrovenwaldes in Nordbrasilien wurden in der Regenzeit untersucht umphysiochemischen Waerforschen. Hier sind starke Überflutungen (Fndel relativ zur Hydrologie und zumrühling: Höhe 4 m) artspezifischen Effekt zu ein wichtiger Faktor,
der mzeigten imit der räum Vergleich eine doppelt so starlichen Veränderung der Make Überschwemmung imngrovenarten verbunden ist. Die Resultate niedriger gelegenen
Terrain (LPgelegenen Terrain (HP), kolonisiert ausschließ), den lich duAvicennia germinansrch WäRhizophora mangleldern an. Das Resultat der , wie im höher
basischem pH-Wstärkeren Überschwemmungen zeigt sich in einemert (6.3-7.5) und ein geringeres Redox-Potential (Eh) (0-100mV) in etwas leicht säurehaltigem bis
it HP (200mV). enten verglichen mden LP Sedim

Bei den Oberflächensedimzur Zeit der Nichtüberflutung stabile Niventen (10 Zentimeter) (HP), zeigten die Eh- und pH-eaus wohingegen während der Überflutung Werte
it pH von 78 auf -253 mV), begleitend mert drastisch fällt (ca. 4.5fachder Eh-WZunahme zuRelation von ~50mV und ~60 m den basischen WeVrten. Wä pro pH-Einheit in den Sedimhrend der Überflutung es zeigte sich eine enten bzw. im Wasser.
im Laufe der Zeit sinusförmiDie sulfite Konzentration, die im ersten Stadium der Überflutung umg wieder auf den zu Beginn gemessenen W 1/3 sank stieg ert an.
Offensichtlich ist die Auswirkung der Überflutung auf diese Veränderungen unwiderlegbar. Allerdings kann die Präsenz der Elekerhöhen. Die Fraktionierung von Phosphor (P) zeigt, dass das an P gebundene trode die Werte der Messdaten
Eisen/Aluminium (P-Fe/Al) die häufigste Chemische Verbindung in den Sedimenten
(0.03 ± 0.01 m(0.35 ± 0.09 to 0.56 ± 0.26 mg.g-1g.g-1) im). In der horizontalen Schicht (30cm Vergleich zu Kalzium) wurde die Dyna gebundenemmi P (P-Ca) k von
(r=0.81, n=29) mverfügbarem Phosphor (extr.-Pit der Überflutungsfrequenz best) durch einen signifikanten positivätigt.. Die höhere Ca: P-Ca ratio (9-en Zusammenhang
niedriger gelegenden Terrain, leich zum Verg höher gelegenden Terrain, im12) imder kurzzeitigspiegelt den niedrigen Caen Überflutung, (2-3,5h) wi-Phosphat Mineralenader. Der signifikante negusfall, der sicherlich verbunden ist mativite
deutet darauf hin, dass die AluminiumZusammenhang (r= -0.74, n=23) zwischen extr.-P und verfügbarer Al : P-Fe/Al Wphosphatfraktion die signifikanteste Fraktion ert
entlösung zu sein scheint. der Sedimfür die Kontrolle der Phosphorkonzentration in

In den vertikalen (50cm) Bodenproben wurden zwei Oxidations- und zwei
undA. germinans WäReduktionsschichten festgestellt. Diese Ergebnisse zeiglder sich in ihrer Reaktioten, dass n auf niedrige Eh- Bedingungen des R. mangle Wälder
oderate aber eine mR. mangleents unterscheiden. Über den Abschnitt zeigte Sedimwurzelinduzierte Oxidationskapazität von konstante wurzelinduzierte Oxidationskapazität (ǻA. germinansEh=66-125 m zwischenV), wohingegen die niedrigeren
(ǻAuswirkung dieser Ergebnisse imEh=22 mV, MP) und höheren W Verhältnerten (ǻis zur Verfügbarkeit von P zeigen zwei Eh=193 mV, LP) variierte. Die
gesamDinge. Erstens die intensive Wurzelinduziete P (tot.-P) und durch Säurebildung fixierte das extr.-P. Zweitens wurde die rte Oxidation erhöht den P-Fe/Al, das
e überwiegender bnahmrzeloxidation begleitet von der AuVerringerung der W

Mengen von tot.-P und P-Fe/Al was sich in einemwiderspiegelte.

VI

en Anstieg von extr.-P extrem

(LP), unterstützte HinwDie niedrige Konzentration von organischem Material imeise von Flutungshydrologie aber keine deutlichen Hinweise Sediment (4.0 ± 0.8 %)
it Zersetzungsprozessen. Der signifikante negative auf einen Zusammenhang mlaren Menge C:N und Eh sichert die oZusammenhang (r= -0.74, n=13) der mVerbindung von höheren Mengen C:N und reduzierten SedimAnstieg des organischen Kohlenstoffs (org.-C) in Beziehung zum gesamtenten, unterstützt vom en Stickstoff
(tot.-N). Unterschiede in den Schwefelfraktionen zwischen beiden Mangrovenarten belegen, dass ihre wurzelinduzierte Oxidationskapazität großen Einfluss auf die Schwefelzirkulation in den Sedimenten hat. Die SO42- Reduktion ist entweder ein
Pfad, oder die Rückoxydation von untergeordneter diagenetischerSchwefelwasserstoff wegen des Eniedrigen Sulfidkonzentrationen im Porenwasser. In den Sedimisen Überflusses führt zu häufig auftretenden enten unterhalb A.
germinansaufgefunden, wohingegen im, wurde Sulfid (34-47 μM) nur im m Vergleich im LP höhere SOittlerem42- Werte (50 gelegenen Terrain (M mM) auftratenP) .
Die Anwesenheit von zwei oxidierten und unoxidierten Grenzflächen (Eh = 0 mV) Grad der Pyritisierung Wbegünstigte die Bildung von Pyrit durch Polysulfiderte (DOP) lässt Anzeichen von Sulfid hemmWege. Der hier aufgefundeneenden
suboxisch KonditionenKristallen, octaedrisch gefor erkennen.m Pyrit komten Kristallen, Frammt in Formboiden und Fram von miboidhaufen vor. nuziös kleinen
BestimmDie grosse Vielfalt an Pyritformung von Ablagerungen nur mationen suggeriert, dass Pyrit als Indikator zur it Bedacht gewählt werden darf. Die Dynamik
it der Aktivität der Mangroven zwischen P, Schwefel und Eisen ist eng verbunden mP-Dynamik und Reaktion von Pflanzen-Sedimund der Sulfat reduzierenden Bakterien. Der kombinierte Effekt von Überflutung menten-Mikroben trägt zur Einteilung derit
untersuchten Mangrove bei.

VIIledgements woAckI thank the Center for Tropical Marine Ecology and Núcleo de Estudos Costeiros for MADAM atheir support. This study is part of the Braziliannd was financed by the Brazilian National Research Council (CNPq) and -German Cooperation Project
the Germ03FO154A. The author thanks the Deutschean Ministry for Education ar Akademnd Research (BMBF) under the code no. ischer Austauschdienst (DAAD)
for the scholarship support. supervising this thesis and his assistance The author thanks Dr. habil Rubén Lara for documin scientific support. As well, the author thanks Pents requests and his patience in the final phase of this work, especially thankrof. Dr. Ulrich Saint-Paul for solves s
e. for his secretary Gaby Boehm staff, in specially to Dr.llaboration of the MadamThe author is grateful for the coThorsten Dittmfor the successfully teamar, Helenice Santos, Cleise Cord work, as well as, Mattheiro, Alexanias Birkicht, Dieter Peterkdre Rossi and Sime anpliciod
Andreas Hanning for the technical support. Especially meand Dr. habil. Matthias Zabel for their interest and imntioned are Prof. Dr. Venugopalan Ittekkot, Prof. Dr. Michael Böttcher prove this work.
support.Last but not least, I would like to mention my parents and brothers for their precious

Tables of Contents

List of Figures
List of Tables
List of Equations
List of Abbreviations and Symbols List of Units

CHAPTER 1 - INTRODUCTION 001
ters a1.1. Phosphorus in coastal w1.2. Phosphorus in mangroves 004

CHAPTER 2 - STATE OF THE ART 2.1. Phosphorus in coastal waters 011
2.2. Phosphorus in mangroves 012

CHAPTER 3 - OBJETIVES 017
3.1. Chapter 6
3.2. Chapter 7

CHAPTER 4 - DESCRIPTION OF THE STUDY AREA 019 Location 4.1. Mangrove 4.1.1. forest 019 Hydrology 4.2. 4.2.1. Amazon region
4.2.2. 4.2.3. Coastal Mangrove waters forest
022vegetation Mangrove 4.3. 4.4. 4.5. Climate Geology 024 023

VIII

X XII XIII XIV XVII

CHAPTER 5.1. 5 – Experimental METHODOLOGY design 025 025
5.1.2. Physicochem5.1.1. Elevation surface of the mical pattern andangrove nutrient status of m angroves
5.1.3. Effect of waterlogging and rhizosphere onphysicochem5.1.3.1. Field experimical parameters ent I: waterlogging
5.2. Field work 5.1.3.2. Field experim 028 ent II: rhizosphere
Sediments 5.2.1. 5.2.1.1 Sensor measurements
5.2.2. Vegetation 5.3. Colorimetry and laboratory analysis
5.3.1.1. pH 5.3.1. Sensor calibration, checking and calculation

002

011

017 018

019 019019021021

025026026026028028030030030 032032

5.3.1.2. Redox Potential (Eh) Sulphide 5.3.1.3. 5.3.2. Water content and organic matter
Salinity 5.3.3. etry Granulom5.3.4. ents 5.3.5. Total phosphorus in sedim fractionation Phosphorus 5.3.6. 5.3.7. Extractable phosphorus analyses Elemental 5.3.8. 5.3.9. Sulphate
5.3.10. Percentage of expected concentration Iron 5.3.11. 5.3.12. Degree of pyritization 5.3.13. Pyrite
ratios Molar 5.3.14.

STATUS OF MANGROVES CHAPTER 6 – PHYSICOCHEMICAL PATTERNS AND PHOSPHORUS 053
053 Results 6.1. 6.1.1. Topography and forest structure 6.1.2. Eh, sulphide and pH: changes with flooding tim e
ngrove roots a6.1.3. Eh, sulphide and pH: changes relative to m6.1.5. Phosphorus fractionation in sedim6.1.4. Eh and pH changes with depth (50 cments (1 m) in sedim) ents
ents interaction 6.1.6. Porewater and solid-phase elem6.2. Discussion 6.2.1. Effect of flooding conditions on sedim 073 ents
6.2.2. Effect of the species on sedim6.2.3. Redox stratification and pH pattern ent properties
hosphorus 6.2.4. Inundation frequency and availability of p6.2.5. Effect of flooding on P-exchange 6.2.6. Effect of redox stratification on P-exchange

IX

033034040040041041042043045046047047050051052

053055056062065070

073075077078079082

083

083 086089094

7.1. Results CHAPTER 7- DYNAMICS OF SULPHUR I 083 N MANGROVE SEDIMENT 083
7.1.1. Substrate characteristics: organic matter, salinity, granulometry, Eh and pH 083
7.1.2. Elemental composition and ratios 086
089 pools Sulphur 7.1.3. 7.1.4. Iron pool and pyrite morphologies 094
097 Discussion 7.2. 7.2.1. Elemental source characterization 097
7.2.2. The sulphur-iron pools and some thermodynamic consideration 100
103 7.2.3. Degree of pyritization and pyrite texture

8 CHAPTER 107CONCLUSION –

CHAPTER 9 – BIBLIOGRAPHY 111

CHAPTER 4

List of Figures

Distribution of mangrove in the Northeast coast of the States of Figure 1Reprinted with permPará and Maranhão. Source: Souza-Filho, P.Wission. Journal of Coastal Research.M and Lara, R. (In
ission)submStudy area on the central sector of the Bragança Peninsula. Figure 2

CHAPTER 5

asure the height of inundation during eDevices utilized to mFigure 3 Figure 4 Comspring high tide. Modified frombination of site topography w Cohen et al. (2001). ith contour levels (cm) and
inundation frequency (days/year) of the total area flooded. Distribution of the transect along the topography and position of ents with sensors. Station. A = fixed station, easuremthe mStation. B and. C = horizontal profiles.Figure 5 DiagramA. germinans of sampling wi stands and bare substrate (th electrodes showing the distribution of Station B), and A.
Figure 6 germinansLog-linear calib and R. mangleration curves stands ( for checkingStation C). the S2- electrode
anceperformFigure 8 Figure 7 PerformSequential extractionance comparison of S schem2- and H2e for quantification of two S electrodes.
-bound P (P-iniumentary phosphorus reservoirs: iron/alumsedimFigure 9 Flow diagramFe/Al) and calcium-bound P (P-Ca). of the sequential extraction scheme for the
reactive- (react.-Fe) and pyrite-iron (pyr.-Fe).

CHAPTER 6

Figure 10 (I)and water height (cm Changes for sulphide in sedim) variation during flood and ebb. ents at 10 cm depth (27.6 o(II)C)
(Eh) of flood water surface Changes in pH and redox potential oC) during spring tide. and depth (25.5Figure 11 sediments at 10 cmVariation in water height (cm depth (26.5 oC) during spring tide. ), pH and redox potential (Eh) in
Figure 12 expedition) and sedimFeature of Eh versus pH values for flood water ent during flood (II)and subm(I)erge(24 h-nce
(12 h-expedition). (III)Figure 13 potential (EVariation of surface (10 cmh) along a horizontal profile in) sediment sulphid: Station B, me, pH and redox angrove

X

020020

025027

029036039044049

057058059060

XI

ent (BS) at and bare sedimA. germinansinated by forest domandR. manglehigh plain (HP), and at miStation C,ddle plain (MP). mixed forest of A. germinans
Figure 14germinans and R. mangleVertical distribution of redox potential (m. HP, high plain; MP, miV) and pH below ddle plain; D, A. 064
Figure 15Comparison between analyzeddepressions; LP, low plain. The figures share the sam and calculated phosphorus e axis. 066
fractions of total P (tot.-P) and inorganic P (inorg.-P). Horizontal distribution of inundation frequency (IF), extractable 068Figure 16 ents and leaf phosphorus (Leaf-P) phosphorus (extr.-P) in sedimFigure 17 (Cordeiro, Mendoza Vertical distribution ofet al phosphorus com., 2003 reprinted with permpounds. The figuresission). 072
e axis scales. e samshare th

CHAPTER 7

Figure 18Figure 19Figure 20

Figure 21

Figure 22 Figure 23 Figure 24

Figure 25

(asediments. ) Lateral view of Eh, (b) Lateral view of pH, both measured in 085
Concentration (wt%) versus depth profiles of organic carbon 087germinans(org.-C) and total nitrogen (tot.-N) below at high- (HP) and low-plain (LP). R. mangle and A.
Relation between the organic carbon (wt %) and nitrogen 088ents. The line represents a imcontents (wt %) of Bragança sedleast-square with a slo15.7 ± 2.3 and 17.2 ± 2.8 below pe corrR.esponding with a m mangle and oA. germinanslar C:N of ,
C:N molar ratio vs. redox potential (Eh). HP, high respectively.(a) Horizontal sedimplain; LP, low plain; D1, depression 1. ent profile of sulphide at 20-25 cm depth. 091
salinLateral tranity (cs) sulphate and (ects of iso-lines for sedimd) percentagent pore-water data of (e of expectedb)
ddle plain; LP, iconcentration (PEC). HP, high plain; D1; MP, mlow plain; depressions (D1, D2). Vertical propercentage of expected concfiles of sediment water content, salinity, sulphentration (PEC), degree of ate, 092
plain, (b) LPpyritization (DOP) and reac, low plain. tive iron (react.-Fe). (a) HP, high
Lateral tran(react.-Fe) and (b)sect of iso-lines sedimdegree of pyritization ent data for (a(DOP). HP, high plain; ) reactive iron 095
, low plain; depressions (D1, D2). Pddle plain; LiD1; MP, mMEV miobserved at the Bragança macrophotographs showing various pyrite textures ngrove sediments: (a) poorly 096
pyrite crydeveloped mstals ini a loci (nute pyrite covering a mc), cluster of fraamngrove root, (boids intob a di) group of atom,
and (dDistribution of the number of m) well developed octahedral crystal.icrocrystals (Nm) in framboids 097
ddle plain; ialong a topographic gradient. HP, high plain; MP, mLP, low plain.

CHAPTER 5

List of Tables

Table 1 Total variability calculated in the measurement of various
ents by concentration of elemeters and ical paramphysicochemand numdifferent meber of replicates (n). thods. Average values for variation coefficient (VC)
s of tion of low concentrationCoefficients utilized for the calculaTable 2 , dilution factor. Fsulphide. D

CHAPTER 6

Table 3

Table 4

depth along a ents at 10 to 40 cmactions in sedimrPhosphate fstandard deviation (SDgradient of inundation frequency (IF). Mean values () and range in mg P.g-1. In parenthesis, x),
Physicochempercentage contribution of the sumical characteristics of surface sed of the fractioimns.ents (0-10 cm),
Mean values (xreactive fractions and phosphate molar ratios from), standard deviation (SD) and range. transect.

CHAPTER 7

Table 5

(15-35 cmPhysical and chem) and deep (45-100 cmical param) sedimeters in surface (10 cment sam), mples across a iddle
tidal gradient. C:N slope refers to the ratios calculated through values (xthe inclination (b=slop) and standard deviation (SD).ae) of regression lines in Figure 20. Mean

XII

031

037

054

069

085

Equation 1 2 6 EquationEquation 3 Equation 4 Equations 5a 5bEquations 6a 6bEquations 7a 7bEquations 8a 8bEquation 9 Equation 10

List of Equations

XIII

6

7 8 9

13

13

38

73 75

List of Abbreviations and Symbols

XIV

ACP Amorphous Calcium Phosphate
Adenosin-tri-phosphate ATP Al : P-Fe/Al Reactive-Al : Iron/Aluminium-bound P
AVS avail.-P Acid Available Volatil Phosphorus Sulphide
BP Before Present
BS Bare Sediment
ȕ-TCP ȕ-Tricalcium Phosphate
Carbon C Circa ca Ca2+ Calcium
Ca : P-Ca Reactive-Ca : Calcium-bound P
a Chlorophyl a Chl CinorConc. g Inorganic Concentrate Carbon
C : P Organic Carbon : Total Phosphorus Ratio
C : N Organic Carbon : Total Nitrogen Ratio
Carbon : Nitrogen : Phosphorus Ratio C : N : P Variation of Coefficient CV DW Dry Weight
1 Depression 1 D 2 Depression 2 D DBH Diameter at Breast Height
DCPA Anhydrous Dicalcium Phosphate - Monetite
DCPD Dicalcium Phosphate Dehydrate - Brushite
DIN DIP Dissolved Dissolved Inorganic Inorganic Nitrogen Phosphorus
Water De-Ionized W DIDOC Dissolved Organic Carbon
DOM Dissolved Organic Matter
Nitrogen Organic Dissolved DON DOP1 Dissolved Organic Phosphorus
Pyritization of Degree DOP DOS extr.-P Degree of Extractable Phosphorus Sulfidation
Fe 2+ Iron
Iron Ferrous FeFe(OH)3 Ferric Hydr(oxide)
PyriteFeS 2 FeSaq Fe : P-Fe/Al Reactive FeS clusters Iron : Iron/Aluminium-bound
Phosphorus Ratio Figure Fig. GIS Geographical Information System

+ Protons H Hydroxyapatite HAp Inundation IF Inorganic inorg.-P Freshwater Q Lower LOZ Lower LRZ Mangrove MADAM Meters m.a.s.l max. M
nmim n. N Minutes
MgNm 2+ M Num
N NNH4+ A
O2OCP Oxygen Octacalcium
OMorg.-C O Organic
Organic org.-P Phosphorus P CalciumP-Ca Iron/AlumP-Fe/Al pH react.-Al p Reactive
Reactive react.-Ca Reactive react.-Fe Radial ROL Particulate POM POPOP R R Particulate
Pyrite pyr.-Fe South S SO42-
Sta. S
Scanning SEM Standard SRM Tab. T
Total TDN Total TDP TFS TEAP`s Total
Total TOC Total tot.-Ctot.-Ptot.-N Total Total
Total TRS

Frequency Phosphorus Discharge Zone Oxidation Zone Reduction ent Managemand ics DynamLevel Sea Above aximal
autical Miles
Microcrystals of ber agnesium
itrogen
mmonium
Phosphate rganic Matter
Carbon Phosphorus Phosphorus -bound Phosphorus -bound iniumHH2O iniumAlum Calciumn IroLoss Oxygen Matter Organic Phosphorus Organic oot Porosity
Iron Sulphatetation
Microscopy Electron Material Reference elba Nitrogen lved Disso Phosphorus Dissolved inal Electron Accepting Processes TermSulphide Free Carbon Organic n Carboen Nitroghorus PhospSulphide Reduced

XV

TSS UOZ URZ VC W ZMT

Total Upper Upper Variation W Zentrum

tse

Solids Suspended Zone Oxidation Zone Reduction Coefficient

für Marine Tropenökologie

XVI

oǻGo Free
14C C D Stable
Hour h km2 Kilom
km ‰ Liter lt m mg.g-1 DW M
Milligrammg/l M nim Millimmm mm3M/s Cubic Millim
Millivolts mV Parts ppm Practical PSU Seconds sec t C/ha/y V v%VC P Variation
13 Stable C į weight wt% V ǻ x mμm

List of Units

Change Energyegree Celsius
Isotope Carbon

eter Square Kiloter emppt (parts per thousand = g/kg)

reteight e per Dry W per GramMilligramLiter per inutes
eter olar Seconds per Meter Million per Units Salinity Tons of Carbon per Hectare per Year stloCoefficient ercentage
e isotopgen itronweight dry of % ariation
Averageeter cromi

XVII

Chapter 1 - Introduction

CHAPTER 1 - INTRODUCTION

1

ically-diverse group of woody shrubs, which possess Mangroves represent a taxonoments ility to survive along sheltered tropical coastline in saline environmba common aOn a global scale, about 60-75 % of under tidal influence (Snedaker, 1982). btropical coasts, with a world-wide ngroves are found throughout tropical and suamextension of about 181,000 km2 (Spalding et al., 1997). In South-America mangroves
cover about 23,800 km2, from which 10,713 km2 are estimated to grow along the
Brazilian North coast in the States of Amapa, Pará and Maranhão (Schaeffer-Novelli
erved but under increasing anthropogenic ngroves are still consa., 1990). These met alpressure (Krause and Glaser, 2003), constituting the study area for an integrated
an cooperation e of a Brazilian-Germe framresearch project initiated in 1996 in thent (MADAM). ics and managemproject on mangrove dynam

Mangroves are highly productive ecosystems, where trees can reach net primary
production rates as high as 30 t C/ha/y (Clough, 1988). One third of this production is represented by plant litter, mainly leaves (Robertson et al., 1992). The leaves, flower
ngroves waterways, where aes into me tidparts and propagules can be flushed by thdecomposition by the processes of leaching, saprophytic decay and fragmentation
occur, resulting in particulate matter net transport (export) to the adjacent coastal
waters. There is also evidence for a substantial net export of dissolved organic mater
(DOM), reaching the same order of magnitude as litter export in some mangrove areas
(Dittmar and Lara, 2001a; Twilley, 1985). Outwelling of DOM from the mangrove
mponents of leaf litter. However, no coy be associated by leaching of dissolvedamngroves play for the ageneral consensus has hitherto been reached about the role mexport and import of nutrients for the adjacent coastal waters (Dittmar et al., 2001).

The main goal of the present study was the evaluation of the nutritional status of
h its control on the iron cycle) and Bragança mangrove forest and oxygen- (trouged ents of aquatic. The performdim sesulphate-controlled phosphorus release fromestimations were integrated to the available researches of mangroves, saltmarsh, fens
ility of on of the availaband wetlands rice soils in order to understand the contributi

Chapter 1 - Introduction

phosphorus by flooding and plant-induced processes to the overall nutrient budget of this tidally-driven costal ecosystem

2

ters aPhosphorus in coastal w1.1. Oceanic fertility is largely dependent on the availability of the limiting nutrient
ty (Redfield, 1958). In surface waters ary productiviphosphorus (P) for sustaining prim1prises a ) often comOPof oligotrophic region, dissolved organic phosphorus (D(Orrett and Karl, 1987; Karl and Yanagi, significant fraction of the dissolved pool portant source of1997); thus regeneration of P from DOM is a potentially imbioavailable-P (avail.-P) (Kolowith et al., 2001). Within pools of marine dissolved
e organic phosphorus (POP) arerates of DOP and particulatand particulate P, turnover faster than dissolved inorganic phosphorus (DIP), and seasonal, enabling low ary production inorganic-P (inorg.-P) concentrations to support relatively high primponents in the (Benitez-Nelson and Buesseler, 1999). The behaviour of dissolved commixture of fresh- and sea-water, can be analysed in a classic, linear, estuarine “mixing
” relationship with the salinity, where a conservative behaviour reflects only adiagramphysical mixing (i.e. no growth or decay) and the non-conservative reflect physico-
tive behaviour suggests the ical and/or biological processes. The non-conservachemaddition of matter gained by degradation of organic matter or desorption of mud, and
the loss of matter within the estuary by adsorption onmud and phytoplankton uptake
estuary to river as a function of eyev, 1996). The fate of nutrients varies from(Artemwater flow, turbidity and biota, as a result of these influences, phosphate usually ., 1981).let a estuaries (Morris behaves non-conservatively in turbid

used to evaluate ) and phosphorus (P) are ental ratios of carbon (C), nitrogen (NElemmarine organic matter (OM) production and diagenesis, as well as nutrient
ental properties in phytoplankton, with ate the elems estimregeneration. Redfield ratiothe atomic C:N:P ratio of marine particulate organic matter (POM). These elements
pounds of algae position of basic functional and structural comare essential in the com(Billenet al., 1991). The two major sources of OM to marine sediments are terrestrial
and marine plants with distinctive C:N:P ratios. Marine phytoplankton has an average
In contrast, terrestrial plants are characterized by ratio of 106:16:1 (Redfield, 1958).

Chapter 1 - Introduction

3

1300 and C:N ratios ranging from 10 to 100 300 to ratios ranging fromorganic C:P 100 to an 1300 and C:N ratios ranging fromfor soft tissue, and C:P ratios greater thberg and ., 1986; Goñi and Hedges, 1995; Ruttemet al1000 for woody tissue (Hedges Goñi, 1997). In river-dominated coastal areas a two end-members mixture of
high ental ratios should trend fromrine phytodetritus occurs. Elematerrestrial and ments closest to the riverine input, to ield ratios) C:N:P ratios, in sedim(above Redfprogressively lower ratios approaching the Redfield ratios, with distance away from berg and Goñi, 1997). This pattern would result in a gradient, from outh (Ruttemthe mnear-shore sediments enriched in riverborne DOM, to more distal in which marine
derived organic material becomes progressively a more important component between
organic material pools. Given the predominant contribution of DOM and the rapid
settling of particulate matter from rivers plumes, presume that much of the cycling of
ounts e rivers also discharge significant ames involve DOM. LargC, N, P in river plumof particulate materials (Dagg et al., 2004). Thus, a large fraction of the river-borne
POM is initially deposited near the mouth. The release of nutrients from fluvial
ineralization or desorption fromparticles into the seawater, occurs through remsuspended particles and resuspended bottom sediment (Mayer et al., 1998). In
ld ratios are most relatedoligotrophic waters, such as the Brazilian region, the Redfieto the ratio of total dissolved nitrogen (TDN) and total dissolved phosphorus (TDP), than the dissolved inorganic nitrogen (DIN) and dissolved inorganic phosphorus 1portant are implying that dissolved organic nitrogen (DON) and DOP(DIP) alone, imnutrient sources (Karl et al., 1993; Kolowith et al., 2001). An understanding of the
, requires rine ecosystemaical factors that regulate the production of coastal mchemalso a detailed assessment of the controls on the availability of the essential nutrients
ents. at the adjacent wetland sedim

Phosphorus in mangroves 1.2. ud traps, essential as physical substrate and nutrient Mangroves are very efficient msource for forest development. Biogeochemical processes within the mangrove forest,
mon, are relatively well understood (i.e. ent horizinly within the upper sedima lant). However the role of flushing time, oxides, interaction soil/pfreduction oics for nutrient inundation frequency, degree of water-logging and stagnancy dynam

Chapter 1 - Introduction

4

trapping and mobilization are major questions in mangrove geochemistry. All of these
, from which the inundation interrelated variables can deeply influence the systeminfluences salinfrequency acts as a controller onity and anoxia (Boto and Wellington, 1984). the nutrient availability, because it strongly

with a good supply on OM and sulphate ents are usually anaerobic,Mangrove sedimions, where iron sulphides (FeS2) are common. As in other tropical marine deposits,
dissolved inorganic nutrients concentrations (except ammonium NH4+) are low in
ents, due their rapid turnover, lower redox potential condition and ngrove sedimamrecalcitrant character of OM (Boto et al., 1989). Additionally, they have very small
ounts of mpools of labile OM, but very rapid DOM turnover, linked to the standing aroot materials (Nedwell et al., 1994; Alongi et al., 2000). Dissolved and particulate P
concentrations in mangrove sediment are generally low (Alongi et al., 1992).
tion, the inorg.-P probably represents the Although organic P (org.-P) is the major fraclargest potential pool of plant-available, soluble reactive P (Boto and Wellington,
1988). In the domain of pH that is relevant to most soils, H2PO4- and HPO42-, are the
inant orthophosphate ions (Lindsay, 1979). To the exchange with the estuarine dom leached by rainfall, tidal inundation or waters, the pore- and groundwater P can bedrainage (Dittmar and Lara, 2001b). Mean concentrations of extractable phosphorus
t decrease with tidal height, it can become ngrove forest gradiena-P) across a m(extr. llington, 1983; Silva and Sampaio, 1998).eiting in elevated areas (Boto and Wlimtolerance is one of the possible cause This supports that the differential flood determining a spatial pattern of distribution(i.e. zonation) of the mangrove species
(Snedaker, 1982).

rhizosphere pH in t uptake and changes inIn this section are described 1) the nutrienistry, 2) activity of roots due the redox conditions developing in s of the P chemtermergence, and 3) precipitation and dissolution of phosphate ents following submsedimnerals. im

pounds as nucleic acids, Phosphorus is present in plants as a constituent of comary sphospholipids and ATP (Marschner, 1995). The primource of P for plants is

Chapter 1 - Introduction

5

ed extractable, exchangeable, labile, rmthodology also teeinorg.-P, depending on the ms available or bioavailable P (Salcedo and Medeiros, 1995). In most natural ecosystemplant growth is limited by the availability of inorg.-P (Smith et al., 2003). The major
intenance of cation-anion balance in plants and aprocesses responsible for the m-+ and 2) accumulation and or OHchanges in rhizosphere pH are: 1) extrusion of Hdegradation of organic acids (Haynes, 1990). The generation of rhizosphere acidity or ounts of nutritive alkalinity is directly influenced by plant roots since unequal amcations and anions are typically absorbed. In order to maintain the electroneutrality at
the soil-root interface, the uptake of the anion P (PO42-, H2PO4-, HPO42-), can be
stoichiometrically balanced by the active excretion of OH-, similarly than the uptake
of a cation (e.g. NH4+, Ca2+, Mg2+, K+, Na+) is balanced by H+ excretion (Haynes,
d odifiethe cells, or pH-stat, can be mical pH maintenance within 1990). The biochem sechanismmby strong or weak organic acids via carboxylation/decarboxylation tals eount of organic acids excreted solubilize P by chelating mm(Davies, 1986). The aions that would immobilize it or forming soluble complexes with P via metal ions, or
., 2000). Such changes in et alann both (Saleque and Kirk, 1995; Kirk, 1999; Neumrhizosphere pH can occur in response to adverse nutritional conditions, such as P (Haynes, 1990).deficiency

A morphological adaptation of mangrove species to flooding is the formation of
various forms of aerial and adventitious roots (Scholander et al., 1955; Blom et al.,
the aerial components are divided in horizontal arches andRhizophora1994). In ns supply oxygen to the undergrounde columvertical columns known as stilt roots. Thn roots through lenticels and roots entering the roots. Air passes into the columsediment are composed by aerenchyma tissue. In Avicennia, shallow, horizontal roots
radiate outwards, often fora distance of many meters. At intervals of 15 to 30 cm
vertical structures known as pneumatophores emerge up to 30 cm above the mud
interface. A single Avicenniatree of 2 to 3 m height may have more than 10 000
pneumatophores. The pneumatophores, like the column roots in Rhizophora, have
lenticels and aerenchyma. Simple physical diffusion trough the lenticels and along the
aerenchyma is probably the main mode of gas movement in mangrove roots (Hogarth,
s is so effective that it even aeratengrove root systems a1999). Gas transport by m

Chapter 1 - Introduction

6

ent in a radial direction the so-called radial oxygen loss (ROL) (De surrounding sedimSimone et al., 2002; Haase et al., 2003; Begg et al., 1994). The radial extension of the
nt) around the root tips eoxidized zone (higher redox potentials then the bulk sedim (McKee, 1993). to 0.5 mextended for < 3cm

supports the oxidation of ammonia to nitrate and sulphide to The leakage of O2e suppression of phytotoxins such as soluble sulphide sulphate, and consequently thspecies (i.e. H2S, HS-, S2-) (Thibodeau and Nickerson, 1986; McKee et al., 1988).
y occur in the apounds mHowever, oxidation of sulphides to non-toxic sulphur comilation (Mendelssohn and McKee, to assimoxidized rhizosphere or in the root prior onstrated roots 1988). Previous studies of flood-tolerant plants such as rice, demuptake sulphur as sulphate (i.e. SO42-) (Pezeshki et al., 1988). Sulphate uptake was a
major factor causing foliage damage and yield reduction in rice plants (Vamos and
Koves, 1972).

bilize iron through the oxidation of omAdditionally, the produced oxygen can imr)oxides, resulting in a zone of Fe(III) as Fe(III) (hyditatedFe(II) which is precipaccumulation near the roots (i.e. iron plaques), generation of protons (H+) and
consequently pH decrease (Begg et al., 1994) (Equation 1, Begg et al., 1994).

4Fe2+ + O2 + 10H2Oo 4Fe(OH)3 + 8H+

Eq. 1

This process may have some effects on the trapping of P as FePO4 (Equation 2, Silva
ould therefore depend on the balance o, 1998). The supply of P to roots wpaiand Sambetween sorption on Fe (III) (hydr)oxides and release from acid-soluble forms, such
-bound P (P-Ca) (Saleque and Kirk, 1995; Silva, 1992).as calcium

Fe(OH)3 + H2PO4-o FePO4 + OH- + H2O

Eq. 2

ents proceeds roughly in the sequence predicted erged sedimThe reduction of a submby thermodynamics (O2, NO3-, Mn(IV), Fe(III), SO42-), when one oxidant (electron
acceptor) is depleted, the next more energetic redox reaction (i.e. higher free energy

Chapter 1 - Introduction

7

change - ǻGo) will occur in a specific redox potential (Stumm and Morgan, 1981;
overlap as the theoretical Libes, 1992). In fact, in a real system these reactions cantter is oxidized, dissolved COaitantly, organic mconsiderations. Concom2accumulates, and the pH of acid sediments tends to increase and that of alkaline
e initial Eh hents to decrease, stabilizing in the range 6.5-7.0 (Kirk, 2003). Tsedimmpanying Oucing substances accoe release of reddrop is apparently due to th2depletion, such as Fe(II), Mn(II), NH4+ and S2-, before Mn (IV) and Fe(III)
(hydr)oxides can mobilize their buffer capacity (Ponnamperuma, 1972). The most
relevant consequence of reduction for the plants is the increase in concentration of
plant-available P. Since most flooded acids soils have a supply of Fe (III)
ds soils is largely due to nt, the increase in pH of aci(hydr)oxides than any other oxidathe reduction of Fe (Ponnamperuma, 1972) (Equation 3, Ponnamperuma, 1972).

Fe(OH)3 + 3H+ + e o Fe2+ + H2O

Eq. 3

ngrove, anoxic conditions affect the aents, as the mittently flooded sedimIn intermequilibrium of Fe(OH)3-Fe2+ system (Ponnamperuma, 1972). An example of this
change is illustrated for FePO4.2H2O (strengite) and Fe3(PO4)2.8H2O (vivianite). Iron
in strengite is present as Fe(III), whereas in vivianite it is present as Fe(II). When acid
ents are flooded and reduction occurs, Fe(II) increases and depresses the sedimsolubility of vivianite, making it the most stable phosphate mineral. When the
submerged sediments are again drained, oxidizing conditions return, causing vivianite
ost likely Fe and/or Al phosphate. to dissolve. The P is released in some other form, mIn this way tightly bound Fe and Al can be solubilised under reducing conditions. In
acid sediments, AlPO4.2H2O (varicite) is shown as the most stable mineral, followed
., 1989). let aent-Fe (Lindsay by strengite in equilibrium with sedim

ents Ca phosphates are the most stable minerals. They decrease in In alkaline sedimsolubility in the order Ca9(PO4)6.nH2O (amorphous calcium phosphate-ACP)
CaH(PO4).2H2O (brushite or dicalcium phosphate dehydrate-DCPD) > CaH(PO4)
(Monetite or anhydrous dicalcium phosphate-DCPA) > Ca4H(PO4)3.2.5 H2O
(octacalcium phosphate-OCP) > ȕ-Ca3(PO4)2 (ȕ-tricalcium phosphate- ȕ-TCP) >

Chapter 1 - Introduction

8

Ca5OH(PO4)3 (hydroxyapatite - HAp) > Ca5F(PO4)3 (fluorapatite) (Lindsay et al.,
in source of inorg.-P in nature a1989). Apatite written in this work as P-Ca, is the m(Dorozhkin, 2002). Autighenic apatite is precipitated in estuarine pore-water and in the water-sediment interface of marine sediments during early diagenesis (Cappellen
aqueous solutions involves and Berner, 1991). The precipitation of apatite frommultiple steps, such as nucleation of a precursor calcium phosphate phase, cluster
itationation to apatite. The net precip and transformaggregation, crystal growth, ionic strength, Ca:P ratios, and the reaction depends of several factors, including pHpresence of other ions such as Mg2+, CO32- and F- (Sahai, 2003). From these ions, H+
catalyze the precipitation and Mg2+ can block the adsorption sites on calcite inhibiting
the precipitation (Brown, 1981; Yadav et al., 1984; Wang et al., 1995). In
2+ental studies the precipitation did not start at a low Ca:Mg ratio (0.2), Mgexperim eek in carbonate-freepede the nucleation and growth of apatite for one wions can imsynthetic seawater (Cappellen and Berner, 1991) and for 700 days in brackish seawater, at 23-25oC and pH 6.75-7.3 (Gunnars et al., 2004). At a given degree of
saturation, the formation of apatite is significantly faster at a high Ca:Mg ratio (4.5-9)
(Gunnarset al., 2004). Moreover, a high Mg/Ca ratio benefits Mg2+ only in the first
step, but because the overall ǻGo for CaPO4 is faster, Ca2+ ultimately prevails (Sahai,
2002). The precipitation is performed in neutral to alkaline environments as follows
, Nriagu, 1976).4Equation(

5CaCO3+ 3HPO42- + H2O + H+o Ca5(PO4)3OH + 5HCO3- Eq. 4

Protons catalyse apatite precipitation by neutralization of its surface charge, point ofd reduction of the surface tension between nuclei 7 to 8, anHrface charge at pzero su., 1989). At et aland solutions, thus promoting nucleation and growth (Christoffersen oC, for pH = 7 the growth rate of fluorapatite is nearly twice than at pH = 8 at 25 identical degrees of supersaturation (Cappellen and Berner, 1991). It has been3-crystal catalyzes the groups on the apatite suggested that protonation of PO4exchange of PO43- between the bulk solution and crystal surface, and, hence,
accelerates both precipitation and dissolution (Christoffersen and Christoffersen,
rner, 1991). This explains why the rate of dissolution at e1982; Cappellen and B

Chapter 1 - Introduction

9

constant pH in non-stoichiometric solutions far from equilibrium, is not uniquely
d anlues of both the calciumagiven by the ionic product, but depends on the actual vChristoffersen and Christoffersen, 1982). (the phosphate concentrations

odels have been already proposed eight ms of apatite dissolution, echanismFor the mitations and odel was found to have limulate the processes involved. Each mto sim was able to describe apatite dissolution in general. Most of draw-backs, none of themthem were elaborated for apatite dissolution in slightly acidic or nearly neutral
oconditions (4< pH < 8) and temC (Dorozhkin, 2002). The peratures of 25 to 37chemical approach describes thphosphates, considering as a hydrogen catalytic me transformaechanistion of apatite inm (Equationsto acidic calciu5a, Nriagu, m
, Dorozhkin, 2002): 5b1976;

Ca5(PO4)3OH + 4H+o 5Ca2+ + 3H2PO4- + H2O

Eq. 5a

Ca5(PO4)3(OH, F) + 7H+o 5Ca2+ + 3H2PO4- + HF, H2O Eq. 5b

Pore waters of sediments deposited on the continental margins are often more acidic
than the overlying seawater as a result of the microbial degradation of OM (Ben-
ents are ngrove sedimaYaakov, 1973). The values of pore-water pH in organic-rich min the range 5.5-8, with a mean value around 6-7 (Middelburg et al., 1996, Clark et
al., 1998; Marchand et al., 2004; Alongi et al., 2004). The lower pH values found in
ce the solubility of apatite iments are unfavorable to apatite precipitation sinthese sedincreases with decreasing pH. This effect, however, is partially offset by the kinetic e residence heffect of the flooding water pH which acts in the opposite direction. Ttimapatite. In manure, the of the flooding water can act as a facte longer the residence otimr controllinge (< 16 weeks), the clo the dissolution kinetics of ser the solution
P levels would approach the chemical equilibrium (Wang et al., 1995). In wetlands, P
(Olilaet alflux in soil drained for 6 weeks was 10-fold h., 1997). The prediction of long-termigher than in P retention in daily-flooded surface soils drained for 3 weeks
horizons of maporewater pH, requires a better knowledge of the stability ngroves, caused probably due the basic flooding water that increase the of phases with which P is
dation degree and creek e inun To date, it is uncertain to what extent thassociated.

Chapter 1 - Introduction

water pH may be relevant for the stabili

ty of P com

., 2004). et althe short scale of a tidal cycle (Cohen

pounds in m

10

ents on ngrove sedima

aic of mThe dynamded soil in relation to nutrient uptakengrove roots growing in floo

is poorly understood. More detailed investigations are needed on the interaction

ent characteristics, flooding, bacteria induced reduction of sedimbetween sedim

ent and the root-induced oxidation and/or acidification controlling the environm

nutrient status of the whole ecosystem.

ent

Art ate of the Chapter 2 – St

CHAPTER 2 – STATE OF THE ART

11

ters aPhosphorus in coastal w2.1. Despite the important role that mangroves play in the biogeochemical cycles of
ents, only a few quantitative long-termcoastal and even off-shore environmland (Australia) (Boto and port balances exist. These are Hinchinbrook Isexport/imBunt, 1981; Boto and Wellington, 1988; Alongi et al., 1998; Ayukai et al., 1998),
.,et alattayakorn Rookery Bay (FL, USA) (Twilley, 1985), Klong Ngao (Thailand) (W1990) and Bragança Peninsula (North Brazil) (Dittmar, 1999; Dittmar and Lara,
ngrove functions as source and also effective a2001a). In Hinchinbrook Island the m); in Rookery Bay was sink for dissolved nutrients and organic carbon (org.-C ents and suspended org.-C; and in Klongined a net export of dissolved nutridetermngrove flux investigations aNgao was reported export of inorganic nutrients. All msotidal regions with exception of Bragança ecro- and miwere carried out in motides enable a net export. Many discrepancies between the acrPeninsula, where mstudied mangroves may have the resulted from methodological differences and
difficulties in accurately determining temporal and annual material fluxes (Boto and
istry or ent chemWellington, 1988), and different tidal range, topography, sedim., 1998). et als (Ayukai community structure of the ecosystem

(2001b) based on the “Eulerian” approach ar et al. In the Bragança Peninsula, Dittme-series of nutrients, identified a net export of dissolved inorganic nutrients and tim(e.g. N, P and silicate) from a creek catchment area (2.2 km2) to the Caeté Estuary. A
2 phosphate with average 0.05 mmol/msignificant export of/d, occurred principally ion of Ca-species during the dry season. Previous studies suggested the immobilizat., 2001), leading to asymmetries in the tidal et alngrove inundation (Cohen aduring mconcentration oscillations (Dittmar et al., 2001a). Tidal range and flooding therefore
determined direction and quantity of nutrient and OM exchange between mangroves
and the ocean.

Art ate of the Chapter 2 – St

12

Phosphorus in mangroves 2.2. ngrove zonation, such as, an expression of plantaVarious factors can contribute to morphic change, a physiological response to tide sucession, a response to geomgradients and differential propagule dispersal (Snedaker, 1982). In mangrove systems,
ics focused two and nutrient dynameation on inundation regimmost available informin topics: amngrove zonation (Snedaker, 1982; aical properties on mpact of physicochem1) the im., 1999).et alBoto and Wellington, 1984; Matthijs tter (Hesse, 1961; a2) the study of various redox processes in relation to organic mClarket al., 1998; Alongi et al., 1998a; Alongi et al., 1999; Marchand et al.,
., 1998). et alllington, 1983; Silva e2004) and nutrients (Boto and W

R.e discrepancies exist about the differential abilities of In these literatures, sommangle and A. germinans to make the sediments less hypoxic. Nickerson and
as a strong correlation between ngrove in BahamaThibodeau (1985) described on a mand proposed that only the distribution of these species and the amA. germinans has a capacity to reduce pore-water sulphide. Inount of hydrogen sulphide in the sediment,
ngroves (McKee, aanother, similar studies, the opposite associated to the Belizean m., 1999) was found. The apparent et al1988) and Gazy Bay (Kenya) (Matthijs differential effect reported by these authors, did not consider previous studies (Scholanderet al., 1955; Gill and Tomlinson, 1977) which demonstrate physiological
a and lenticels). All ecies (e.g. aerenchymadaptations to flood tolerance for both spthese studies, utilized electrochemical measurements with electrodes for the
l, inserted in extracted pore-water or ination of sulphide and redox potentiadetermically thodology is physical-chemeents. Strictly, only the latter mdirectly on sedimeaningful, but is also subjected to various error sources.m

-AvicenniaA redox stratification model was proposed by Clark et al. (1998) for an dominated Australian mangrove sediment profile (<70 cm depth), where the
concentration and chemical speciation of metals are influenced by the distribution of
ically distinct horizons. In the reduction horizons with a pH > 7 and an Eh <geochem-150mV, metals are largely present as sulphide-bound species (e.g. FeS2), whereas in

Art ate of the Chapter 2 – St

13

the oxidation horizons with pH < 7 and an Eh > +100 mV, most metals are present as
exchangeable or (hydr)oxide-bound species (e.g. Fe(OH)3, Fe2O3). The upper
ents ly drying and re-oxidation of sedimoxidation zone was related to the periodicaland also the bioturbation. The lower oxidation zone was related to oxygen intrusion in ent via adventitious roots of plants and via diffusion of air through pores, the sedimdrained during low tides. As a result of rapid changes in redox potential in the upper horizons of mangrove sediments, sulphate reducing bacteria (e.g. Desulphiovibro
) utilise the organic molecules and obtain energy by reducing sulphate desulphuricansto potentially metal-biding sulphides, from ferric compounds (e.g. Hematite Fe2O3 or
Fe(III) (hydr)oxide) to iron sulphides (e.g. Mackinawite FeS(1-x), Greigite Fe3S4,
Pyrite FeS2) (Equations 6a, Kirk, 2004; 6b,2+ Clark et al2+., 1998). In equation 6b M2+
represents any divalent metal, usually Ca and Mg. The dissolution of iron +
and large sulphides occurs when it reacts with oxygen, this oxidation generates H7aamandount of s 7b, Kirk, 2004). ulphuric acid, lowering pore-water pH in the oxidation zones (Equations

2Fe2O3 + 8SO42- + 16CH2O + O2ĺ 4FeS2 + 16 HCO3- + 8 H2O Eq. 6a

9CH2O + 4SO42- + 4Fe(OH)3 + 4M2+ĺ 4FeS + 4MCO3 + 15H2O +5CO2Eq. 6b

4FeS2 + 15O2+ 14H2Oĺ 4Fe(OH)3 + 8SO42- +16H+ Eq. 7a

FeS + 3/2 O 2 + H2Oĺ Fe2+ + SO 42- +2H++ 2e- Eq. 7b

The formation of pyrite is an indicator of the oxic-anoxic transition (Wilkin et al.,
1996) or the iron oxide barrier (Clark et al., 1998). Pyrite formed in mangrove
sediments has the dominant textural forms of euhedral cristals and framboids
ics in other wetlands, e) and P dynam, 2003). Studies of sulphur, iron (Fet al.(Gleasonindicates that a strong sulphide/reactive Fe dependency controls P solubility under ., 1989). The latter hypothesis et al, 1984; Caraco et al.reduced conditions (Boto ation of iron sulphides can prevent the re-supply of iron oxides to suggests that formle for the vegetation. ation of extr.-P availabents, enhancing the formimsurface sed

Art ate of the Chapter 2 – St

14

obilization and diated by plants on meThe effect of oxidation and acidification mremobilization of nutrients, has been seldom studied in mangrove species. Worldwide,
several studies on this effect were done ments. Saleque and ly in wetland rice sedimina the acid-Kirk (1995) suggested that the bulk of P taken up by the plants was from ferrous hydroxides and orphous-bound P, amsoluble pools included calciumcarbonates. Since the zones of acid-soluble P depletion coincided with the caused P solubilization. e acidificationthacidification zones, they concluded that alternately flooded and drained Another studies investigated the P uptake by rice from ., 2003). They concluded that in the continuously floodedet alsoils (Huguenin-Elie was consistent with acidification as a result of oxidation of soil, the P solubilizationFe2+ by oxygen (O2) released from roots (Equation 1), contrary to the moist sediment
tion was coincident with organic anionooded where the P solubilizalalternately f roots. excretion from

-bound P iniumIn Brazil, pioneer studies showed that P-Ca (Silva, 1992) and iron/alum(P-Fe/Al) (Sthe tidal inundation appears ilva and Sampaio, 1998) were the mto influence these chemaiin inorg.-P species. In both studies cal reactions involving P
retention and release. The hypothesis that differential flood tolerance is one of thepossible causes determining a pattern of mangroves distribution has been studied in
Bragança Peninsula, where macrotides constitute the main hydrodynamical features of
thods eation was based on chemical m. The available informthe coastal ecosystem(Reise, 2003), GIS tools (Cohen et al., 2004) and empirical models (Berger et al., in
ass and productivity studied by Reiseission). The variation in structure, biomsubm(2003) showed an interdependency between inundation frequency, bulk density, odels, the structural patternent salinity and P concentration. In the empirical msedimL. racemosawas characterized by a disproportionate change in growth rates beteween andA. germinans during early forest succession, which might be explained by
species-specific differences in nutrient-uptake efficiency (Berger et al., in
ission). subm

pography, the inundation Our study considered for the plant availability of P the toation/acidification) to tide-frequency and the physiological plant response (i.e. oxid

Art ate of the Chapter 2 – St

maintained gradients, in agreement with the m

st relevant mo

ngrove za

15

onation factors

considered by Snedaker (1982). Physiologically, the effect of the substrate conditions

is ultim

leaves.

reflected inately

the d

tion and accumstribui

ulation of nutrien

ts in roots and

Art ate of the Chapter 2 – St

16

apChes etivter 3 – Obj

CHAPTER 3 – OBJETIVES

Physicochemical patterns and phosphorus status of mangroves 3.1.

17

jor aimaThe mplore the general hypothesis that differences s of this chapter, was to ex

ngrove forests colonized aistry in relation to phosphorus uptake by ment chemin sedim

byR. mangle and A. germinans, are a function of physical environment (e.g. low- vs.

high-intertidal position) and the species-specific capacity to alter sediment conditions.

ical parisons of physicochemil, the comAlong a topographic gradient in North Braz

paired stations at tion) parameters in ical (i.e. phosphorous fractiona(Eh, pH) and chem

a) identify the species-specific effects toequivalent topographic elevations were aim

relative to the effects of the hydrology (i.e. inundation frequency), and b) assess the

intensity and importance of each parameter (e.g. species-specific effects and

ove zonation. ngrahydrology) as factors influencing the phosphorus availability and m

ation upon the following specific points: pts to gain informThis chapter was attem

ical parama) Evaluation of the variation for physicochem eters (e.g. Eh, pH, sulphide)

in surface (10 cm) sediments subjected to submergence and non-submergence of

acro-tides, to examm pact of short-time inundation on the precipitation andine the im

nerals (e.g. P-Fe/Al, P-Ca). idissolution of phosphate m

eters (e.g. Eh, pH, ical parament physical-chemination whether the sedimb) Determ

ngrove aerating structures (i.e. prop roots aity to msulphide) were affected by proxim

forR mangle and pneumatophores for A. germinans), to compare the relative effect of

the presence of one and twical icochement physo species aerating structures on sedim

eters (e.g. Eh, pH, sulphide). param

ter 3 – ObjapChes etiv

Dynamics of sulphur in mangrove sediment3.2.

18

In this study, R. mangle and A. germinans forests inhabiting different tidal elevations

ine the termesettings in the Bragança Peninsula (North Brazil) were addressed to d

ents of both forests. The distribution spatial pattern of iron sulphur speciation in sedim

ngroveaof m forest allowed us to test the hypothesis that difference in the sulphur and

andh-intertidal position low- vs. higiron pools are a function of physical settings (e.g.

fferent of dian the capacityent composition) rather thregional differences in sedim

nt conditions. emspecies to alter sedi

This chapter was aim to provide additional information on the following specific

points:

ination of the differences in organic ma) Examong the present tter diagenesis ama

plant species precursors, assessing the origin and degree of organic matter

preservation by C:N ratios. The data on carbon and nitrogen pools allow the

estimation of their efficiency of mineralization at some of the habitats.

iting factors foration of iron sulphides, the importance of reactive iron and limb) Form

of pyrite used as a redox pe stabilityer, thrthpyrite formation. Fu

depositional environments.

entify y to idoxr

Chapter 4 – Description of the study area

CRIPTION OF THE STUDY AREA CHAPTER 4 - DES

19

4.1. Location The research region is the mangrove ecosystem located Southeast of the Amazon
(Pará) and São The sector between Beléme State of Pará, Northeast Brazil.delta in thLuis (Maranhão) represent 83% of the total mangrove area of the Amazonian States
Amapá, Belém and Maranhão (Fig. 1). This sector is considering the world’s largest
unitary mangrove system, estimated to cover a total area of about 8,900 km2 of
coastline (Kjerfve and Lacerda, 1993; Kjerfve and Macintosh, 1997; Kjerfve et al.
1997).

Mangrove forest 4.1.1. st side of theaThe study area is situated on a higher part of the central sector to the Engrove belt along about 350 angrove peninsula of Bragança, which is part of the mamkm southeast of the Amazon estuary (46o42’16’’W - 46o42’02’’W to 0o55’40’’S -
0o55’58’’S, 2.4 m.a.s.l.) (Cohen et al., 1999). The MADAM project area is located
.(Fig. 2)between the mouth of the Rivers Maiaú and Caeté

Hidrology 4.2. region azon 4.2.1. AmThe mouth of the Amazon estuary, which is the dominant freshwater source influence
in Northern Brazil, is composed by two major systems: The Amazon and the Pará
Rivers, whose discharge greatly varies with season. The Amazon has large enough salt water into the river mouth (Gibbs, flow year-round to prevent the intrusion ofnt to the continental shelf. The empeded transport of sedim1970), allowing uniazon River varies from 23 to 123 mg/l during dry and the Amsuspended load froments adjacent to these rainy season, respectively (Gibbs, 1967). The coastal sedimrivers contains organic carbon to nitrogen ratios (Corg:N) which exceed 10, reflecting
the terrigenous or freshwater POM, these ratios decrease seaward remaining higher
ddle and inner shelf South of ients on the man, 1975). Superficial sedimthan 5 (Millimposed with greater than 90% sand (Shepard, 1932). the Pará River are com

Chapter 4 – Description of the study area

Figure 1:Figure 2:

Distribution of m

ngrove in tha

e North

Maranhão. Source: Souza-Filho, P.W

ission. perm

Journal of Ceast coast o

the States of Pará anf

.M and Lara, R. Reprinted with

oastal Researchission). (In subm

Study area on the central sector of the Bragança Peninsula.

20

d

Chapter 4 – Description of the study area

21

waters 4.2.2. Coastal A large number of rivers drain the coastal area Southeast of the Amazon-Pará system.
é and Turiaçu, which are the Caeté, Gurupi, Maraçumost amongst these riversForemflow throughout the year. The plumes of these rivers are advected to the northwards-tern boundary current-the North Brazil-Guyana System flowing onto a flowing wes) shelf, which led to transport the suspended load to the Northwest broad (200 kmalong the coast. Its inner edge reaches within 20 to 40 km of the coast and the strength
of the current depends greatly on the local wind stress (Milliman et al., 1975; Edmond
et al., 1981). The Amazon-Pará system can also receive sediments of the Amazon
en and coastal ional regimbasin, especially at the rainy season due the more eroslation.circu

forest4.2.3. Mangrove The mangrove system of Bragança Peninsula can be described as a riverine /fringe
mangrove according to the criteria of Lugo and Snedaker (1974), which the tide
is oodroffe, 1992). The Caeté EstuaryWinated with strong bidirectional flux (domcharacterised as macrotidal, with tidal amplitudes of about 4 m range, where a major
ann, 1998;ngrove is inundated only during spring tides (Schwendenmapart of the mthe mCohenaet al.ngroves about 3000 km2 1999). The Caeté River is about 100 kmof the basin (Schwendenmann, 1998). long and drains upstream from

the Bragança Peninsula is across several ngrove creeks of aThe circulation in the m with the estuary of the Caeté tidal channels which link the forest flooding systemRiver and the Maiaú Bay (Cohen et al., 1999). The construction of a embankment 35
km road at the middle of the peninsula, from 1986 to 1974 impides the transversal
flow of the water tide at several locations (Cohen et al., 1999) (Fig. 2). As suggested
in others mangroves (Wolanski et al., 1992), the pronounced asymmetry between the
intaining aents and mcurrents exporting sedimebb and flood tides, results in larger ebb deep the tidal channels. This tidal motion was considered as a possible vehicle of
idiurnal withm different sources (Lara and Dittmar, 1999). Tides are semnutrients fro., 1999), resulting in no vertical salinity et als) (Cohen /strong tidal currents (1.5 m

Chapter 4 – Description of the study area

large areas of active erosion and deposition in the external part of the t andgradienestuary, with strong shifting of sand banks.

22

vegetation 4.3. Mangrove ote sensing along the Northeast of Pará antification with remngrove areas quaThe me world`s largest unitary ering as thand Northwest of Maranhão, sector considmangrove system, estimated an area of 7500 km2 (Kjerfve and Lacerda, 1993) (Fig.
1). This area represents 57% of the total Brazilian mangroves for still conserved
ngroves, but under amng anthropogenic pressure. increasi

rshes, al units are found, such as salt mIn the Bragança Peninsula several vegetationaassland in stable and mobile dune rngrove types, restinga and coastal gadifferent mrsh areas localized in the central peninsula region occur at avegetation. The salt mslightly higher elevations than mangroves (Cohen et al., 2004). The mangroves are the
2la Bragança Peninsuinant vegetation type covering about 90% of the 180 kmdom(Krauseet al., 2001). Three typical mangrove species dominate this region:
(Avicenniaceae) and Avicennia germinans(Rhizophoraceae),Rhizophora mangle ed forest with mprise a well-developbretaceae) and co(ComLaguncularia racemosa ., 2003). These species are representative for et al height (Menezes tree of 10-25 m eida, 1996). Concerning the pattern ofothers coastal regions of the State of Pará (Almmangrove distribution, Avicennia dominates in high elevated areas, Rhizophora and
e on lower areas, including borders of ediate areas, whil co-occur in intermAvicenniatidal creeks and channels, Rhizophoradominates (Menezes et al., 2003).
inates in disturbed sites (e.g. occurs in a wide distribution, but domLagunculariagher isition zones to the h2000). In the tranchannel borders) (Thüllen and Berger,elevated salt marsh area, scrub or dwarf Avicennia stands of2-3 m tall occur
., 2003). et al(Menezes

Chapter 4 – Description of the study area

23

4.4. Geology A large proportion of the Amazonia consists of Cretaceous or Tertiary sedimentary
gion, occurs along the Bragança-Viseu deposits (Sombroek, 1984). The Bragança repaleotopography of the basin deposits controlled by the coastal basin, with Quaternaryentary evolution of this coastal plain began (Souza Filho, 2000). The Holocene sedimat aprox. 5,100 14C years BP, when the relative sea level reached a maximum. From
14 entary evolution is a result of coastline progradationC years BP the sedim5,100with development of the mangrove system (Souza Filho and El-Robrini, 1998;
.,let aBehling2001).

azon Basin is widely covered by lateritic rocks, especially ferruginous and The Ambauxites (Costa, 1991). Laterites are defined as products of intense sub-aerial rock weathering whose Fe and/or Al content is higher and silica content is lower than the
bedrock (Schellmann, 1983; 2003). They consist predominantly of mineral
assemblages of goethite, hematite, aluminium hydroxides, kaolinite minerals and
quartz. The relative positions of the fields of kaolinisation, weak or immature
ture lateritisation vary, aoderate lateritisation and strong or mlateritisation, mdepending on the parent rock material (Bourman and Ollier, 2002). In the Proterozoic
Gurupi Group (Pará-Maranhão) exists a wide diversity of lateritic rocks, prevailing
plateaus as Serra do Pirocaua between the terial is localized inathe phosphoric. This mé Rivers (Costa and Araújo, 1996), in islands as Trauira near the Gurupi and Maraçum.,et al(Costa near the Gurupi RiverTuriaçú River (Brandt, 1932) and as Itacupim2004a).These laterites are derived from basic-ultrabasic and meta-sediment bedrocks
plateaus and islands are sphoric laterites found in ture phoa(Costa, 1991). The m-7 m characterized by 1) a ferruginous crust generally at the top of the profile (1thickness), 2) an Al-phosphate horizon lying below (maximum 2-6 m thickness), and
thickness) above the bedrock. Contrary, the 3) a saprolitic clay horizon (30-50 md are characterized by 1) a sand/clay atures laterites occur only in low-land animm thickness), and 3) acovering (0.3-2 thickness), 2) a ferruginous horizon (0-3 mmmatures laterites saprolitic clay horizon (< 15 m thickness) above the bedrock. The i

Chapter 4 – Description of the study area

24

ent sources from the Barreiras ied by Costa (1991) as part of the sedimwere identifsediments.

alluvial, estuarine and coastal plain.The Bragança coastal area can be subdivided ined by the ateau, is formlent of these plains, which forms the coastal pThe basements of the Barreirase, occupied by Tertiary sedimation of Miocene agrmPirabas FoFormation and post-Barreiras Formation (Bigarella, 1975; Goés et al., 1990; Arai et
al., 1994; Souza Filho et al., 1997). The chemical composition of the Barreiras
Formation display two signatures 1) a terrigenous one with material interpreted as
weathered products (laterites) of the source area (most made of SiO2, Al2O3 and
Fe2O3) and 2) a robust marine influence, as indicate the strong salinisation and
of Na and Cl as halite, K as feldspars ce e presens presence, reinforced by thdiatomand Mg as complex clay mineral (Rossetti et al.,1989, Souza Filho et al., 1997; Costa
et al., 2004b, Behling and Costa, 2004; Berrêdo, 2006). Additionally, higher-plant,
algal- and animal-derived organic matter promotes diagenesis and the new formation
(autigenic mineral) of minerals (mainly pyrite) (Marchand et al., 2003; Behling and
Costa, 2004). It was also proposed thhigher-plant and can be available after de degradation ofat the source of phosphor the organic mus is accumulated in the atter (Costa
).personal communication

Climate 4.5. diaracterised by a tropical warm and humThe coastal region of Northeastern Pará is chclim oate. The average temperature of 25 oC has a narrow annual thermal amplitude
C) characteristic of intertropical regions. The average rainfall is 2500 mm (22-32ecipitations between January and May the pr(period 1973-1997) with 75 % ofcreased freshwater runoff by the Caeté (INMET, 1992). During the rainy season, inRiver reduces salinity in the mouth of the estuary (17 o/oo), while during dry season
salinity 38 o/oo prevail (Dittmar, 1999).

Chapter 5 - Methodology

CHAPTER 5 – METHODOLOGY

5.1. Experimental

design

5.1.1. ngrove arface of the mElevation su

25

ic in the transect, a topographic survey of In order to evaluate the inundation dynam

ber 2001. The ngrove floodplain was undertaken during a spring tide at Septemathe m

water level relative to the sedim

ined through water-ce was determent surfa

ents devices installed at 32 stations along the transect (measurem

e ). ThFig. 3

measurements of the maximal high-tide were taken and these values were adjusted to

tidal level after Cohen et al. (2001). Afterwards, inundation frequency (IF

expressed as the number of times that the tides reached each topograph

the tide table. lated fromthis transect in one year calcu

Figure 3:

), is

ical quota on

easure the height of inundation during spring high Devices utilized to m Cohen et al. (2001). tide. Modified from

Chapter 5 - Methodology

26

5.1.2. Physicochemical pattern andnutrient status of mangroves
ine the differential flood tolerance as The field work was carried out in order to determber 2001 to April 2002. ngroves during Septemaa factor influencing zonation of mrface, each of e transect of the elevation suThree forest zones were chosen in the samthese extending approximately 200 m. At a total of 11 sampling stations,
approximately 200 sediment samples were collected for physicochemical and
chemical analysis. The first zone began at the edge of the central road (Fig. 2) through
anA. germinans monospecific forest, the second zone was a mixed stand of A.
germinansandR. mangle, and the third zone was a R. mangle monospecific forest
ending at the Caeté Estuary (Fig. 4). To encompass the nutrient status with a
st structure, a standard Point Centered prehensive description of the spatial forecomQuarter (PCQ) method was used (Cottam and Curtis, 1956). At each sampling station,
the species less abundant (i.e. principal) was identified and considered as the center of the species more abundant in the quarter a 10 to 10 m quadrant. Three to five trees ofighbors) were also considered. In this with the canopy around the principal tree (i.e. neapproach the structural attributes for the neighbor group of trees was assessed (Staupendahl, 2001). In the mixed stand alternate species positioned as central tree
were regarded.

5.1.3. eters ical paramter logging and rhizosphere on physicochemEffect of wa

ent I: water logging Field experim5.1.3.1.

pecific forest onos mA. germinansA fixed station was established at the high plain in to examine the instrumental variation (i.e. electrodes) with time of measures and to
ics during two response to flooding dynamical changes in ine the physicochemdetermspring tides ofApril 2002 (Fig. 4). Station A was localized at 50 cm distance from the
ine the possible roots influence during the tree to examR. mangleprincipal trunk of an inundation (00o55’40.5’’ S, 46o42’16.4’’ W).Sensors of pH, redox potential (Eh),
sulfide (S2-) and temperature fixed at 10 cm depth on the sediment took measures
every 10 minutes, with higher resolution during submergence (3-5 min.), during 12
and 24 hours. The level of flooding water was recorded each 5 min. during both
eters of the ical paraminundation periods. Mobile sensors recorded the physicochem

Chapter 5 - Methodology

ately 2 cmimfloodwater at approx

. )mrface (bottothe water su

Figure 4: Com

above the sed

bination of site topography with

ent layer (surface) andim

contour levels (cm

2 cm

27

below

) and inundation

the total area flooded. Distribution of the transect frequency (days/year) of

along the topography and position of the measurem

ents with sensors.

and. C = horizontal profiles. Station. A = fixed station, Station. B

Chapter 5 - Methodology

28

ent II: rhizosphere Field experim5.1.3.2. ents ical modifications in sedimine the baseline of the physicochemIn order to determby the presence of roots, two areas with different inundation frequency were selected. tals profiles of pH, redox potential (Eh),At each area with different length, horizonsulfide (S2-) and temperature for the evaluation of the effect of root superposition were
monospecific A. germinansdone. The high plain area (Sta. B) was localized in the forest at 50 m from the road (00o55’40.9’’ S, 46o42’15.8’’ W) (Fig. 4). Here, the
measurements (n=14) were taken at 10 cm depth along a 22 m transect perpendicular
atophores branches for two to the road, following one of the network of pneumoppositeAvicennia trees separated by bare sediment (Fig. 5).

The middle plain area (Sta. C) was located in a mixed stand of A. germinans andR.
mangle at 280 m of distance from Sta. B (00o55’50.8’’ S, 46o42’09.5’’)(Fig. 4). The
measurements (n=12) were taken at 10 cm depth along an 11 m transect following the
same scheme as for Sta. B. The transect followed one aerial root from a R. mangle
tree, crossed a bare sediment until reaching one pneumatophore branch of a central A.
R. mangleto the opposite side site until reaching another tree, continuinggerminans).Fig. 5(

rk oField w5.2. ents Sedim5.2.1. A stainless steel sediment corer, 1 m long by 8 cm diameter, was used at each station
to sample near (<1 m) the trunk ofthe central tree and one of the neighbor trees.
inundation frequency, allowing e the samThese paired stations are subjected toseparately comparing the biological effect of each species on the sediment. The core
situIndiately after extraction to avoid oxidation. ewas covered with a plastic foil immples were taken centred at the depths of 5, 10, 15, 20, 30, 40, 50 thick sub samm5 cples were put into intervals down to 1 m depth. The sam and thereafter at 25 cmcmoC.dark airtight polyethylene bags, returned to the laboratory on ice and stored at 4ogenized, oven ent cores were hom sections of the sedimBefore analysis, the 5 cmdried at 60oC until constant weight and powdered in a porcelain mortar to pass a 1
mm mesh.

Station C

Station B

A.

A.

and

stands (

; BS, bare sedimddle plain; Rhi, ient; MP, m

.

). Abbreviations: HP, high plain; Avi,

pling with electrodes showing the distribution of

), and

stands and bare substrate (

of sam

: Diagram

29

Chapter 5 - Methodology

R. manglegerminansStation CmangleR.A. germinansStation BgerminansFigure 5

Chapter 5 - Methodology

30

ents easuremSensor m5.2.1.1. To compare the effect of roots (i.e. oxidation) on the sediments, paired measurements
were made in the proximity (50-100 cm) to the adventitious roots (i.e.
e species. Vertical profiles were ectivres for the respatophores) or aerial structupneumdone across the transect by insertion of steel needle electrodes (Microscale Measurements) in 11 stations during low tide. Each profile was measured in situ down
. The electrodes were inserted for several m depth with resolution of 5 cto 0.5 mminutes into the mud until stable values were reached, in the same range of the
stabilization time needed in the calibration. After each insertion, they were thoroughly
washed and subsequently dried with tissue paper in order to prevent poisoning by sulphide. The instruments employed were a 50 cm length combined Eh / sulphide
de ectrode, both connected to a separate reference electroelectrode and a pH el to one of the r was linked Each senso(Ag/AgCl) filled with a 3M KCl solution.sockets of a digital millivolt meter and fixed in a depth gauge with a fine-scale
resolution until 50 cm depth. All the sensors were checked and calibrated before,
during and after the field work.

Vegetation 5.2.2. Height and breast height diameter of A. germinans and R. mangle from the principal
tree and the canopy neighbor’s trees were recorded.

Colorimetry and laboratory analysis 5.3. The colorimetric method for the determination of phosphate, is basedon the reaction
olybdate of these ions with an acidified molybdate reagent to yield a phosphomcomplex, which is then reduced by ascorbic acid to a blue color proportional to P
e analysis of orthophosphate, aliquots of 10 ple. For thconcentration present in the samml sample were reacted with 0.2 ml of the reductant (acidified ascorbic acid solution)
and 0.2 ml of mixed reagent (Grasshoff et al., 1983). The P concentration in sediment
samples was determinedutilizing absorption quartz cells of 1 cm, using a Hitachi U-
. eter at a wavelength of 880 nm2000 Spectrophotom

Chapter 5 - Methodology

31

/L was utilized to A phosphate standard stock solution (Titrisol) of 1000 mg PO4), used for different calibration prepare two stock standards (526 and 52.6 μM PO4standards. The calibration curve prepared at the same day of the analysis considered
h tions of increasing concentrations. Eactwo blanks and five to eight working solubatch of samples (n = 10) included two blanks. Most of the chemical analyses were
average for variationconducted on duplicate and triplicate, its experim coefficient (% VC) (Tab. 1). The chemental error was expressed as an ical and physical-
chemical analyses without replicate were not considered in this table.

Table 1: Total variability calculated in the measurement of various physicochemical
ents by different methods. Average emeters and concentration of elparamvalues for variation coefficient (VC) and number of replicates (n).

Methodology

PorewaterpH(Eh) Potential Redox idehSulp

SedimentTotal Phosphorus
Inorganic Phosphorus
Iron/Aluminium-bound P
bound P -mCalciuExtractable Phosphorus
Total Carbon on bnic CargaOr enNitrogSulphate Reactive Iron, Aluminum and Calcium
Pyrite Iron

VC) (%

5 2.1310

8 2.6 2.0 8.0 9.5 4.2 3.6 3.7 3.6 1.0 1.2 1.

hors Autn

37Grasshoff et al., 1983
Küster & Thiel, 1982 37107 Cline, 1969; DIN 38405-D26

Black, 1955 Leeg & 2Black, 1956 Leeg & 22Kurmies, 1972
2Kurmies, 1973
2Hesse, 1957; 1961
& Stern, 1984 Hedges 22Verardo et al., 1990
& Stern, 1984 Hedges 2Hesse, 1957 22Lord, 1982
2Lord, 1983

Chapter 5 - Methodology

32

All lab ware used in the determinations was washed with HCl solution (10%) and
etry of ). Glassware used for the colorimter (DIWathoroughly rinsed with de-ionized wP and sulphate/sulphide was washed separately. Traces of organic matter from the
teflon bottles were completely removed by submitting them to a preliminary digestion
ntrol of bottle air tightness ple analysis. The coas for samilarwith acid persulphate simith screw caps during the oxidation, was checked by weighing the teflon bottle wbefore and after the autoclaving. In all samples the colour interference of suspended
particles was corrected by filtration of the extract before the addition of the
chromogenic reagent. Only in the fractionation analysis of sediments, the brown
colour interference of the unfiltered sediments extracts, was corrected with a “colour
blank” obtained by the measure of the samples a second time with all reagents except
ogenic reagent.the chrom

Sensor calibration, checking and calculation 5.3.1. pH 5.3.1.1. of five standard phosphate buffers (200 lThe pH electrode was calibrated in 50 m hydrogen phosphate ). The acid-base buffers were prepared with potassiumMm(K2HPO4) and potassium dihydrogen phosphate (KH2PO4), using the Henderson-
Hasselbach equation to compute the ratio of the base form to acid form and a pH-
gression analysis (pH as independent, mV ter to guide to the correct pH. The reemsignal as dependent variable) calculated the slope as a response to the ion being detected and this was compared with the linear theoretical Nernstian slope 59.16
mV/pH at 25oC (i.e. for each pH unit the potential change 59.16 mV) (Cheng and Da-
correction, when idering 1) the slopeMing, 2005). The electrode was regenerated consctive output) and 2) the of the effethe slope dropped to values <50 mV (85 % = Vetric potential correction, when the deviation of the isopotential point (0 masymmperature compensation was controlled pH 7) ± 30 mV increased. Additionally, the tem length (HI 8757, ments using a steel temperature electrode of 50 ceasuremby m the sensor perature error, resulting in a drifting signal as response ofHanna). The temto different environments temperature, can be reduced by the correction coefficient of
d for Grasshoff et al. (1983). For this 0.0114 pH units per each grad Celsius proposelied, considering that the buffers and study the temperature correction was apptuiin s

Chapter 5 - Methodology

33

measurements performed in different media resulted in an average temperature
difference of approximately 3oC, which corresponds to a correction coefficient from
0.0342 pH units.

Redox Potential (Eh) 5.3.1.2. h) electrode was calibrated using seven separate phosphate The redox potential (Ebuffers (200 mM) in 50 ml DIW saturated with chinhydrone (~1 crystal/5 ml
paigns, the seven pH-equidistant buffers were prepared phosphate buffer). For all camtaken in account the field pH range (6-8). The potential of this saturated redox buffers values of Vter regression analysis, the mperature, thus afdepends only on pH and tempared to the respective standard n at each pH were comthe chinhydrone solutiosolutions at 25oC and calculated from the expression as follows (Näser, 1976; WTW,
1997):

o:CStandard solutions at 25 VpH 1.68 = 383 m VpH 4.01 = 255 m VpH 4.60 = 220 m VpH 7.00 = 78 m:CalculationU (mV) = Uo (mV) – slope (mV) x pH – U ref (mV)
Uo (chin) (mV) = 0.7175 – 0.00074 x t (oC) x 1000 (mV)

Slope(mV) ln10RT1000mV§ 59 at 25oC
Ftential easured electrode po= m Uo Uo chin = Temperature compensation of the potential of chinhydrone solution
o 0.0699 V§C)(0-37 reference f= potential o U ref bs per mole) = the Faraday Constant (96,485 coulom F ole) = the Gas Constant (8.3145 joules/degree/m R T (K) = the Absolute Temperature (273.15 + t (oC))
2.303 = the conversion factor from natural to base 10 logarithm

U ref F R T (K) 2.303

portant focusing two options: is highly im) depends on pH, the value of the latter 2-+ S-S + HS2HȈide (TFS) concentration ( fraction of the total free sulph2-the fact that the S (Rickard and Luther, 1997). Due inant (reducing) at pH >7-dom2-neutral at pH=7, to Sinant in -dom-inant (oxidizing) at pH <7, through HSS-dom2 Hchanging pH, fromrely by e oxidized to reduced ms may change from systems is that reducedsystem naturalIn ). Equations 8 a, b) (-S) or partly dissociated sulphide (HS2sulphide (H) ions in aqueous solutions, but not to the concentration of undissociated 2-sulphide (Stration of fully dissociated d to the concene is relatece electrodd a referenelectrode an. The potential between the sulphide S), mounted on top of a conducting wire2(Ag/Age idctrode is a crystal of silver sulph of the ion-specific sulphide ele tipThe sensingSulphide 5.3.1.3. Chapter 5 - Methodology

34

The rH-value (hydrogen-redox exponent by Clark) is used to characterize the redox strength for a redox system, setting the redox potential and the respective pH-value
. e systemrelated to thUo / 0.029 V + 2pH = rH- value

inples and the second option for ended for water samThe first option is recomm situis work. ents, the latter applied in theasuremm

e standard hydrogen electrode, by adding The reported redox data are relative to th207.7 mV to the original millivolt reading obtained with an Ag/AgCl 3M KCl
oC (Küster and Thiel, 1982). 5e at 2reference electrodUH(mV) = Uo (mV) + U ref (mV)
UH = redox potential (Eh)

1)convert all forms of free sulphide to S2- by raising the pH to e.g. 12, applying
ffer (SAOB), or ant-bu sulphide anti-oxidthe so-calledbient pH value rate the electrodes at the amcalib2)

Chapter 5 - Methodology

35

The calibration of the S2- electrode was performed under anaerobic conditions
wasMoving the oxygen by autoclaving. A sulphide stock solution of ±100 mrem e andlose tissu sulphide, wiped dry with a celluprepared washing crystals of sodiumweighed (± 2.4 g Na2S.7-9H2O). The crystals were added to 100 ml hot DIW and
subsamples were dispensed in 5 or 10 ml aliquots. After allowed to room temperature
the concentration of TFS with an iodometric titration was assayed (Jander and Jahr,
1963) and colorimetrically performed with the methylene-blue method (Cline, 1969)
thods (DIN 38405-D26). ean standard mafter Germ

2- electrode, five pH-adjusted (4.5 - 8.5) phosphate buffers For the calibration of the Sof 200 mM strength in triplicates equilibrated at the same temperature (± 1oC) were
beaker, 50 ml of phosphate buffer was added. The lprepared. To a clean, dry 100 mbeaker was placed on a magnetic stir plate, added a magnetic stir bar and stirred at
slow speed (no visible vortex). After rinsing the reference, sulphide and pH electrodes with DIW and blot dry, the electrodes tips were immersed in the buffer solution. To
wise 100 μl of the sulphide stockeach pH-adjusted buffer concentrate seven stepsolution were added, resulting in a TFS increase from 162.3 to 1077.8 μM (log TFS
2.2-3.0 μM). By each of these serial additions, after stabilization readings (3-5 min),
the mV directly from the meter display was read. This procedure was repeated for all
the buffers.

re constructed by plotting an average weOne calibration curves per pH-adjusted buffer V) as dependent variable as a function of the sulphide ured potential (measof the mions activities (or effective concentration) in an increasing series (log TFS 2.2-3.0
potential, corresponding to the level ofμM) as independent variable. The measured2- ion in solution, is dSscribed by the Nernst equation: eUo UrefSlopelog (A)

Uo U ref A

easured potential = mf= potential o reference = level of sulphide in solution

Chapter 5 - Methodology

Slope(mV) 2.3RT|29at 25 oC
nF

charge ionic = n

36

an ion in ncentration ofity or effective cotivThe ionic level (A) refers to the acsolution. The sulphide ion activity is related to free-ion concentration in a specific
easure of the number of ions taking part in any given a mionic strength. In fact, is by the electrode (Orion, e ion selectedith those interacting wreaction, in this case th2003).

To check the performance of the electrodes the slope value was calculated using linear
regression. Log-linear calibration curves with a theoretical slope of -29.5 mV per concentration decade at room temperature were obtained (Boudreau and Jørgensen,
oC was acceptable between 25 to 30 V2001). The slope in the range of -25 to -30 m(Orion, 2003)(Fig. 6). If more as two calibration points not provided a slope within
r this case the electrode was not considered. Foe calibrationthe appropriate range th paper, afterward with polishing paper for a few with cottonmbrane was cleanedemnutes. i for about five mWseconds and rinsed with DI

0

50-3

2,02,22,42,6

-700Potential (mV)

y = -31,1x - 81

y = -25,5x - 374

y = -23,5x - 571

,28

log S2- (μM)
2,3,03

5 5.pH 6.pH05 7.pH

Figure 6: Log-linear calibration curves for checking the S2- electrode performance.

Chapter 5 - Methodology

37

In the electrodes manual, an illustration of S2- calibration shows for concentrations
slope as the theoreticallower than 162.3 μM (log TFS < 2.2 μM) an evident higher 2-) V signal (i.e. < -200 mvalue (data non-showed). To encompass the lower field Sthod, the sulphide stock solution was diluted 10 to elibration mwith the standard caes with the respective pH-adjusted phosphate buffer. Calibration curves per 100 tim V) as a function oftial (measured potenpH-value were constructed by plotting the mμM) and 100 (log TFS 0.2- the dilution factors 10 (log TFS 1.2-2.0the log TFS froms rationg series. In fact, the slopes obtained by log-linear calib1.0 μM) in an increasinwith the dilution factors of 10 (-134 to -82 mV/decade) and 100 (-80 to -44
al. mV/decade) were higher as the theoretic

Table 2DF, dilution: Coefficients utilized for the calcu factor. lation of low concentrations of sulphide.

log Sul (μM) fide (mBuffer V)

pH 4.5 DF 10

1.1.21 51 --237228
1.1.68 80 --242242
242-90 1.1.2.97 03 --245243

5 . 6pH

2.21 51 2.68 2.80 2.90 2.97 2.03 3.

432- 482-502- 512- 522- 525-- 533

log Sul(μM) fide (mBuffer V)

5 . 4pH

2.2.51 21 --289286
2.2.80 68 --303300
307-90 2.2.3.97 03 --311316

pH5 . 7

21 2.51 2.68 2.80 2.90 2.97 2.03 3.

427- 449- 514- 532- 544- 551- 561-

Coefficient

pH4.5 / pH 4.5 DF 10

25 1.22 1.24 1.25 1.27 1.27 1.30 1.

6.5 pH 7.7 /pH

99 0.93 0.02 1.1.04 04 1.05 1.05 1.

Chapter 5 - Methodology

38

tween diluted and non diluted consecutive nce beMoreover the potential (mV) differee . Four averag(Tab. 2)pH-buffers results in a constant coefficient of 1.12 ± 0.07 tentials lower e field pohcoefficients between pH-values 4.5 to 7.5 were calculated. Tthan -200 mV were multiplied by these coefficients until attaining the respective pH-
ple before the standard sulphide calculation.value of the field sam

atic due to S electrodes can, however, be problemThe use of well-functioning Ag/Ag2Sination of the second dissociation constant of Hthe large uncertainty in the determ2(Meyeret al., 1983, Millero et al., 1988) calculated for the conversion of S2- activity
to total sulphide concentrations (Equation 8 a, b Kühl et al., 1998). The very high
pK2 of the sulphide system also precludes the application of Ag/AgS2 electrodes in
acidic environments, where S2- is practically non-existent (Kühl et al., 1998).

Eq. 8a

Eq. 8b

H2S + H2Oļ HS- + H3O+ Eq. 8a
K1 = [HS-][ H3O+] / [H2S]
andHS-+ H2Oļ S2- + H3O+ Eq. 8b
K2 = [S2-][ H3O+] / [HS-]
with[Stot2-] = [H2S] + [HS-] + [S2-]
leads to [H2S] = [Stot2-] / ( 1+ K1 / [ H3O+] + K1K2 / [ H3O+]2)
Considering these limitations, in a test performed in cooperation with Julian Oxmann
and Ing. Matthias Birkicht (ZMT), the sulphide ion activity derived from our S2-
voltametric macroelectrode (tip 1mm) was compared with an H2S amperometric
microelectrode (tip <30μm), designed by AMT (Analysenmesstechnik GmbH,
Rostock) (Jeroschewski et al., 1996) and hired from UFT (Zentrum für
Umweltforschung und Umwelttechnologie, Bremen). The comparison was performed
perature in a defined phosphate buffer solution of 200 mM strength at pH 6.7 and temof 23oC. Both sensors were inserted in this buffer solution and the concentrate ions
(S2-, H2S) were simultaneously measured by 17 stepwise additions (7x25μl, 4x125μl,

Chapter 5 - Methodology

39

6x50μl) of sulphide stock solution, resulting in an increasing concentration range (3.6 check the previous separately calibrations to 518.2 μM). This curve considered a wide concentration range, with the aim(i.e. non- dilution, dilution factors 10 and to
100).

0

00-1

00-200-3)Vml (iatnetoP-400

00-5

00-602

1412M)10S (μ286 Hgo llactieroeth
420

02

4

46

862-(μM)og Sl

8loS (μM) g H2

01

01

Figure 7: Performance comparison of S2- and H2S electrodes.

21

21

41

41

Chapter 5 - Methodology

40

pared the procedure to recognize the concentrations data by In this test, it was comeach electrode. The results gave evidences for its good measurements performances
S glass electrode with wider curve linearity. Attention was given to the H(Fig. 7)2than the S2- needle electrode. Therefore, the limit of determination and detection for
es lost the linearity at th electrodboth electrodes are different. Moreover, boconcentrations less as 64 μM (log TFS 1.8 μM). It is important to note, that the H2S
amperometric microelectrode is suitable for a liquid medium and the water-sediment
interface. Newer development designedelectrodes, are able to measure in soft
sediments, when the electrode can be inserted into the sediment with a micro device.
The use of needle (stainless steel) electrode can be a possibility, especially for in situ
e t thecessary to find several linear areas, buents depth profiles, also when it is nsedimpH should be theoretically >12, what make their use under field conditions un-
le. practicab

tter aWater content and organic m5.3.2. is often used for dry weight correction of other ttenination of water conThe determanalytical data (e.g. salinity), where soil moisture is usually determined by thermal
ple (3 g) was dried in an oven at 105ºCsediment samdrying. For this study, a fresh until constant weight. The loss in weight (i.e. water) was expressed as percentage of
tter. adry m

The total organic matter (OM) content of the sediments was determined from the loss
in weight when OM was destroyed. It was estimated in the same dried sample after
e weight loss was expressed as ignition or ‘loss-on-ignition’ at 550ºC for 4 hours. Thper cent OM. Both methods were run according to the German standard methods
DIN38414-S2 and -S3, respectively.

5.3.3. SalinityTotal soluble salts of a 1:5 soil-water extract were homogenized with a glass stirrer
ined in the clear supernatant and allowed to stand 12 hours. Conductivity was determ

Chapter 5 - Methodology

41

soil conductance were expressed as salt -LF 197). The results of the specific (WTWinger, 1996) by equation: ple (Ensment samcontent of the sedimKs Ke(((WWMM100100))Vm)

Ks (o/oo) = calculated salt concentration of the sediment sample
Ke (o/oo) = salt concentration in extract
ple ent sam = weight of sedimW (g) M (o/oo) = sediment moisture content
Vm (ml) = volume of water used to prepare the 1:5 sediment-water extract

etry Granulom5.3.4. ine the particle size-distribution of termthod of Folk (1974) was used to deeThe msand and silt/clay fractions of the core samples for 5 cm intervals. 15 g DW sediment
etaphosphate (calgon solution), hexam of 0.5 g/l sodiumlwas dispersed with 250 maining sieve. The proportion removernight under stirring and sieved through a 64 μm was recorded as considered as sand fraction, was oven-dried and the DWe,in the siev fraction was ple weight. The proportion of the silt/claye of initial samtaga percentage. difference and expressed as percenylated bcalcu

ents Total phosphorus in sedim5.3.5. It was used Leeg and Black´s ignition method (1955), on the experimental basis
peratures. They verified the obtained by these authors with different ignition temaccuracy of an ignition procedure focussing two critical points 1) that the solubility of
at the entirethe original inorg.-P does not change as a result of ignition, and 2) th ple appears in the extract of theount of org.-P present in the non-ignited samamoC they found an equal (10 %) of ignited sample in the form of inorg.-P. At 240 positive and negative errors related to the increase in solubility of Pi and incomplete
decomposition of org.-P, respectively. Additionally, this method was selected for a
suitable standard of comparison with the P status reported for other Brazilian
mangroves (Silva, 1992; Silva et al., 1998; Silva and Sampaio, 1998; Oliveira, 1999).
In this procedure 0.5 g DW of sediment sample was placed in a muffle furnace (240

Chapter 5 - Methodology

42

oC, 1 hr). To the ignited samples were added 4 ml of concentrated hydrochloric acid
(HCl), heated in a steam plate (60 oC, 20 min) and allowed to stand at room
HCl was added, allowed to stand room lperature. An additional quantity of 5 mtemtemperature, after which was agitated for 30 minutes. The suspension was centrifuged
the supernatant, the P concentration n) and after dilution (1:100) ofi, 10 m(3000 rpmwas measured by colorimetry. A same sample was extracted and analyzed by the
same procedure but without ignition. The increase in extracted inorg.-Presulting from
ple. ignition was taking as the content of org.-P in the non-ignited sam Phosphorus fractionation 5.3.6. To investigate the fractions of inorg.-P that react with Ca, Fe and Al, it was used the of the various inorganic’s procedure based on the differential solubilityfractionationforms in the extracts. The basic assumption of specificity of this procedure is that the
reagents used are able to selectively extract one phase without any solubilisation of
., 1983; et alitations underlined by several authors (Etcheber the others. The limMartinet al., 1987) are basically matrix effects and re-adsorptions. The first occurred
ent in the first fractions is largely reproduced when unsatisfactory recovery of an elemin the other extraction phases, resulting in correlative errors in the last fractions, and
plexation, precipitation, adsorption and loss on the vial the second occurred by com the extraction. walls during

The original fractionation procedure (Chang and Jackson, 1957) was improved by several authors (Williams et al., 1967; Kurmies, 1972; Hieltjes and Lijklema, 1980;
Ruttemberg, 1993). The inorg.-P fractionation in this study was performed according
to the Kurmies (1972) scheme (Fig. 8). This method was utilized in some Brazilian
paio (1998). In this estuaries and specifically in the Amazon estuary by Silva and Sammethod 0.5 g DW sediment was washed three times with 7.5 ml of an alcoholic-KCl
1N solution to eliminate calcium ions. This solution was prepared by the dilution of
74.55 g of dry KCl (110 oC, 2 h) in 100 ml DIW, completed with etilic alcohol to 1 lt
ment from the KCl extraction was used (Morita and Viegas, 1981). The residual sedi -Fe/Al) in two sequential-bound P (Pinium“as is” for the extraction of iron/alum l NaOH 1N and 7.5 mlphases. These phases consisted of the extraction with 7.5 m

Chapter 5 - Methodology

43

Na2SO4 1N, followed by a heating period for 2 hrs at 95oC (only second phase) and
the addition of 2.5 ml H2SO4 1N favouring the flocculation during one hour. The
residual sediment for the second phase was washed with 7.5 ml Na2SO4 4% to remove
xt phase the extraction ofe In this nNaOH excess before the third phase started.calcium-bound P (Ca-P) utilized 7.5 ml H2SO4 1N in a heating period for 2 hrs at
95oC. After the analytical procedure of colorimetry for each phase, the phosphate
P-Fe/Al concentrations of the first and second ined. Theconcentration was determphases were added.

Extractable phosphorus 5.3.7. The amount of tot.-P in the sample that might be readily available for biological use,
termsolution and which is almed in this work as extractable P (extr.-P), is the inorganic forost exclusively orthophosphate (Hesse, 1971). The m occurring in soils
ination of extr.-P, is to shake the soil with a solution to underlying principle for determe extract is analyzed for ts, afterwards thle for plant rooailabdissolve the fraction avthod ece of an adequate extraction msoluble phosphorus. In our study, the choiconsidered the significant correlation between the extr-P and the Australian mangrove
growth (Boto and Wellington, 1983). This method (Hesse, 1957; Hesse 1961),
ents with dim of seed in cooperation with Cleise Cordeiro, extracted 10 g DWperform100 ml of a weakly acid (pH 4.8) 10% w/v sodium acetate in 3% v/v acetic acid the
ant (Allen, 1989). The use of an acid buffer solution known as Morgan’s extractextract soil P does make sense in weakly acid soils, especially in agriculture for which
this meand analyzed by colorimetry.thod was widely used. After shaking for 30 minutes the solution was filtrated

Chapter 5 - Methodology

Figure 8: Sequentia

l extra

ction s

e for quantificahemc

phosphorus reservoirs: iron/alum

bound P (P-Ca).

n of two sedimtio

44

entary

bound P (P-Fe/Al) and calcium-inium

-

Chapter 5 - Methodology

45

In the molybdenum blue method by the colorimetry, the reduction stage by ascorbic
acid is crucial. The overall acidity is one important parameter which must be
controlled, due the color suppression caused in strong acid solutions (Grimshaw et al.,
1989). In spite of this precaution, a sediment extract was spiked to test the possible
interference of the acid, Morgan’s extractant with the molybdenum blue method. The
P) to of a P working solution (50 ppmlspiking solution was prepared by adding 5 m10 ml sediment extracts sub-samples (30.2 ppm P). The added concentrations were
confirmed by determining P concentrations before and after spiking. The result
lated as follows: ) was calcuexpressed as concentration (ppm) = (C1- (C2 –C3)) Concentration (ppmC1 = concentration of non spiked sampleple C2 = concentration of spiked samC3 = concentration of standard

P (-The difference found between the spike and unspiked values results in -0.47 ppm1.55 %). This evaluation exhibited similar error acceptance of concentration as tested
inot and Kérouel (1997). terial by Amafor P references m

5.3.8. ental analyses Elemental analyser Carbon (C) and Nitrogen (N) content were quantified with an elem(Fisons, NA 2100) after Hedges and Stern (1984).

Analysis for total carbon (tot.-C) and nitrogen (tot.-N) not required pre-treatment. In
the procedure, a pair of sediment sub-samples (10-15 mg DW) was weighed into tin
cups and sealed. For the analysis of organic carbon (org.-C) a second pair of sub-samples was prepared by weighing into silver cups, which are more resistant to acid
mples in the open silver cups were acidified with 200 μl HCl attack than tin. The saoC to loss the acid-soluble carbon during 1N and heated in an oven for 24 h at 40 carbonate dissolution (Verardo et al., 1990). After allowed to room temperature the
ped closed. cups were crim

Chapter 5 - Methodology

46

s ples were handled separately, because residual acidied samThe untreated and acidiftes in untreated ples could react with carbonavapour in the HCl treated sam the difference counterparts. Inorganic (carbonate) carbon (inorg.-C) was derived frombetween tot.-C and org.-C and calculated as [8.33 x (tot.-C – org.-C)] (Verardo .,et al1990).

The samples, encapsulated in sample vials, were introduced into a combustion column
pler. The samples were oxidized in oxygen flow by eans of an auto samreactor by mhigh temperature flash combustion (1050 oC). In a helium flow the oxidation products
atographic column, in which C and N oxides were were transported to a chromreduced over copper quantitatively to CO2 and N2, respectively. Detection was carried
ent response was calibrated for each al-conductivity cell. The instrumout with a thermsample batch (n=10) using 10 standards (Acetanilide) and 1-2 empty vials (blanks) to
of the standard was tested after everyear calibration curve. The qualityobtain a linple using a certified soil standard (LECO 1009).fifth sam

Sulphate 5.3.9. Sulphate (SO42-) was measured by Hesse’s (1957) modification of the turbidimetric
method of Chesnin and Yien (1950). This turbidimetric determination of SO42-
sulphate. A 10 g DW involves the precipitation under controlled conditions of barium or extracted with 100 ml acid (pH 4.5) sodium acetate buffer solutionple wassamn. the suspension was iMorgan solution (Allen, 1989). After shaking for 30 mcentrifuged (2500 rpm, 5 min) and filtered through Whatmann No. 41 filter paper into
the extract was transferred into a 25 ml ofletric flask. An aliquot of 10 ma volumgraduate flask, and 1 ml ferric chloride (5%) and 1 ml sodium hydroxide (40%) were
es. The extract was x several timiadded. During the addition the flask was gently mfiltered through Whatmann No. 41 filter paper to remove of not only the colloidal
organic matter, but of Fe (III) and soil color as well, as suggested the modification of
lHesse (1957; 1971). The original flask, precipitate and paper were washed with 7 mDIW, and to the combined filtrate and washings, 1 ml of concentrate glacial acetic
of clear extract was transferred into a 25 mllacid was added. An aliquot of 10 m chloride crystals was added. The flask was graduate flask and 1 g of graded barium

Chapter 5 - Methodology

47

gently mixed during 1 minute and 1 ml of gum acacia solution 0.25% w/v was added.
The solution was diluted to 25 ml and again gently mixed. The turbidity was finally
easured by the decrease in intensity of light that goes through the suspension of the mbarium sulphate with a spectrophotometer at 670 nm using a 1 cm cell. SO42-
concentration in the samples was estimated by comparing the turbidity readings with a
2- standards (10, 25, 50, 75, 100, 250, 500 calibration curve prepared by seven SO42-).mg/l SO4

Percentage of expected concentration 5.3.10. In order to discriminate the influence of the pure mixing of brackish and marine
2-, salinity was considered as a waters from other effect on the variability of SO4eter (biologically and chemconservative paramically non-reactive) and used as a tracer to determine whether decrease in SO42- concentrations were due to either
ater as the only source of ical processes. Considering seawhydrological or biogeochemSO42- in mangrove sediment, the SO42- concentration of 2712 ppm and the standard
salinity 35 o/oo of seawater were used as reference (Libes, 1992). Thus, equation will
provide the percentage of the expected concentration (PEC) at a given salinity under a conservative condition as the ratio of measured/expected concentration (Cohen et al.,
1999).

2-PEC (%) = (Salinity standard x [SO _____________________________ x100 4] sample)
2-ple) ] standard x salinity sam([SO4

Iron5.3.11. The term “reactive” fraction was utilized to describe that fraction of metals in
sediment that are likely to be readily available for participation in chemical reactions
under normal early diagenetic conditions (Huerta-Diaz, 1989). Iron (Fe), which has
is sense can illustrate this concept. been widely studied in th

Different definitions for “reactive” iron (react.-Fe) exist in the literature (Canfield,
have been applied in the past. We1989; Kostka and Luther, 1994), an as such, different techniques for its extraction consider react.-Fe as the fraction of iron, mainly
non-silicate-bound iron, that readily reacts with dissolved sulphide to produce iron

Chapter 5 - Methodology

48

phasis of ., 2003). The emet almonosulfides and eventually pyrite (Roychoudhury

nerals are reactivi(III) mseveral studies has been of which Fee for reduction to Fe(II)

and subsequent iron sulfide precipitation (Berner, 1970; Jorgensen, 1977; Raiswell

and Berner, 1985). Generally, the most “reactive” Fe(III) is described as amorphous

Fe (ferrihydrite or Fe(OH)3) and less crystalline Fe (lepidocrocite). Less “reactive” Fe

Fe (ferrihydrite or Fe(OH)3

is defined as the more crystalline Fe(III) phases (e.g. goethite and hematite) (Canfield

., 1992), in addition to Fe which has et aland Berner, 1987; Canfield, 1989; Luther

ed to pyrite (pyr-Fe) (Poulton and Canfield, 2005).been transform

ed, in cooperation with Helenice Santos, by a ination of Fe was performThe determ

sequential extraction following Huerta-Diaz (1989), a mthod of Lord eodification m

ent Fe fractions are characterized: react.-Fe (1982). Two operationally defined sedim

and pyr-Fe. Both Fe fractions were cold sequentially acid extracted under oxic

conditions from 0.5 g DW of sediment sample. The react.-Fe was leached with 4 ml

HCl 1M and the pyr-Fe was subjected to a series of leaching procedures with 6 ml HF

10 M, 1 g H3BO3 and 2 ml HNO3 conc. (Fig. 9). In the same solution of react.-Fe was

determined the reactive aluminium (react.-Al) and calcium (react.-Ca), also named as

) the extracts for each phase were available or extractable. After dilution (1:500

etry (AAS). ectromic absorption spanalyzed separately by atom

Chapter 5 - Methodology

Figure 9

: Flow diagram of the sequential extraction

Fe) and pyrite-iron (pyr.-Fe).

49

e for the reactive- (react.schem

-

Chapter 5 - Methodology

Degree of pyritization 5.3.12.

The initial redox proxy based on Fe speciation was the degree of pyritization (DOP)

ed in this work as DOP and defined as: (Berner, 1970), nam

Pyrite-Fe
__________________ x 100 DOP (%) = Pyrite-Fe + Reactive-Fe

Where Pyrite-Fe was defined by Berner as:

Pyrite-Fe = 1/2 (TRS-AVS)

TRS = Total Reduced Sulphide

hide AVS = Acid Volatil Sulp

50

ents where the AVS was present as a significant fraction of TRS Afterwards, in sedim

ovide a good accounting of Fe availability during early and the DOP approach not pr

mdiagenesis, the degree of sulfidation (DOS) was proposed (Boesen and Postn, a

ed according to Berner (1970). This orm, was perf1988). The redox-proxy in this study

e Brazilian estuaries of the Sepetiba Bay by Bastos (2002) utilized.thod was in somem

The DOP is a m-Fe with of the reaction of the react.mpletenesseasure of the co

aqueous sulfide (Leventhal and Taylor, 1990). Normal marine well-oxygenated

bottoments covered by water give a DOP value generally less than 0.4 on the sedim

nts (also called dysaerobic or dysoxic – very low ethis water, whereas suboxic sedim

O2 to no O2, but no H2S) implied DOP values around 0.5 to 0.7, and euxinic

sediments (containing H2S in place to O2- term derived from the Black Sea whose

ancient Latin name was Pontus Euxinius) refers to DOP values greater than 0.7

suggesting anaerobic depositional conditions (Raiswell et al., 1988). Under euxinic

inerals can occur both before and after S with iron-mconditions the reaction of H2

itself (Leventhal, 1983). entation burial, even during sedim

Chapter 5 - Methodology

51

Pyrite 5.3.13. 6 cores (1 m ent samples fromcroscopy work, three disaggregated sedimiFor the mdepth), sampled for both tree species in each forest type, were mounted in aluminum
stubs and gold coated. These impregnated samples were examined using a standard
method of Scanning Electron Microscopy -SEM (LEO 1450VP) in the Goeldi
Museum, State of Pará (Goldstein et al., 2003). The diameter size and number of
microcrystals of pyrite and framboidal morphology observed in the sample were
measured using an optical microscope. The grains were measured to ± 0.25 μm with a
standard ocular calibrated with a stage micrometer.

e, spherical to sub-spherical aggregates -packed, raspberry-likboids are denselyFramstutz, 1966; Rickard, 1970). They cro-sized pyrite crystals (Love and Amiof equal mrange from a few to several tens of microns in diameter (Wilkin et al., 1996), with
ramboidal clusters or aggregates” boids” (Love, 1971) or “frare “polyfram(Sawáowicz, 1987) occasionally reaching several microns to millimeters. The
boidal morphology, are generally thought to processes causing pyrite to have framoccur during the replacement of progressively more S-rich phases: disordered
mackinawite (FeS(1-x))ĺ ordered mackinawite (Fe9S8)ĺ greigite (Fe3S4)ĺ pyrite
(FeS2). Within individual framboids, microcrystals are remarkably uniform in size and
shape, therefore the diameter is independent of orientation (Wilkin et al., 1996). The
number of microcrystals (Nm) in spherical and oblate (dumbbell) framboids
composed of uniformly sized spherical microcrystals in a cubic closest-packed (ccp)
arrangement was calculated. This equation considered the framboid diameter (D),
microcrystal diameter (d) and packing coefficient of the microcrystals (ø = 0.74)
., 1996) as follows: et alilkin (W

3 = ø (D/d)Nm

The notion of pyrite stability has resulted in the wide use of pyrite formation as a
redox proxy for identifying depositional environments and a cyclic transition
(Raiswell and Berner, 1985; Raiswell et al., 1988, Wilkin et al., 1997).

Chapter 5 - Methodology

52

In fact, very slow deposition under euxinic conditions increase the amount of pyrite

formed, because slowly reacting Fe compounds are given more time for reaction with

H2S (Berner, 1984). An important distinction has been made by Raiswell and Berner

(1985) between syngenetic pyrite formed in the water columns of euxinic

environments that settles to the sediments-water interface prior to burial and

diagenetic pyrite formed in situ within the pore-waters of anoxic marine sediments

diagenetic and syngeneticns. This distinction between underlying oxic water colum

sedimentary pyrite affects the interpretation of the DOP observed in ancient sediments

., 1988).et al(Raiswell

5.3.14. Molar ratios

-Fe, react.-Al, react.eters org.-C, tot.-N, react.The param-Ca, P-Fe/Al and P-Ca were

expressed in weight % and their ratios were calculated on w/w basis. The latter can be

obtained multiplying these parameters (w/w) values by their moles (mol/mol) values.

This calculation was applar ratios: oe following mlied for th

-

-

-

-

C : N

Fe : P-Fe/Al

Al : P-Fe/Al

Ca : P-Ca

= organic carbon versus total nitrogen

bound Phosphorus -iniume versus iron/alum= reactive-F

inium= reactive-Al versus iron/alumbound Phosphorus -

-bound Phosphorus = reactive-Ca versus calcium

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

CHAPTER 6

53

AND PHOSPHORUS STATUS CAL PATTERN PHYSICOCHEMI

OF MANGROVES

6.1. Results

6.1.1. Topography and forest structure

es could be identified: high in topographic featuraace, three mAt the elevation surf

ediate (MP) and low plain (LP), where the hidro-edaphic conditions of(HP), interm

Fig. 4topography and flooding showed a clear gradient (). The HP was defined into

the boundary of 41-67 days/year and was inundated only during spring tide. The MP

had a boundary of 80-101 days/year, whereas LP 124-162 days/year. At the

pth occurred, demboundaries of HP and LP with MP, slight depressions of 5-8 c

identifains of a paleochannel appearing to be remied as depression one (D1) and two

the depressions do not correspond to . IF values from(Tab. 3)(D2), respectively

e of the flood water, anency timhigher frequencies, they indicate higher perm

moreover they were into the trend.

At the study area, situated at a higher part of the central sector of the Bragança

terized by a tree-height gradient including Peninsula, the forest structure is charac

monospecific stands of A. germinans and R. mangle divided by a transition zone of

both species. At the drier and more saline zone A. germinans dominated, belonged to

eter at breast height diamallest stemthe sm (DBH) (1-16 cm) and height size (4.5-10.5

cm) classes. Conversely, in the most frequently inundated area R. mangle trees with

highest DBH (9-19 cm) and height size (7-13 cm) classes occurred (Menezes, 2006).

topographic niaSuch pattern of species zonation and forest structure along three m

orphic changes and physiological features can been attributed as a response to geom

aintained gradients (Snedaker, 1982).response to tide-m

ical Physicochem 6– Chapter

pattern

and phosphorus status of mangroves

54

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

55

6.1.2. During tide stability conditions thEh, sulphide and pH: Changes with flooding time values of fixed electrode es in sediments (10 cm)
) were taking as an indicator of the e period (24 h- and 12 h-expeditionsalong a timmeasurements variation. This instrumental variability, expressed as variation
10 % (0.80 ± 0.08 μM), pH 2.5 % (7.4 ± 0.19) ecoefficient (VC), results for sulphidand Eh 13 % (70 ± 9 mV) (Tab. 1,Fig. 10 (I), Fig. 11). The VC of sulphidedidnot
consider the initial measure period (14:45-17:25 h) due adaptation of the electrode
signal to lower sediment sulphide concentrations (Fig. 10 (I) A). The mean values of
the fixed stations were calibrated with the downcore electrode dates performed in the
sediment (10 cm) at the same station (HP) one month latter (Fig. 17). The comparison
ilar ), pH (7.7) and Eh (81 mV) resulted in simwith sulphide (1 μM, data not showvalues suggesting reliable measurements in the profiles of the entire transect.

ical ergence on physicochemergence and non submThe overall effects of subm24 h-expedition are shown in Figures 10 and ent and water for the eters of sedimparamooth sulphidements, a sergence of sedim11. During flood, but before the submincrease was seen, which kept stable for 12 min (Fig. 10 (I) A, C). At the same
ergence (20:15-20:45 and 8:30-8:50 hrs), a bmue of ssectors, during the course timlargest sulphidedecrease occurred. This ~3-fold sulphidedepletion was sharper at
higher (C, 23 cm) than at lower (A, 8 cm) water height, and was characterized by an
initial drop of 0.87-0.59 μM (A) and 0.97-0.44 μM (C) followed by a gradual and 0.44-0.28 μM (C). During ebb a gradual sinusoidal depletion of 0.59-0.31 μM (A) ing to steady state (B, C). crease was seen returninsulphide

At the 12 h-expedition, the Eh and pH were characterized by stable values during non submergence (9 hrs) (Fig. 11 B, C). The submergence of sediments (20:00-21:00 hrs)
reaching 34 cm water height caused an intense opposite change of these parameters
itant with the pH increase (4 units), (C). The Eh depletion (194 mV) was concomshowing a relation of ~50 mV per pH unit, maintained also during ebb conditions
(Fig. 11 D, Fig.12). During the 2.5 hours of submergence (20:00-22:30), the Eh
dropped drastically ~4.5-fold from 78 mV to -253 mV, concomitant with the pH
height of 2 cm, the returning trend to 6.8 to 13.3 units. In the water increased fromstable values occurred.

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

56

e influence of the The pH and Eh of floodwater was controlled to check the possibl

ents and the percolation of floodwater s in the sedimelectrode tip causing fissure

during submeters showed a smooth opposite (II)). The water paramFig. 10ergence (

intained a relation of ~60 mV per pH change of ~0.4 unit and stable values afeature m

the presence of slightly , D). During inundation s) (Cduring ebb begin (9:02-9:16 hr

ent pattern ) floodwater affecting the sedimVbasic pH (7.7-8.6) and oxic (195-71 m

ergence the pH and Eh values of ). At the end of submFig. 12was not evident (

floodwater showed a tendency to reach the sediment stable values (Fig. 12 (I)(II)).

intained during aent was mThe relation of 0.5 volts per 1 pH unit in sedim

(III)) supporting the experimFig. 12ergence (submental data of Gotoh and Patrick

ez et al. (1999). (1974) and Gom

ngrove roots aEh, sulphide and pH: Changes relative to m6.1.3.

In the comparison of two areas with different inundation frequencies, a higher

sediment Eh and lower sulphide concentrations occurred in the surface sediments (10

cm) of the monospecific A. germinans forest (HP) (Fig. 13,Fig.4B, C). Eh ranged

2- concentrations ranged 0.4-1.0 and 0.4-3.7reas S12 to 349 and 260 to 298 mV, whe

μM, both for HP and MP respectively. Topographic elevation and inundation

and 60 days/year, frequency between stations locations B and C differ 20 cm

respectively (Fig. 4). At MP, in the bare substrate at 0.5-1.5 m distance of the

ent pH ranged vegetated zone, sulphide values (2.1-3.7 μM) were highest. The sedim

spectively. Non significant differences in Eh and e6.6-7.0 and 6.2-6.5 in HP and MP, r

nt (BS) and vegetated area. empH were found between bare sedi

pattern ical Physicochem 6– Chapter

Figure 10:

(I)and phosphorus status of mangroves

Changes for sulphide in sedim

ents at 10 cm depth (27.6

57

oC) and water

height (cmpotential (E) variation during flood and ebb. h) of flood water surface and depth (25.5(II) o Changes in pH and redox C) during spring

potential (E

tide.

h) of flood water surface and depth (25.5

Chapterpattern ical Physicochem 6–

Figure 11

and phosphorus status of mangroves

: Variation in water height (cm o depth (26.5at 10 cmC) during spring tide.

depth (26.5at 10 cm

), pH and redox potential (Eh) in sedim

C) during spring tide.

58

ents

Physicochem 6– Chapterpattern ical

Figure 12:

Feature of E

and phosphorus status of mangroves

versus pHh

sediment during flood

values for flood water

(III)and submergence I) (I

(I)

59

(24 h-expedition) and

). edition (12 h-exp

Figure 13Station B

Station

xed forest of i m

and

ddle plain (MP). i at m

ide, pH and redox potential

m) sediment sulph

c

inated ngrove forest doma m

by

ent (BS) at high plain (HP), and and bare sedim

(Eh) along a horizontal profile in:

: Variation of surface (10

60

and phosphorus status of mangroves

pattern ical Physicochem 6– Chapter

R. mangleA. germinansC,A. germinansStation B,

ical Physicochem 6– Chapter

Station C

pattern

Continuation Figure 13

and phosphorus status of mangroves

61

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

62

) in sediments Eh and pH changes with depth (50 cm6.1.4. ent Eh (-121 to 296 mV) and pH (6.4 to 8.0) Along the transect, changes in sedimexhibited a wide range in the profile (50 cm). This vertical variation in Eh and pH was
, vegetation zonation and depth. In severalstrongly associated with the different IFodel proposed by ilar to the redox stratification mstations, the Eh zonation was simoxidation anClark et al. (1998) for an d two reduction horizons were propoAvicennia-dominated mased, each accomngrove forest. In this mpanied by acidodel two ic and
).Fig. 14basic pore-water pH, respectively (

In HP, the upper horizon of sediments (5 cm) named by Clark et al. (1998) as “upper
oxidation zone - UOZ”, was characterized by oxic conditions in R. mangle (262 mV)
andA. germinans (106 mV). Beneath the oxreduction zone - URZ” was characterized by lower Eh conditions in idized surface layer (10 cmR. mangle), the “upper (207
mV) and A. germinans (82 mV) in relation to the UOZ. At the “lower oxidation zone -
LOZ” (10-35 cm) an Eh increase in R. mangle (ǻEh1=89 mV) and A. germinans
(ǻEh2= 139 mV) was recorded. Beneath the roots zone (40-50 cm) merge the “lower
Eh depletion and/or stabilization. Z” characterized by reduction zone - LRy be appreciable two different pH features. The first pH feature aAdditionally, there model. At URZ was consistent with the redox stratification mA. germinansobserved at basic (7.8) to neutral (7.0) values, at LRZ the slightly and LOZ, this trend ranged fromR.curred below clished. The second pH feature obasic (7.3) conditions were re-estab by pH depletion (6.9-6.4).mangle

species with a mV) in bothe (57-184ediate plain (MP), Eh kept positivAt the intermmaximum localized under R. mangle trees. Below R. mangle,a moderate intensity of
the redox stratification was maintained with Eh increasing at LOZ (ǻEh5=50 mV,
ǻEh7=125 mV). Beneath A. germinans non specific Eh pattern was appreciable
characterized by low Eh variation at LOZ (ǻEh6=22 mV, ǻEh8=38 mV). Concerning
ilar to the rked decrease (7.8-6.4) was evidenced, simapH profiles, in both species a msecond pH pattern described in HP.

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

63

A.At the depressions (D1, D2) an intense redox zoning was established, except by e lower Eh beneath terized by extrem in D1. These depressions were characgerminansR. mangle (-110 mV) and A. germinans (-79 mV), in D1 and D2, respectively. Here,
at LOZ, the Eh increase accounted moderate variation for R. mangle (ǻEh3= 76 mV,
ǻEh9= 59 mV) and A. germinans (ǻEh4= 25 mV, ǻEh10= 65 mV). The redox profiles
belowA. germinans (D1) and R. mangle (D2), showed a peculiar stable Eh values at
depth, respectively. The pH was recorded only at D1, showing a 7-25 and 15-30 cmA. and 7.7-7.1 for R. mangledecrease feature (second pH pattern) ranged 7.3-6.7 for .germinans

At the low plain (LP), R. mangle was localized in more oxidized sediments than A.
germinans similar as in D2. Below R. mangle, a peculiar stable Eh for 7-13 cm depth
followed by a shorter Eh increase (ǻEh11=66 mV) was observed. This trend was
, a clear A. germinans6.8) values. In contrast, beneathooth acidic (6.6-associated to smredox stratification was exhibited, accounted by the minimum Eh value (-121 mV)
and the highest Eh increase at LOZ (ǻEh12=193 mV) of the transect. This feature was
characterized by pH of 7.4-6.9.

ation with respect ong species are of interest since they provide informDifferences am (LOZ), an intense A. germinansto their relative tolerance to flooding stress. Belowroot-induced oxidation was observed from oxic (HP) (ǻEh2= 139 mV) to anoxic
sediments (LP) (ǻEh12=193 mV). Opposite to A. germinans, beneath R. mangle
(LOZ), the root-induced oxidation was attenuated from oxic (HP) (ǻEh1=89 mV) to
anoxic conditions (LP) (ǻEh11=66 mV). At medium plain (MP) A. germinans did not
showed any oxidizing power. These data suggest that low and high flooding would .R. mangle than in A. germinanslead to a higher root-induced oxidation in

pattern ical Physicochem 6– Chapter

Figure 14:

and phosphorus status of mangroves

Vertical distribution of redox potential (m

V) and pH below

64

A. germinans

i. HP, high plain; MP, mR. mangleandddle plain; D, depressions; LP, low

e axis. plain. The figures share the sam

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

65

)ents (1 mPhosphorus fractionation in sedim6.1.5. The analyzed tot.-P was compared with that calculated from the sum of org.-P + P-
Fe/Al + P-Ca + extr.-P. Similarly the analysed and calculated inorganic P (P-Fe/Al +
ed fractions were considered as 100 %. pared. The analyzP-Ca + extr.-P) were comparisons indicated that the calculated P fractions were 76 % and 74 % These comlower than the analyzed tot.-P and inorg.-P, respectively (Fig. 15). The relationship
between the analysed and calculated parameters showed highly significant
(r=0.90, n=93, p<0.005) and inorg.-P (r=0.88, n=93, p<0.005). correlations for tot.-Ppounds by each specific of other P comThese results suggested non dissolution atic variation. These ed to a low and systemthodology, summeanalytical mparisons indicated the convenience of using the analysed tot.-P for the general com the calculated fractions only for describe the inorg.-P results description andproportions.

depth (LOZ) are summarized mValues of tot.-P and fractionation between 10 and 40 cinTable 3. Along the transect, tot.-P levels in sediments decreased with increasing
e fraction of P in en though tot.-P is not representative of the reactivvelevation. Epared to existing data based on ponent that can be coments, it is usually the comsediminations. ical determless detailed geochem

e tot.-P. Thre than 80% tooThe percentage of analysed inorg.-P contributed mconcentrations of inorg.-P increased markedly from HP to LP (0.21-1.28 mg P g-1
trend as tot.-P. The concentrations of org.-P in HP and LP e), following the samDWwere in the same range (0.01-0.08 mg P g-1 DW), with exception by the increase at the
depressions (0.03-0.12 mg P g-1 DW) resulting in a proportion for 12-17% of tot.-P.

pattern ical Physicochem 6– Chapter

Figure 15and phosphorus status of mangroves

parison between analyzed and calc: Com

). ) and inorganic P (inorg.-PP (tot.-P

ulated phosphorus fr

66

actions of total

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

67

The dominant mineral fraction was the iron/aluminium bound phosphate (P-Fe/Al)
-1 DW. At LP, this fraction accounted for the higher ranging for 0.11-1.12 mg P g proportion (81 %) of the calculated inorg.-P. At both depressions, the depletion of this

ost enriched with org.-P. The ents mpool (67-70 %) was coinciding with sedim

bound phosphate (P-Ca), with a concentration range in one order of calciummagnitude lower than those of P-Fe/Al (0.01-0.12 mg P g-1 DW), accounted for 6-

12% of the calculated inorg.-P, showing the highest proportions at LP.

Along the transect, the levels of extr.-P varied for 0.02-0.19 mg P g-1 DW, with
).Tab. 3highest proportions (18-27 %) of the calculated inorg.-P at the depressions (

).Fig. 16The extr.-P showed a high dependency with position along a tidal gradient (

nd inundation frequency was best described ) aThe relation between extr.-P (25-30 cm

by a positive correlation (r=0.81, n=29, p<0.005).

Chemical analysis of surface sediments (5-10 cm) yielded means of 0.04 ± 0.03 extr.-
P, 24 ± 1 react.-Fe, 13 ± 1 react.-Al and 0.19 ± 0.03 mg. g-1 DW react.-Ca (Tab. 4).
HP, contrasting to Eh’s (185-262 mV) inThe surficial layer was characterized by high

derately acidic to basic pH o) in LP. In both forest a mVlower Eh’s (-121 to 136 m

avalues (6.5-7.5) was mintained. Marked differences in extr.-P and react.-Ca between

aerobic (HP) and anaerobic (LP) sediments were observed. LP sediments have more

extr.-P and react.-Ca in solution than HP sediments. The react.-Al was constant and

the react.-Fe do not showed well-defined horizontal pattern. Concerning the mi-1nerals
DW) P-Fe/Al concentration (0.49 ± 0.10 mg. gpools, LP accounted for the highest opposite to HP (0.30 ± 0.05 mg. g-1 DW). Regarding P-Ca, the concentrations along
the transect ranged 0.02-0.14 mg. g-1 DW, increasing (0.09 ± 0.03 mg. g-1 DW) in LP.

pattern ical Physicochem 6– Chapter

Figure 16

and phosphorus status of mangroves

: Horizontal distribution of inundation frequency (I

68

F), extractable

ents and leaf phosphorus (Leaf-P) (Cordeiro, phosphorus (extr.-P) in sedim

Mendoza

ission). ., 2003 reprinted with permet al

ical Physicochem 6– Chapter

pattern

and phosphorus status of mangroves

69

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

70

a -C-Al : P-Fe/Al (Al : P-Fe/Al) and react.-Fe : P-Fe/Al (Fe : P-Fe/Al), react.The react.: P-Ca (Ca : P-Ca) ratios gave the highest values in HP (Tab. 4), reflecting in general
at correlated HP. The only ratio ththe lower P-Fe/Al and P-Ca concentrations insignificantly with extr.-P was Al : P-Fe/Al which was negatively related (r= -0.74, showed r values of -0.63 (n=24, p<0.05), ent Fe : P-Fe/Al ratio n=23, p<0.05). Sedimwhich were close to significant, while Ca : P-Ca ratio showed poor correlation with-0.49 (n=24, p<0.05).

ents interaction 6.1.6. Porewater and solid-phase elempounds varied significantly among forest types and was The concentration of P com. Along the (Fig. 17)ent depthical variability with sedimaffected by the physicocheme variation pattern of tot.-P. In general, the sect, P-Fe/Al showed the samentire tranextr.-P pattern varied as a function of dissolution and precipitation for P-Fe/Al and P-
Ca pools (Saleque and Kirk, 1995; Wang et al., 1995). Due the low proportion of
lation to inorg.-P (> 83 % of tot.-P), the organic fraction ) in re % of tot.-Porg.-P (<17pressions were did not seem to play a key role on the extr.-P, with exception of the de. ulation occurredterial accumamore m

a short decline at URZ,At HP the trend of tot.-P and P-Fe/Al was characterized byR.ooth increase at LOZ and a substantial depletion at LRZ. Below followed by a sm the P-Ca concentrations reflected a greater depletion with depth (0.08-0.03 mg mangleP g-1 DW). Contrary, beneath A. germinans, occurred at LOZ a P-Ca increase (0.02-
0.08 mg P g-1 DW) and at LRZ a decrease (0.08-0.03 mg P g-1 DW). In both species,
the extr.-P (0.01-0.07 mg P g-1 DW) was controlled by the dissolution and
R. mangle precipitation for the minerals pools. The org-P concentration belowdecreased sharply (0.08-0.01 mg P g-1 DW) and under A. germinans remained stable
(~0.03 mg P g-1 DW).

In the intermediate plain (MP), the tot.-P increased at LOZ (0.35-0.58 mg P g-1 DW)
and depleted at LRZ (0.58-0.17 mg P g-1 DW), the extr.-P follows the same pattern
(data not showed). The lower tot.-P range may reflect the moderately and slightly
oxidation intensity in R. mangle and A. germinans, respectively (Fig. 14). In both

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

71

e org.-P sharpspecies th depleted with depth. This forest was lacking of P-Fe/Al and P-Ca data.

ediate plain, at the waterlogged areas, the extr.-P, P-Fe/Al Contrasting to the interments re anoxic or oxidized sedimoand tot.-P showed two different features, linked to m(Fig. 14). The first trend, beneath R. mangle (D1) and A. germinans (D2), was linked
to the redox stratification, reaching anoxic values in LRZ (Fig. 14). This trend was
by higher extr.-P, P-Fe/Al and tot.-P ilar as the HP pattern, differing heresimconcentrations (Fig. 17). In the second feature, beneath A. germinans (D1) and R.
mangle (D2), the extr.-P increased concomitant with P-Fe/Al dissolution and tot.-P
pool depletion. Here beneath the dominant species of each forest side, the extr.-P
accounted the maximal levels of all transect (0.20 mg P g-1 DW). Although the Fe/Al
dissolution occurred in oxidized sediments, seem to be linked to the remarkably stable
Eh values with depth (Fig. 10). In both cases the P-Ca rem-1ained stable and the org.-P
). DW ggcontent was higher than average (0.02-0.13 m

ilar to D2,At low plain (LP), the features of P-Fe/Al, tot.-P and extr.-P were simcharacterized below R. mangle by dissolution of Fe/Al-P (0.60-0.23 mg P g-1 DW)
and tot.-P depletion (0.78-0.42 mg P g-1 DW); contrasting to higher P-Fe/Al (0.21-1.1
mg P g-1 DW) and tot.-P concentrations (0.32-1.6 mg P g-1 DW) beneath A.
germinans. Here again the physicochemical parameters seem to play a key role in R.
mangle by stable Eh influencing the P-Fe/Al dissolution and in A. germinans by a
by roots) affecting the P-Fe/Al greater biological influence (i.e. oxidationprecipitation (Fig. 14). Additionally, at the uppermost sediments (< 25 cm), the P-Ca
pool accounted the highest concentrations of the entire transect in R. mangle (0.05-
0.12 mg P g-1 DW) and A. germinans (0.03-0.10 mg P g-1 DW). In this plain the org.-
P did not show any specific pattern.

pattern ical Physicochem 6– Chapter

Figure 17

and phosphorus status of mangroves

: Vertical distribution of phosphorus com

pounds. The figures share the

72

-bound P; extr.-P, extractable P; org.-P, organic e axis. P-Ca, calciumsam

-bound P; tot.-P, total P; HP, high plain; MPiniumP; P-Fe/Al, iron/alum

imddle plain; D, depressions; LP, low plain.

,

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

73

Discussion6.2. Effect of flooding conditions on P-exchange 6.2.1. ent changes in ergence the sedimIn the 24- and 12 h-expeditions, during submsulphide, Eh and pH shows a general pattern with rapid initial changes, followed by ergence continuing more gradual changes, returning to stable values during non subm).Fig. 10, 11(

ility before the stabe-confined sulphideIn the 24 h-expedition, after a timsubmergence, the sulphiderapid initial drop and gradual depletion (20-30 min.) in the
floodwater, or b) the consumearly stage of the submergence couption of SO42-ld be apparently due to a) sulphide during the degradation of the organic dilution by the
matter (Fig. 10 (I) A, C). Unfortunately, the control of the salinity and Cl-/ SO42-
ents could not be evaluated, so the absence of these data molar ratio data in the sedimxing of inot supported this hypothesis. This hypothesis can be supported by the m(Caetanofreshly tidet alal water in the interstices of., 1997). Meanwhile, the gradual sulphide depletion was sim sediment surface (2 cm) with sea-water ilar to the
theoretical H2S decrease as the result of consumption of SO42- (Postma and Jakobsen,
(I) B, D), considering the tidal Fig. 10 ergence (1996; Kehew, 2001). During subm2-ents, the sinusoidal sulphide in these sedimsupply seawater as the source of SO4increase over time, reflects the reduction of the electron acceptor SO42- resulting in a
2006).gradual production of sulphide in the porewater (Equation 9, Tribovillard et al.,

(CH2O)106(NH3)16(H3PO4) + 53SO42-oH3PO4 + 53S2- + 16NH3 + 106CO2 +:106H2O
Eq. 9

In the 12 h-expedition, the shift to anoxic conditions during sediment submergence
reaching a minimum of -253 mV is consistent with recent studies in lowland rice
fields in Philippines (Kirk, 2003) (Fig. 11, D). Meanwhile, our values are extreme
reduced as compared to laboratory studies (240-60 mV) performed at 10 cm depth
with 16 cm of water height above the sediment surface (Ellison and Farnsworth,
1997). In the Bragança sediments, the Eh minimum was reached in a shorter time (2.5

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

74

hrs) than in rice fields (few days), suggesting a difference on the “intensity” factor duction (Kludze and DeLaune, 1995), caused ined by the relative ease of the redetermng the transect, at HP where the 12-h Alondition.probably due the water-saturated cothan at LP for the samexpedition was performed, the sedime depth (60 %). Thus, we expect thent humidity at 10 cm depth waat the “intens lower (38 %) sity” factor of
the reduction in LP will be higher, due the high inundation frequency (146 ± 13 days/year) resulting in longer time of minimum Eh following each inundation.

e the Eh decrease is consistent with the ergence expedition, during the submIn the samchanges in pH attained a maximum of 13.3 units after 2.5 h (Fig. 11 C, D). In fact, a
pH of 13.3 not occurred in natural systems. In the next section is discussed the
ation about this topic. available inform

d is responsible for the es protons anate reduction consumThe Fe(III) and sulpherved in rice enon is commonly obs). This phenomEquation 3 and 9increase in pH (and normal acid sulphate soils after submergence (Konsten et al., 1994;
Ponnamperuma, 1972; Kirk, 2004). However, these experiments were conducted over
ergence (several weeks), under such conditions an initially low pH prolonged submwill increase converging between 6.5 to 7 at steady state (Pavanasasivam and Axley,
ent was 1980). In the current study, the relation of 50 mV per 1 pH unit in sedimmaintained during submergence supporting the experimental data of Gotoh and
e . (1999). The sharp Eh reduction and pH increaset alez Patrick (1974) and Gomcould be due the high concentrations of Fe/Al-P (0.35 ± 0.09 mg.g-1) and SO42- (17.7
± 3.2 mM), summed to the organic matter (12.9 %, 0-10 cm) accumulation due the
topographical high position of this station (HP).

Under anoxic conditions, Gotoh and Patrick (1974), and Gomez et al. (1999) proposed that the redox limit for the reduction of Fe(III) (hydr)oxides to Fe(II) change with pH:
between 100 and -50 mV at pH 7, -100 mV at pH 8 and -150 mV at pH 9. Under this
e reduction of Fe(III) apparently take wide range of pH-redox potential conditions theen zones of Fe(III) ologically active waterlogged soil. The interface betwiplace in a b(hydr)oxides and sulphate reduction is rather poorly defined. From a thermodynamic

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

75

an sulphate reduction, the gher Eh thipoint of view, iron reduction takes place at a htwo reactions may proceed simultaneously over a wide range of environmental
ining whether Fe(III) (hydr)oxides or inant factors determconditions. The domsulphate reduction is energetically most favourable, are the stability of the Fe(III) all (hydr)oxides and the pH, while the effect of the sulphate concentration is sm(Postmwritten as (Equation 10a and Jakobsen, 1996). An equilibrium, Appelo and Postma, 2005): reaction for the two couples can be

8Fe2+ + SO42- + 20H2Oļ 8Fe(OH)3 + HS- + 15H+ Eq. 10

Hents) and ~60 mV (water) per pergence the relations ~50 mV (sedimDuring submunit were maintained (Fig. 11 C, D; Fig. 12). The first relation agreed with that
ld ents of fieproposed by Gotoh and Patrick (1974) and Gomez et al. (1999) for sedim a batch reactor, rice and coastal lagoons, under oxic and anoxic conditions in olation of floodwater troughon of our results by the percatirespectively. The overestimderate, due to the local (HP) food ocrabs burrows and bioturbation should be m., 2006). et alitation reflected in low crab density in the study area (Nordhaus limClearly, the effect of flooding on these trends is irrefutable. Meanwhile, the presence ents. easuremtensity of the mplify the inof the electrode can am

ent properties Effect of the species on sedim6.2.2. A.parison at equivalent elevations, of the root-induced oxidation capacity of The comgerminans pneumatophores alone and in combination with the stilt roots of R. mangle,
ents with one species evidenced a greater oxidation effect on redox status of the sedim(Fig. 13). A similar study performed by McKee et al. (1988) described spatial patterns
of Eh contrary to these reported here, characterized by greater redox status on ght explain theients with two species. The distance between species msedim and in ) 1 mdifferential patterns, in McKee et al. (1988) was over short distances (. These results indicated that the )-22 mthe present study was over longer distances (5ntal design does not seem occurred. esuperposition of roots in our experim easured here were higher than those reported for McKeeNevertheless, the Eh values mas (Nickerson and Thibodeau, 1985) for et al. (1988) and for Twin Cays in the Baham

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

76

pth in me dethe samngrove substrates. The range of values obtained for sulphide a concentrations (0.4-3.7 μM) were considerably lower. The reason for the lack ofagreement between these studies with respect to S2- is the different sulphide species
analyzed, total soluble sulphide (H2S, HS-, S2-) and H2S values, respectively.

forest Avicenniant plains, the finding beneath Across the topographically differeents exposed to ents, reflects well-drained sedimstands (HP) of less reducing sedimlow IF, because the submergence results in a reduced state (Ponnamperuma, 1972)
in the topographic gradient (e.g. ). At the small-scale discontinuities(Fig. 14noespecific stands, the odepressions), coincident with the transition between me. the increase of water residence timely evident duehydrological effect was extremilarly to LP, where due to higher IF the water remSimains constant.

icant greater ents (LP) a signifduced sedimealed for intense reThe present findings revroot-induced oxidation capacity in A. germinans (ǻEh12=193 mV) (Fig. 14). This
ckerson and Thibodeau de by Niaobservations mresult is in accordance with previous (1985), McKee et al. (1988), Andersen and Kristensen (1988) and Alongi et al. ents (HP) the species-specific oxidation was also (2000). For low reduced sedimhigher in Avicennia (ǻEh2= 139 mV) (Fig. 14). For instance, Kludze and DeLaune
s tion between root radial oxidation los(1995) reported for rice field a high correlaent (ROL) rates and sediment Eh intensity, where a higher ROL in reduced sedim ROL was proposed by Scholander et al. (1955) and defined by e termhoccurred. TColmer et al. (1998), Armstrong et al. (2000), and De Simone et al. (2002). Other
es ngrove wetland speciastudies in mangrove (Youssef and Saenger, 1996) and non-mtrong, 1987; Koncalová, 1990), observed a little effect of the degree sm(Justin and Are ent. In a seedling staga developmy or aerenchymof substrate aeration on root anatomAvicennia roots are more porous than Rhizophora, this root porosity (POR) is related
to the air space in the aerenchyma. The changes in POR and ROL in A. germinans can
ents (Youssef ide sedimlphpensate for substrate changes associated with acid-sucom along the R. manglewer but constant oxidation of and Saenger, 1996). However the lo ),Fig. 14transect was only weakly affected by the different substrate conditions (

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

consistent with data on this genus reported by McKee (1996), Youssef and Saenger

(1996), and Ellison and Farnsworth (1997).

77

Redox stratification and pH pattern 6.2.3. del overtical profiles agree with the stratification mRedox potential values of the ). The thickness of the UOZ, depend upon the Fig. 14described by Clark et al. (1998) (balance between the rate of O2 diffusion into the soil and the rate of O2consumption
(Howeler and Bouldin, 1971). Soils with high OM content tend to have a high crobial respiration and consequently a thin UOZ (House, 2003). At the study area, imthe HP was characterized by an UOZ thickness of ~5 cm, as indicative of sediments
with low biological heterotrophic activity. At MP and LP, the UOZ thickness depleted ), suggesting a greater biological heterotrophic activity, principally at D1 and (1-3 cmD2 (~1 cm) coincident with the second vertical P trend (Fig. 17). Below UOZ, the

URZ was restricted to a thin layer ( 10 cm depth). A deeper URZ ( 15 cm depth)
was reported by Clark et al. (1998) also during rainy season. At Bragança, the LOZ merge for 10-35 cm depth, suggesting a maximum root activity and the outward
ngrove below-ground root adiffusion of oxygen. This LOZ was lower than other m

systems extended to 50 cm depth (Komiyama et al., 1987; Alongi et al., 2004).

rying as a function of the revealed two patterns, one vaThe vertical pH distribution

redox stratifdistribution centered around 6.7 (ication mFig. 14odel and a second displaying one monotonous smooth acid ). In the first pattern, the tidal supply

tes to an increase in pH (pH > s at UOZ and URZ contribuseawater of basic cation7.0). At LOZ the pH fall (pH = 7.0) exhibited a link with the redox increase. This pH 2++ generated by oxidation of Fedepletion trend is probably the result of 1) H(Equation 1, Beeg et al., 1994) and/or FeS2 (Equation 7a, 7b, Kirk, 2004) by root
released O2, and/or 2) H+ released from the roots to balance excess intake of cations
to over anions (Haynes, 1990). Concerning the second pattern, the acidic values seembe derived from both organic decay and sulphur oxidation (Marchand et al., 2004).
ann (1998) for pore-range (4.5-7.5) reported by SchwendenmThe pH values are in the

water surface sediments at Furo do Chato creek, Bragança Peninsula. Additionally,
ents ents indicated a seasonal shift with sedimngroves sedimaprevious studies of m

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

78

being more acidic in the wet season (Alongi et al., 1999; Marchand et al., 2004;
., 2004).let aAlongi

Inundation frequency and availability of phosphorus 6.2.4. ents is directlyents that the phosphorus status at the sedimThis study clearly documcorrelated to the IF. Across the three floodplains, the decrease pattern of tot.-P with increasing topographic elevation (Tab. 3), was similar to the pattern found for the
ngroves studied by Boto and Wellington (1984). The tot.-P data, with aAustralian mexception of the depressions, are within the ranges previously reported for the upland e, 2001) and for the samet al.and seaward sectors of the Bragança Peninsula (Medina study area (Reise, 2003). At the LP, our higher tot.-P value (1.58 mg P g-1 DW) is in
one (Hesse, 1961). ethe range of Sierra L

ed by a ic behaviour of extr.-P was confirmThe effect of the IF on the dynamsignificant positive correlation. This pool fits very well with leaf-P for the same study
the IF (r = 0.86, n=33,e correlation witht positivarea, showing also a significanp<0.005) (Cordeiro et al., 2003) (Fig. 16). This effect was attributed to the solubility
s ). The general low P statuditions (Shapiro, 1958 reducing conof P-Fe/Al brought bye species in Sierra Leone by Hesse (1961) and for e saments reported for thin sedimthe Australian mangrove by Boto and Wellington (1984), both used the similar
phosphate extraction medium, reflecting the same dependency on the inundation
gradient. Recent results for Micronesia reported also low P status (Gleason et al.,
ents 2003). Our high values, support the abundant plant-available P detected for sedime) of the of tropical lowland evergreen rain forest and dry land forest (terra firm., 2003).et alBragança Peninsula (Frizano

the larger fraction domBased on the fractionation analysis, the percinated by P-Fe/Al (Tab. 3entage of calculated inorg.-P accounts for ). These results are analogous for a
lva and Sampaio (1998) iazon reported by Sbu Island at the Amlow plain in Comusing the same methodology, but considerable higher as the values (ca 55 %) reported
by Fabre et al. (1999) for French Guiana. The inorg.-P data suggests that P-Fe/Al is
ents in all parcels. These high pound present at the sedimin insoluble P comathe mresults might be caused under tropical climate. The strong chemical weathering of the

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

79

lateritic source area supply oxidized par, besides in water ticles (Fe(hydr)oxides)plexes), that are transported and cominiumical components (alumdissolved chemdeposited partly as mangrove muddy sediment and/or transported to the ocean (Costa
., 2004b).et al

Effect of flooding on P-exchange 6.2.5. Previous hydrological studies on a Caeté tidal creek, suggested that the inundation mapresumy influence the chemably pH-dependent (Cohen et alical reactions involved in the retention and release of P ., 2004). At this creek, the basic pH range of
by Cohen et al. (2004), can potentially the flooding water (7.2-7.9) registered ., 1994; Dorozhkin, 2002). For the et alnerals (Diaz iprecipitate P as Ca-phosphate mP-Ca crystallization in seawater, longer periods of time (8-20 month) will be n2+eeded
) and., 1984), and the presence of significant calcium (Caet al(Gulbrandsenmagnesium (Mg2+) levels is determinant (Brown, 1981; Yadav et al., 1984), the latter
.,et alsking the adsorption sites on apatite (Martens and Harriss, 1970; Wang am surface waters due tooval (60 %) fromally, effective P rem1995). Additionprecipitation, requires Ca2+ concentration of >100 mg.l-1 at pH >8 (Diaz et al., 1994).
2+ (1062-sula, reported pore-water levels of MgPrevious works at Bragança Penin3124 mg.l-1) in one order of magnitude than those of Ca2+ (269-1176 mg.l-1) (Furtado
2+on apatite. da Costa, 2000), suggesting an inhibitory effect of Mg

At HP, the surface (5-10 cm) sediments with low react.-Ca content and slightly acidic
cted in higher ecipitation of P-Ca refleto neutral pH (6.5-7.4) values, showed low prCa : P-Ca ratios (9-12) (Tab. 4). Any precipitate formed and settled on the sediment
surface could also be rapidly solubilised, due the slightly acidic to neutral nature of
this plain. These intermittently flooded sediments, shift to extreme basic conditions
). However, Fig. 11ation of P-Ca ((13.3) during flooding periods, suggesting the formby considering the relative duration of inundation (2-3.5 h), longer reactions periods may be needed to form more stable Ca-phosphate minerals. The calcium phosphates
phate (ACP), Dicalcium Phosphate Dehydrate Phosphases Amorphous Calcium Phosphate (DCPA), are the initial muus Dicalci(DCPD) and occasionally Anhydro an and Rowell, 1981). Bell and Black (Freemreaction products of P with CaCO3

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

80

Phosphate (OCP), (1970) found that DCPD changed to Octacalciumthermodynamically more stable mineral, within 44 weeks in slightly acidic to alkaline
(6.9-7.9) sediments. Murrmann and Peech (1969) reported sediment extracts to which
equilibriuma constant Ca(OH) with OCP a2f had been added, displaying a vater 108 h. Thus, in LP higher P-Ca concentrations and thencelue about pH 8.5, nearly in
lower Ca : P-Ca ratios (4-8) (Tab. 4), reflected longer time of maximum pH (>8)
ed and these later reorder to are formfollowing each inundation, where freshly P-Ca more crystalline forms (Kirk, 2004).

At LP, Al ain high plains (Silva and Samnd Fe phosphates, which are presumpaio, 1998), increased concomably the dominant foritant with Eh depletion in ms of P minerals
moderately acidic to neutral pH (6.6-7.2) values, reflected by lower Fe : P-Fe/Al (60-95) and Al : P-Fe/Al (52-93) ratios (Tab 4). Besides, anoxic conditions provided a
release of extr.-P associated with Fe and Al (hydro)oxides resulting in negative correlations with their ratios. These data support the early hypothesis of Patrick and re P tooents released mKhalid (1974) and Khalid et al. (1977) that anaerobic sedimsolution and sorbed more P from solution than did aerobic sediments. These authors
mpounds in ferrous cocedrface area of the gel-like redusuggested a greater suanaerobic sediments resulting in more solubilisation and sorption. In intermittently
flooded sediments, Huguenin-Elie et al. (2003) suggests that in reducing conditions
nts increase, though with prolonged eergence the labile P in sedimfollowing submflooding the P became re-immobilized and in the sub-sequent drying the re-oxidation
pounds immobilize P. of reduced com

between extr.-P and Al : P-Fe/Al ratio, The higher significant negative correlation suggest that the aluminium phosphate fraction seems to be the most significant
ents solution. Thus, this P control fraction in controlling the P concentration in sedimis reasonable to assume due 1) the terrigenous signature (laterites) of the source area
(Aluma decrease in logarithminium complexes) (Costa et alic solubility potential (K., 2004b), and 2) the msp) in the oerder Al(OH)tal reactivity expressed as 3(33.5) >
Fe(OH)3 (38.8) > FeOOH(Į) (41.5) > Fe2O3(Į) (42.7) > Ca5(PO4)3OH ( 55.6) at 25oC
rine aibes, 1992; Butler, 1998), the lack of Ksp data for mand in ionic strength cero (L

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

81

ing interms do not support this hypothesis. Concernsystements, ittently flooded sedim

Murrmann and Peech (1969) described the labile P increase with varying pH (4-10)

involving the dissolution of Al (hydr)oxides containing occluded P. Inasmuch as P is

eadsorbed by Al (hydr)oxides, it is likely that P becom s distributed throughout the

(hydr)oxides structure during the repeated dissolution re-precipitation process.

The solubility of Al and Fe phosphates decreases at lower pH, directly opposite to the

st available around pH 6.5 o phosphates. Therefore, P is msolubility of calcium

et al(Havlin. At lower pH ., 1999) and pH 5-6 (Georgantas and Grigoropoulou, 2006)

levels, P retention is high due to P-Fe/Al precipitation, and at higher pH levels due to

P-Ca minerals precipitation. On the other hand, aluminium is efficient in trapping P in

a wider pH range (4-7) and when water pH is above 8 aluminium become soluble
-) (Georgantas and Grigoropoulou, 2006). Across the transect, in surface (Al(OH)4ents with a pH range for 6.5-7.5 we expect P availability to be relative high. sedim

Before submergence, neither sediment react.-Ca, Ca : P-Ca ratio nor P-Ca seemed to
influence the release of extr.-P from sediment, supporting that adsorption of Ca2+ is of

minor importance when Al(hydr)oxides and amorphous FeOOH is present

an, 1988). As the pH increases during flooding periods, Al and Fe species in (Golterm

the soluble form being negatively charged cannot trapping P successfully, suggesting

e. ited inundation tim phosphate phases due the limation of instable calciumthe form

Effect of redox stratification on P-exchange 6.2.6.

The present study indicated that stands of R. mangle and A. germinans differs in their

responses to low sediment Eh conditions and such differences may be partially

explained by the distribution pattern along the transect and the species-specific

oxidation ability.

In fact, the greater sensitivity of A. germinans to low Eh conditions was reflected on P

compounds (Fig. 17). As above-mentioned, under anoxic conditions there is a flush of

easily extr.-P attributed to solubility of P-Fe/Al brought by reduction (Shapiro, 1958).

BelowA. germinans (feature 1), the intense root-induced oxidation re-immobilized

the extr.-P by increase of P-Fe/Al, tot.-P and acidification. At D2, this acidification

Chapter 6– Physicochemical pattern and phosphorus status of mangroves

82

caused the dissolution of acid-soluble P (P-Ca). But countering this, the freshly
2+ oxidation would have a large P sorbing ed by Feprecipitated ferric hydroxid form

capacity (Saleque and Kirk, 1995).

BeneathR. mangle (feature 2), occurred the diminution of the root-oxidation

panied by the depletion of tot.-P and P-Fe/Al pools. The reduction of P-Fe/Al accom

itant with this extr.-P e increase of extr.-P. Concompool was reflected in an extrem

intained stable oraincrease, the P-Ca was me cases increased, probably due the in som

basic pH caused by the reduction, suggesting the re-immobilization of P by the

carbonates pool.

supply of P to roots would therefore depend on the balance betweeThe diffusen

acid-soluble forms increased sorption on ferric hydroxide and increased release from

, 1995). The vertical profiles elucidated the extr.-P behavior in the et al.(Saleque

horizontal profile (Fig. 16). Here, the inversion of the extr.-P curve at MP, suggest

that the higher root-induced oxidation in A. germinans limited the availability of P

and consequently their distribution. Thus, the adaptive response of to R. mangle

inance in low plain. nutrient uptake in water-logging could contribute to its dom

Chapter 7 – Dynamics of sulphur in mangrove sediment

CHAPTER 7

LPHUR IN MANGROVE SEDIMENT UCS OF SDYNAMI

83

Results 7.1. 7.1.1. Substrate characteristics: organic matter, salinity, granulometry, Eh and pH.
The sediments at the high plain (HP) were characterized by a silt and clay content of
96-98 %, this amount decreased along the transect to 90-97 % at the Rhizophora-
ents forest in the low plain (LP), probably due to tidal effects on the sedimained inundated on average 51 ± 13 days/year and remetry. HP wasgranulomcontinuously air-exposed during neap tides, whereas LP was inundated on average 146 ± 13 days/year, with the latter exposed to tidal currents. At the boundaries ddle plain (MP), slightly depressions of 5-8 cm ibetween both HP and LP, and the m paleo-channel identified as ains of a to be remdepth occurred, which appeardepression one (D1) and two (D2).

The salinities differed spatially among HP and LP, varying from 43.7 ± 16 o/oo to 34.6
± 10 o/oo(n = 47), respectively (Tab. 5). Salinity data have been plotted in contour
plots along the transect (Fig. 21 (b)). In the surface layer, the iso-lines 20 o/oo were
discontinuous, due to the influence of sub-surface layers with higher salinity (30 o/oo).
-lines distribution showed a strong ity iso-forest (HP), the salinAvicenniaIn the gradient with depth, resulting in a profile with higher stratification. Conversely, at LP, being permanently under tidal influence, the salinity iso-lines 20-40 o/oo sharply
deepened into the Rhizophora-forest toward the estuary, characterized by sediments
with low stratification.

The sedimentary organic matter (OM) ranged between 4 to 13 % with no systematic
trend along the transect (Tab. 5). At HP, higher OM contents in the top 10-cm-thick
. At LP, the decrease was only moderate (4.1- cmlayer (13 %) decreased (6 %) at 100, whereas at D2 R. mangleined only below t was determ3.8 %). At D1, the OM contenbelow both species, characterized by an average of 3.5 ± 1.1 % (n=7) and 4.9 ± 0.8 % (n=11), respectively.

Chapter 7 – Dynamics of sulphur in mangrove sediment

84

esented in Figure 18, with the stationsRedox potential (Eh) and pH iso-lines are prarranged with respect to an increasing inundation gradient. In this figure, the redox odel proposed by Clark et al. (1998) related to the oxidation and stratification m) in Ves not evidenced. The Eh pattern was higher (200 mreduction zones, becomAvicennia-forest (HP) than in Rhizophora-forest (0-100 mV) (LP) (Fig. 18 (a)).
ination of the depressions (D1, D2) revealed two oxic-anoxic interfaces: 1) Eh = Exam at LP. These sectors agree both with the V HP; and 2) Eh = 0-50 mt0-150 mV afunctional causes have to evaluated. physical separation between the plains, the Additionally, the pH range was 6.3-7.5, showing slightly basic values in HP and MP, (b)).Fig. 18 and slightly acidic values in LP (

Table 5: Physical anddeep (45-100 cm chem) sediical paramment samples acroeters in surface (10 cmss a tidal gradient. C), mi:ddle (15-35 cmN slope ref) and ers
to the ratios calculated through the inclination (b=slope) of regression lines a) and standard deviation (SD).xin Figure 20. Mean values (

Parameters nDepth HPD1MPD2LP
) (cm

IF(days/year) 1-4 -41 - 62 6780 - 101 124 128 - 162

Salinity (o/oo)6-10 0-10 20 ± 2 27 ± 6 28 ± 11 26 ± 7 24 ± 7
10-20 15-35 38 ± 9 37 ± 9 40 ± 9 34 ± 6 33 ± 8
10-20 45-100 55 ± 16 53 ± 9 49 ± 8 38 ± 4 44 ± 7

) (%OM

(mC:Nol/mol)

C:Nope b = sl

6-10 0-10 12.9 ± 1.7 3.8 ± 1.8 6.9 ± 0.2 -4.1 ± 1.3
10-20 15-35 5.8 ± 2.7 3.7 ± 0.7 6.6 ± 0.8 6.1 ± 1.7 4.0 ± 0.8
10-20 45-100 5.7 ± 1.7 3.0 ± 1.3 3.9 ± 1.5 4.8 ± 0.9 3.8 ± 1.1

6-10 0-10 15 ± 1
10-20 15-35 15 ± 3
10-20 45-100 15 ± 2

515-9-12 5-Rh15i 14
9-13 Avi15

17 ± 1 18 ± 2 16 ± 0 17 ± 3
15 ± 3 18 ± 2 16 ± 3 16 ± 3
16 ± 4 19 ± 3 17 ± 1 18 ± 2

--

--

--

2125

Chapter 7 – Dynamics of sulphur in mangrove sediment

a

Abbreviations: n, number of samples; HP, high plain; D1, depression 1; MP, m

plain; D2, depression 2; LP, low plain; OM, o

Rhi,

R. mangleFigure 18: (

; Avi,

aA. germinans.

) Lateral view of Eh, (

ents. sedim

ganic mr

tter; C:N, ma

85

iddle

lar C:N ratioo

b) Lateral view of pH, both m

;

easured in

Chapter 7 – Dynamics of sulphur in mangrove sediment

86

7.1.2. The elemElemental comental composition considered the sedimposition and ratios ent below each species separately,
c carbon (org.-C) and total nitrogen (tot.-ical differences. The organidue to their chem-0.22 wt %, respectively. N) exhibited a wide range between 0.82-3.24 wt % and 0.09A. germinans and R. mangleThe down core distribution of org.-C and tot.-N below was examined in detail (Fig. 19). At HP, for both species, org.-C decrease from the
surface to 50 cm depth increasing in deeper substrate (100 cm),whereas tot.-N showed
a decreasing trend from the surface to 100 cm deep (Fig. 19). At LP, the org.-C down
e upper 10-cm-thick layer with highest core profile differed among species in thconcentrations below R. mangle in comparison to A. germinans. At the middle layers
), both species showed a pronounced org.-C increase. Conversely, tot.-N(15-50 cmbelowA. germinans was maintained higher in the upper 15-cm-thick layer, in relation
a R. mangle, which showed a smooth decrease with depth; below R. mangletouniform feature with slightly increase with depth occurred.

The molar C:N ratios showed lower values in HP (15 ± 2, n=46) in comparison to LP
). At HP, the C:N ratio Tab. 5(17 ± 3, n= 45), this difference was not significant (remC:N difference between both plains and to attenuate the efained almost stable, whereas at LP increased with sedimfect of outliers, thent depth. To confirm the e C:N
ratios for surface sediments (5-15 cm) were calculated on the basis of the inclination
(e.g. slope) of the regression line (Fig. 20 (a),Tab. 5). A significant correlation
between org.-C and tot.-N in A. germinans- (HP, r=0.95; LP, r=0.94, n=9, p<0.05)
andR. mangle-dominated sediments(HP, r=0.86; LP, r=0.88, n=11, p<0.05) occurred.
parison to LP ed lower values in HP in comnfirmes coThe slopes of the regression linand also showed lower C:N values below R. magle in relation to A. germinans (Tab.
5). Additionally, at intermediary layers (25-50 cm) the molar C:N ratio correlated
significantly with Eh which was negatively related (HP, r=-0.74; LP, r= -0.74, n=11-ents are those ediary layers the reducing environm19, p<0.05), showing that in intermC:N values(with highest C:N values,Fig. 20 (b)). whereas the environments with positive Eh record the lowest

Chapter 7 – Dynamics of sulphur in mangrove sediment

Figure 19:

87

Concentration (wt%) versus depth profiles of organic carbon (org.-C) and

total nitrogen (tot.-N) below

P). low-plain (L

and R. mangle

A. germinans at high- (HP) and

A. germinansmangleR.(Eh). HP, high plain; LP, low plain; D1, depression 1.

and

r ratio vs. redox potential alo, respectively. C:N m

en contents (wt %)

d nitrog

t %) an

ts a least-square with a slope en

corresponding with a molar C:N of 15.7 ± 2.3 and 17.2 ± 2.8 below

Relation between the organic carbon (w

ents. The line represof Bragança sedim

Figure 20:

88

Chapter 7 – Dynamics of sulphur in mangrove sediment

Chapter 7 – Dynamics of sulphur in mangrove sediment

89

Sulphur pools 7.1.3. ), and the iso-lines for ations (20-25 cmThe horizontal profile of sulphide concentrsalinity, sulphate (SO42-) and the percentage of expected concentration (PEC) of SO42-
gure the stations are arranged with respect to an iare presented in Figure 21. In this fincreasing inundation gradient.

At D1 and MP, sedimnts have higher sulphide concentrations (10-47 μM) than those efor HP and LP below the detection limit (<1 μM) (Fig. 21 (a)). At MP, below A.
e μM), compared to th-47, sulphide reached higher concentrations (34germinansextreme lower concentrations (<4 μM) below R. mangle. Here, the sulphide increase
was concomitant with the lowest SO42- concentrations (5 mM) and PEC (30 %) values
(c)(d)).Fig. 21 of the transect (

2- concentrations reflected the changes in salinity (i.e. source) across In general, SO4the transect (Fig. 21 (b)(c)). At HP, a sharp SO42- gradient with depth was coincident
with high salinity stratification, whereas at LP, the SO42- iso-lines 10-15 mM enters
e plain. If exist ilar as the salinity for the samre inundated a rea, simodeeper into the med a high significant correlation between a conservative behaviour it will be expectsalinity and SO42-, although a close to significant correlation (r=0.70, n=143, p<0.05),
deviation is found for MP, where the low several deviations occurred. An evident SO42- iso-lines 5-15 mM enter into deeper layers, differing for the salinity iso-lines
pattern.

ically non-eter (biologically and chemSalinity, considered as a conservative paramreactive), was used as a tracer to determine whether decrease in SO42- concentrations
etical processes (Madureira mcal and chewere due to either hydrological or biologial., 1997; Sherman et al., 1998; Cohen et al., 1999). In the present study, the
percentage of expected concentration (PEC) of SO42- showed most values below 100
% indicating that the sediment contains a lower amount of SO42- as would be expected
for a given salinity, suggesting biological uptake, reduction, sorption and/or precipitation processes (Fig. 21 (d)). The absence of PEC values above 100 %
indicate the lack of SO42- input from other sources beyond seawater or net re-

Chapter 7 – Dynamics of sulphur in mangrove sediment

90

a loss (65-70 %) in icating indsharply,oxidation processes. At MP, the PEC decreased SO42- relative to salinity across the 30 to 60-cm thick layer. In contrast, at LP, the PEC
increased attesting a lower SO42- loss (35-40 %) in 30 to 80-cm thick layer.

ong the forest ent status amTo elucidate if the difference in the sulphur pools in sedimtypes is a function of tree species or differences in tidal elevation, the down core ediateeters for two intermdistribution of the sulphur pools and associated paramstations at HP and LP were examined. At HP (Fig. 22 (a)), a different picture was
displayed a A. germinans the species, where for water content betweenobtainedsubsurface (5-15 cm) water content maximum in relation to R. mangle. The salinity
and SO42- distribution, similarly as the humidity, increased with depth. Below A.
er, the PEC depletion (50-36 %) point out lower , at 5 to 15 cm thick laygerminansSO42- values in relation to salinity. In addition, below A. germinans, the reactive iron
depth 15 cm-(react.-Fe) profile was characterized by a subsurface depletion at 5e itant with a DOP increase (45-52 %), which indicated the iron sulphidconcomformation trough the reaction of sulphide with react.-Fe. At LP (Fig. 22 (b)), below A.
germinans the water content displayed a deep (50 cm) maximum, in comparison to
the subsurface maximum at HP for the same species. Here, PEC values (100 %)
belowA. germinans evidenced the maximum SO42- value in relation to salinity,
suggesting a continuous supply of SO42. The highest DOP values in the lower part of
), coincided with the react.-Fe depletion, indicating reduction the profile (50-100 cmprocesses.

Chapter 7 – Dynamics of sulphur in mangrove sediment

Figure 21: (

) Horizontal sedima

ent profile of sulphide at 20-25 cm

transects of iso-lines for sedim

ent pore-water data of (

de

91

pth. Lateral

)c) salinity (b

sulphate and (d) percentage of expected concentration (PEC). HP, high

ddle plain; LP, low plain; depressions (D1, D2). iplain; D1; MP, m

Chapter 7 – Dynamics of sulphur in mangrove sediment

Figure 22

: Vertical p

iles of sedimfor

ent water content, salinity, sulphate, percentag

concentration (PEC), dof expected

iron (react.-Fe). (

) HP, high plain, (a

gree of pyritization e

) LP, low plain. b

(DOP) and reactiv

92

e

e

Chapter 7 – Dynamics of sulphur in mangrove sediment

Continuation Figure 22

93

Chapter 7 – Dynamics of sulphur in mangrove sediment

94

phologies roIron pool and pyrite m7.1.4.The react.-Fe was characterized by minimum and maximum concentrations occurring
at the surface and deep sediments, respectively (Fig 23 (a)). At surface layers (D2,
-1oxic interface itant with an oxic-an) concomLP), the react.-Fe decreased (22.5 mg.gat redox potential Eh of 0 mV (Fig 23 (a),Fig. 18 (a)). At deep layers, react.-Fe
decreased from 35 mg.g-1 (MP) to 25 mg.g-1 at the inland (HP).

e degree of pyritization (DOP) was th pyrite iron (Pyr.-Fe),-Fe andMeasuring react.used to determine whether Fe availability limited pyrite formation (Berner, 1970).
pleteness of the reaction of this react.-Fe (Fe III) with easure of comDOP is a mylor, 1990). Over the first 100 m of the transect aaqueous sulphide (Lenventhal and Taining (HP), the DOP was 52 ± 2 % indicating that less than half (48 %) of the remreact.-Fe could be potentially sulphidized (Fig. 23 (b)). At D1, the DOP increased to
55 ± 1 %, than decreased to 51 ± 3 % at MP. At LP, a pronounced increase for 61 ± 3 le on pool was still availabat about 38 % of the total ir% occurred, which indicates thation. potentially for pyrite form

croscope analyses iscanning electron med through The pyrite presence was confirm(SEM-EDX). Pyrite grains were hard to locate below the interface sediment-water (<5
cm); they were restricted to the medium (10-25 cm) and deep (40-100 cm) layers in
the vertical profiles. The single pyrite grains in the mangrove sediments average a
mean diameter of 1.2 μm. Two range of pyrite grains were identified: 1) 0.6-2.3 μm-
size at medium layers (Fig. 24 (a)(b)), and 2) 0.8-4.1 μm-size at deep layers.
Framboids morphology ranging in size from 2.7 to 21.5 μm-size with a mean
diameter 7.4 μm, dominated in the reducing sediments at greater depths (40-100 cm).
At LP, below R. mangle (100 cm), the framboids clustered in a diatom (45 μm-size)
(Fig. 24 (c)) and an octahedral crystal (5.2 μm-size) were observed (Fig. 24 (d)). The
latter textures were concomitant with the highest DOP value (67 %) at 100 cm depth
(b)).Fig. 22(

ationboid formrite framIn order to asses the reducing conditions with respect to py(Wilkin et al., 1996), a survey on the number of microcrystals (Nm) in framboids

Chapter 7 – Dynamics of sulphur in mangrove sediment

along a topographic gradient was m

). Because the sample mFig. 25de (a

95

ounts

represent a random cut of the sediment sample, the measured distributions are only an

approximation of the true distribution. Individual framboids from sediments of the

Bragança Peninsula contain less than 100 microcrystals arranged in a decreased

(HP), ind LP to the inlandmt frogradien

growth are different in the plains.

Figure 23 that the factors controlling framicating

boid

and (: Lateral transb)ect of iso-lines sedimdegree of pyritization (DOP). HP, high plain; D1; MP, ment data for (a) reactive iron (react.-Fiddle e)
plain; LP, low plain; depressions (D1, D2).

crophotographs showing various pyrite textures observed at thei

, and (boids into a diatommof fra

) well developed octahedral crystal.d

nute pyrite i

ster

), cluc

ngrove root, (acovering a m

) group of pyrite crystals in a loci (b

a) poorly developed m

ents: (

ngrove sedima

Bragança m

: MEV m

96

Chapter 7 – Dynamics of sulphur in mangrove sediment

Figure 24

Chapter 7 – Dynamics of sulphur in mangrove sediment

Figure 2597

: Distribution of the number of mtopographic gradient. HP, high plain; MP, microcrystals (Nim) in fraddle plain; LP, low plain. mboids along a

7.2. Discussion

ental source characterization Elem7.2.1.

Sediment organic matter (OM) serves as nutritive substrate for micro-organisms and

diated processes such as reduction-oxidation and the einfluences biologically m

cycling of nutrients. The suitability of OM as a carbon or energy source can vary

an, 1985). Therefore,depending on the source of vegetal tissue (Anderson and Colem

er the litter and root production, the retention and portant to considit is im

position of these tissues within the mdecom

individual species.

ngroves when evaluating the effects of a

Chapter 7 – Dynamics of sulphur in mangrove sediment

98

Concerning HP, which was flooded only at spring tide, a high OM occurred; this ). These Tab. 5highly flooded, obtained low OM (contrasted with LP, which being results act as supporting evidence with respect to tidal hydrology, which predicts that e to higher tidal flushing rates of litter export du fringe forests have highriverine and(Lugo and Snedaker, 1974). The OM data do not clearly behaves with respect to A. gradable OM in sediments belowposition, where the presence of more dedecomgerminans can sustain a higher rate of microbial activity than sedimentsbelow R.
mangle, and as a consequence, more decomposition of OM occurs (Lacerda et al.,
1995). Decomposition in the mangrove forest can be regulated by the quality of the
organic substrate (Sherman et al., 1998). Decomposition studies have consistently
demonstrated that Rhizophora leaves decompose more slowly than Avicennia leaves,
centrations and higher tannin content. The tannin produced aas a result of lower N condecrease in the activity of benthic and microbial organisms, and as a consequence, a
higher OM accumulation in sediments occurred (Cundell et al., 1979; Boto et al.,
1989; Robertson et al., 1992; Lacerda et al., 1995). Previous studies in Bragança
Pponent, where as Hportant litter comshowed that the leaves were the most improduced yearly the lowest amount of litter (7.2 t.ha-1.y-1) compared to LP (10.0 t.ha-
-11ith these results probably because of ) (Menezes, 2006). Our OM data contrast w.y the tree ture what falls fromrily capathe fixation of the litter traps which not necessterial in aaining m(i.e. production). In the present study, OM values reflect rement after flooding during the rainy season (February) which is characterized by sedimthe minimum litter production of the year (Reise, 2003).

The org.-C and tot.-N dynamics in the surface sediments were most likely linked to
the decomposition and export of the litter produced by both species (Fig. 19). The
ents of Bragança (0.82-3.24 wt %) was higher than range in org.-C contents in sedimat other mangrove sediments (0.4-2.2 wt %) (Alongi et al., 1993; Kristensen et al.,
1994). At HP and LP, although both org.-C and tot.-N are released during mineralization, the sub-surface (10 cm) clearly marked org.-C concentrations below
R. mangle reflected a different control as mineralization was probably linked to 1)
root exudation or degradation of fine roots (Boto et al., 1989; Holmer et al., 1999),
and/or 2) higher necromass-root biomass for R. mangle, in comparison to the higher

Chapter 7 – Dynamics of sulphur in mangrove sediment

99

live-root biomass for A. germinans, reported for the same study area(Reise, 2003).
) are consistent with mHere, high org.-C concentrations in deep sediments (50-100 cde for aprevious studies of dissolved organic carbon (DOC) in groundwater mchwendenmann, 1998). Higher DOC ninsula during the rainy season (SeBragança Pents indicate a relatively higher degree of ngrove sedimaconcentrations in mre abundant bacterial population and detrital oanaerobiosis and stagnation as well as mmatter on the forest floor (Alongi, 1988; Boto hypothesize that during flood tide, groundwater is recharged by infiltration of water et al., 1989). Ovalle et al. (1990)
grates (vertical fluxes) back to the creek, iand solutes, and at ebb tide, groundwater m below-ground processes. The deeper and acts decoupling vegetation pattern fromitted to anently under the influence of the pore water and subm permzones beingpermanently anoxic conditions would be favourable for the preservation of the org.-C
., 1998). et alcontent (Lallier-Verges

The sedimentary C:N ratios range of 15 to 18 (Tab. 5), implies that the organic input
rine sources (i.e. algae and phytoplankton) that is maes completely fromdoes not comcroscopic study i 1994), although the mcharacterized by a ratio about 7 (Meyers,(c)). During senescence of the mangrove leaves Fig. 24 revealed algal-derived MO ((i.e. before litterfall), about 64 % of the N is reabsorbed by the mangroves (Rao et al.,
1994). Consequently, the sediments receive mangrove material that is depleted in N
, 1996), resulting in C:N ratios higher than the overlying vegetation. et al.(Middelburg paredw if comentary C:N ratios are rather loIn contrast, our data indicated that sedimies (Cordeiro and these specngrove fresh leaves C:N ratios (22-32) fromato mmilar observation about nitrogen excess with respect to carbon Mendoza, 2007). A siin sediments related to mangrove material, has been made in the East African
mangrove (Middelburg et al., 1996). Previous studies in Bragança Pensinsula,
suggested that low C:N ranges indicated that woody tissue did not contribute entary OM and, also, that this fact was indicative of a substantially to the sedim 2001c). At respiration (Dittmar and Lara,ing or ss of carbon due to leachselective loR. mangle core distribution for tot.-N below and slightly rising downLP, the uniformFig. 19suggests that their immobilization () relates to the high tannin content in Rhizophora leaves (Lacerda et al., 1995). Meanwhile, previous studies in the same

Chapter 7 – Dynamics of sulphur in mangrove sediment

010

en compounds can not account itrogins and other nat the tannstudy area suggested th fact, at the with depth (Schmitt, 2006). In distribution of tot.-N valuesfor the uniformlar C:N ratio rising at surface and deep oinated plain (LP), the m-domRhizophoralayers (Tab. 5), reflected a proportionally higher increase in org.-C than in tot.-N
values (Fig. 19). At middle layers (25-50 cm), the highest C:N ratios associated with
ents (LP) and the lowest C:N ratios with positive Eh (HP) mmore reducing environ(Fig. 20 (b)), support the studies of Lallier-Verges et al. (1998),suggesting that the
pounds. For the ation of nitrogen comore rapid formoxidative conditions allow a me transect, the degradation of proteins and the partial loss of the corresponding samnitrogen only drop from HP to LP during rainy season (Schmitt, 2006), confirms the
results of the present study.

7.2.2. The sulphur-iron pools and some thermodynamic consideration
In Bragança, A. germinans and R. mangle grow under a wide range of salinities (18-
68 o/oo) (Tab. 5, Fig. 21 (b)). This range was higher as compared with seawater o
),/ the Caeté River during tidal inundation (8-32 diluted by freshwater fromooestimated in a contiguous creek during rainy season (Cohen et al., 1999). Surface
salinity (20 o/oo) varied, responding to tidal inundation, intensity of the rain and
luencing the evaporation. The inundation gradient controlled the surface salinity infooded during stability of the different forest stands. The elevated plain (HP) is flremaining water. Here, the higher salinspring tide only and in between evapotranspiration causes a salinity inity stratification reaching hyper salinecrease of the
conditions (70 o/oo) in deep (50 cm) sediments, suggests the lixiviation from the upper
layers to the deeper layers during flooding. These results, were consistent with studiesin Bragança, which reported higher salinity (60 ± 7 o/oo) for 50 cm in deep than for the
o) (Furtado da Costa, 2000). At LP, the low salinity stratification /surface (35ooreaching 50 o/oo indicated that R. mangle requires an occasional fresh inundation,
(2004) in French Guiana..ith Marchand et alconsistent w

), VOn a whole-forest scale, the sediments had a high redox potentials (0-200 mindicating that the sulphate reduction is a less important diagenetic pathway in
Bragança or that the reoxidation of hydrogen sulphide due to the high abundance of

Chapter 7 – Dynamics of sulphur in mangrove sediment

110

iron keeps accumulating pore water sulphide concentrations low (Moeslund et al.,
). In addition, very low sulphide pool was found at HP and LP sites g. 18iF1994) (ith a low sulphate reduction, a large degree of aeration during low tide consistent w(HP) and probably due to enhanced sulphide reoxidation due to intense bioturbation by benthic fauna (LP) and reaction with oxides (Fig. 21 (a)). Alongi et al. (1998b)
estimated that sulphate reduction represented a considerably smaller proportion (42
%) of total organic matter oxidation and Holmer et al. (1999) reported similar results
ngrove forests, a(<2 to 44 %) at a 2-years forest, in a Malaysian and Thailand m., 1991) and in et alrespectively. Other studies conducted in Thailand (Kristensen iration and sulphate reduction ., 1994) suggest that aerobic respet alaica (Nedwell Jamare the major pathways of OM diagenesis in mangrove sediments. The low SO42- iso-
lines (5-10 mM) in the surface layer (5-10 cm) during wet season (Fig. 21 (c)),
gradients as suggested by the salinitye association toprobably occur due to 1) thsalinity variation (Fig. 21 (b)), 2) an intensive bacterial dissimilatory sulphate
the reduction to the rapid mineralization of OM, and/or 3) diffusion of sulphate fromoverlying water and from the underlying sediments to deep layers. Holmer et al.
(1999) have similarly observed low sulphate values (11 mM) in the top sediment
layers.

In spite of the oxic-anoxic interfaces (Eh = 0 mV) due to the rough surface
) of the upper oxidation topography at the depressions (D1, D2), the thickness (~1 cmzone (UOZ) suggests a greater biological heterotrophic activity (Fig. 14, Fig. 18). At
MP, the SO42- loss (5 mM) and minimum PEC (30 %) iso-lines, suggest that
biological and/or chemical processes controlled SO42- concentrations in this plain
ate reduction, higher free sulphide concentrations (c)(d)). As result of sulphFig. 21((10-47 μM) were observed in the pore water below A. germinans (Fig. 21 (a)). The
decrease of sulphide concentrations below R. mangle was probably due to 1) higher
oxidation capacity by iron (hydr)oxides related to A. germinans (Fig. 14), and/or 2) to
the corresponding formation of metal sulphides. Apparently the amount of oxidants
released from the living roots is sufficient to avoid the sulphate reduction (Hinnes et
al., 1989). At our study area, the higher necromass-root biomass for R. mangle in
relation to A. germinans (Reise, 2003), can be places for pyrite formation and trace

Chapter 7 – Dynamics of sulphur in mangrove sediment

210

tal precipitation due to enhanced concentration of metals in dead roots (Alongi emet., 2005). al

eters (water ithin tree-scale, the down core distribution of sulphur and related paramW (a)(b)). Fig. 22C) was consistent among the stations (Econtent, salinity and PRegardless of whether the species-specific differences were caused by the species ent conditions crosite differences, a closer look on sedimiselves or by existing mthemfor the establishment, growth and survival of mangroves is necessary. In the present
study, the down core distribution of SO42- below A. germinans significantly correlates
with salinity (r= 0.90, n= 8) (Fig. 22 (b)) and the maximum PEC values (100 %) at 50
cm depth suggests SO42- diffusion from the overlying layers, both intensified by the
e presence of crabs burrows. Here, the hydrological processes in LP and probably ths bioturbation trough the presence of deep deposit-feeders, inhabiting burrow with a living cave lying at the burrow end penetrating to a depth of 164 ± 52 cmaker, 1998) facilitate oxygen transport and adem of water (Rlcontaining about 500 moxic decomposition (Alongi et al., 2001) in deep layers. Two points are clarified in
relation to salinity as SO42- source: 1) the presence of surface (HP) and deep (LP)
id layers resulting in saline intrusion, and 2) within tree-scale differences the humwater content can be greater than within forest-scale differences, these tree-scale
ity in a forest-scale. crease the average salinuate or to indifferences tend to atten

Because pyrite precipitation and iron solubility are partially dependent upon sulphate
reduction, a sediment shift from oxic to anoxic conditions (Eh = 0 mV) was chosen to
indicate conditions below which significant sulphate reduction occur, consistent with
Schulz (2000) and Gleason et al. (2003). In a first scenario, at the surface (D2, LP), the react.-Fe depletion concomitant with an oxic-anoxic interface (Eh=0-50 mV)
onstrate that the mcorresponds with observation of Luther et al. (1992), which deolecules onditions (Eh<0 mV) and produce mreact.-Fe depletion occur under anoxic c (a), Fig. 23with reactive functional groups such as hydroxyl- and carboxyl-groups (Fig. 18 (a)). Meanwhile, in a second scenario (D1), the depletion of react.-Fe
concomitant with an oxic-anoxic interface (Eh=0-150 mV) do not occur. For both
scenarios (D1, D2), the predominant alkaline range (7.1-7.5) in deep sediments (Fig.

Chapter 7 – Dynamics of sulphur in mangrove sediment

310

18nerals in the first ilution of Fe(III) m (b)) support the sulphate reduction. The dissoligands produced by exudation of algae and scenario could be explained by 1) organic etary productivity (Luther plants (Stumm and Sulzberger, 1992), and 2) a higher primean stand age of ., 1992), as indicates the robust marine influence and the higher mal stand (22 years) (Menezes, 2006). The A. germinans (26 years), related to R. manglee e oxic-anoxic interfaces can point out thplication for a functional cause of thesimlimitation of the temporal equilibrium of the co-ocurrence of both species. In principle
ation via is favourable for pyrite form)Van oxic-anoxic interface (Eh = 0 m., 1996; Rickard and Morse, 2005), this exclude et alilkin polysulphide pathway (Went redox potential at the ffect for sedimthe possibility of a species-specific edepressions.

Degree of pyritization and pyrite texture 7.2.3. The degree of pyritization (DOP) across the transect (Fig. 23 (b)), fall between 50 and
60 % which is between the range of values (DOP = 40-70 %) typically reported for
modern saltmarshes sediments (Giblin, 1988; Raiswell et al., 1988; Sherman et al.,
1998). The DOP values obtained here are indicative for sulphide-limiting suboxic
S), but no Hconditions (also called dysoxic or dysaerobic - very low to no O22(Raiswellet al., 1988), in agreement with the measured low sulphide concentrations
(a)).g. 21iFin the sediments (

by tidal flushing andely controlledThe DOP values across the transect are largsulphate reduction (Fig. 21 (c)), and to a lesser extent by root oxidation and
bioturbation (Fig. 22 (b)). Although Berner (1970) argued that Fe may control pyrite
itation of react.-Fe ation at DOP values above 50 %, our data do not showed a limmfor (a)). The DOP Fig. 23in LP, in relation to HP, where pyrite burial was greatest ( ood accounting of iron availability during earlys do not provide a gapproach seeme diagenesis (Rickard and Morse, 2005). Thus, it appears that DOP values are reflectivent sulphate reduction and pore-water redox conditions, supporting of the sediman et al. (1998) and Roychoudhury et al. (2003), previous studies of Shermlative constant value of DOP with depth can be ore, the re Furthermrespectively.interpreted as pyritization being complete below the sediment-water interface (Fig. 23

Chapter 7 – Dynamics of sulphur in mangrove sediment

410

ation can be regulated by 1) the rooting form(b)). Meanwhile, the pattern of pyrite ngroves vegetation and their active radial oxidation loss (ROL), and 2) azone of the mthe iron concretions found on mangrove rhizospheres (Lacerda et al., 1993). At LP,
the more efficient reduction and sulfidation of Fe, relative to the HP, increased the
DOP (Fig. 23 (b)) and consequently the number of microcrystals in framboids (Fig.
).25

nute dispersed grains, irphologies occurs as moAt Bragança Peninsula pyrite moctahedral crystals, framboids and framboidal cluster (stoichiometry Fe2S2) (Wolthers
et al., 2003) (Fig. 24). Diagenetic pyrite formation in the sediments significant
ents sedimentary pyrite in Bragança, contribute to the partially oxic to suboxic sedimsyngenetic pyrite in the water column was not evaluated. The scarce pyrite
pointed out the strong influence of the depth interval, distribution in the 5 cmrrent and also the diffusive input of advective transport of pore water by tidal cuergence. Studies of Aragon et al. (1999) in mangroves of oxygen during non-submaller advective transport of pore water at the South-East Brazil, suggested that a sm depth interval occurred. In our parison to the 5 cm depth interval in com10-25 cmstudy area, the first range of pyrite grains (0.6-2.3 μm-size) predominated at medium
g their rapid ), providing evidence supportinofiles (10-25 cmlayers of the vertical prformation (Fig. 24 (a)(b)). The rapid formation of small (0.1-2 μm-size) euhedral
rsh by Howarth aents of salt mpyrite particles have been observed previously in sediments provide ngrove environma. It is reasonable that m(1979) and Luther et al. (1982)dynamic conditions for pyrite growth because the surface is exposed to the
atmosphere and the rhizosphere of mangrove trees can oxidize deep sediments (10-35
., 1998). The second range of pyrite grains (0.8-4.1 μm-size) at deep et al) (Clark cm unaffected by the root oxidation and can s to be relative) seemlayers (40-100 cme to nucleation and growth, suggesting that further be interpreted as sufficient timpyrite was stable.

At the surface and middle layers (<35 cm), the lack of framboids accumulation
suggested that the mangrove sediments are exposed to the active root systems
controlling their formation. Although the rhizosphere environment of salt marsh may

Chapter 7 – Dynamics of sulphur in mangrove sediment

510

provide places for the growing of framboids (Luther et al., 1982), this hypothesis is
de in this study. In comparison, the anot consistent with field observations m) pointed out the reduction of the boids in deep layers (40-100 cmmabundance of fratransport processes and turnover ranges in these sediments (Wijsman et al., 2001).
crocrystals iber of m HP to LP of DOP (52.5-65 %) and the numThe increase from(12.5-70), indicated the framboids formation in reducing mangrove sediments (Fig.
23,Fig. 25). The average mean framboid diameter (7.7 μm-size) is representative of
an oxic and dysoxic environment (Wilkin et al., 1996). The approximate growth time
ated ents can be estim-size) for Bragança sedimboid (21.5 μmof the larger framcomparing with the maximum 7 years for framboids of 25 μm-size (Wilkin et al.,
1996). The maximum size of framboids formed in environment with typical
concentration of dissolved Fe (<100 μm) in pore water would be 50 μm (Raiswell et
., 1993). al

l and euhedral grains was cited as very boidaciation of a framThe close spatial assoowicz, 2005). The peculiar link between these two áawcommon in nature (Sowicz (1993) and áutz (1966), Sawstmorphologies was studied by Love and AmWilkin and Barnes (1996). In the process of infilling “Sammelrekristallisation”
(German term for fusion-recrystallisation), new pyrite fills the spaces between the
microcrystals in the framboids (Love and Amstutz, 1966). Similar to the
transformation from framboid to euhedral, the cluster framboid (FeSaq cluster) may
recrystallize to a massive grain (Sawáowicz, 1993; Butler and Rickard, 2000). At LP,
belowR. mangle (100 cm), the simultaneous presence of FeSaq cluster and an
(c)(d)). Fig. 24owicz (1993) (áoctahedral grain corresponds with observations of SawIn this station (100 cm) (Fig. 22 (b)), below R. mangle, the maximum DOP (67 %)
and lower react.-Fe (24.7 mg.g-1), in relation to A. germinans, suggests that the
continuous supply of constituting material to the microenvironment of framboid
plest way of leading to a euhedral pyrite ts the simation occurred, this represenmforcrystal, with framboidal pyrite as an intermediate stage (Sawáowicz, 1993).
ent of octahedral grains of pyrite was the poor developmAdditionally, apparentlyrelated to slower rates of pyritization (Wilkin and Barnes, 1996). The FeSaq cluster
(Fig. 24 (c)), seems to infill the free space in a fossil diatom, this may not always be

Chapter 7 – Dynamics of sulphur in mangrove sediment

genetically related to fossilization, b

sediment (Saw

á

owicz, 2005).

ecause th

e fossil serve o

610

ly as available space in n

ons Chapter 8 – Conclusi

CHAPTER 8 – CONCLUSIONS

710

e (2-3.5 h) that the water flooded the HP of the During the short period of timmangrove area, there was a significant transient effect on surface sediment chemistry.
ergence, the sharp Eh reduction and pH increase could be due to the high During submFe/Al-P and SO42- concentrations, summed to the high mineralizable OM in HP. The
sulphiderapid initial drop and gradual depletion was related to the SO42-consumption
dilution by the floodwater. The sulphide during the OM degradation and/or sulphidedilution supports the mixing of freshly tidal water in the interstices of sediment
surface (2 cm) with sea-water. The re-establishment of non-submergence conditions
ply oxidation and n or simy involve biological processes in the water columam., 1997). Clearly, the effect of et alents (Caetano reduction inside the sedime ergence on these trends is irrefutable. Meanwhile, the presence of the electrodsubmcan amplify the intensity of the measurements.

ent pore ical characteristics between sedimThe interaction of different physicochemthe concentration of P in the aqueous s to control water and flooding water, seemphase by the chemical equilibrium of several minerals for the lateritic source area. As
being mergence, Al and Fe species in the soluble for submthe pH increases during ation ofnegatively charged cannot trapping P successfully, suggesting the formited inundation timinstable Ca-phosphate phases due to the lim e. The last hypothesiswas confirmed by the ratios between reactive metals and phosphate minerals, and the
negative correlations of theses ratios with extr.-P, the latter confirmed that Al-P was
ents. In addition, st significant fraction controlling the P concentration in sedimothe mthe positive correlation between extr.-P and inundation frequency, summed to the
acidic pH at LP, supports the acidic range proposed for the highest P availability related to Fe/Al- and Ca-P minerals (Havlin et al., 1999; Georgantas and
ics appear d effects of flooding with the P dynambineGrigoropoulou, 2006). The comto be a major factor responsible for the zonation of the investigated mangroves.

ted how several oxidation and reduction The redox model presented here, illustrants. The “upper oxidation zone - UOZ” ehorizons can develop within the forest sedim

ons Chapter 8 – Conclusi

810

ospheric oxygen and their thickness depend upon the balanceis a result of atmbetween the rate of O2 diffusion into the sediment and the rate of O2consumption
- LOZ” develops due the (Howeler and Bouldin, 1971). The “lower oxidation zone release of oxygen from mangrove roots (Armstrong et al., 2000). The resulting
oxidation is accompanied by a reduction of pore water pH, as iron sulphides are
, 1998). Both the “upper et al.converted to (hydr)oxides and sulphuric acid (Clark reduction zone - URZ” and the “lower reduction zone - LRZ” are characterized by
tal-binding eposition of OM and sulphate reduction to produce potentially mdecomation and distribution was regulated , the pattern of pyrite formodelsulphides. In this mboidal mnute dispersed grains at LOZ and fraiby the active roots, resulting in mZ. Rre at Ltextu

The redox model revealed for A. germinans a significant greater oxidation at LP,
ot-induced oxidation capacity was oderate but constant ro the m R. manglewhereas forweakly affected by the different substrate conditions. The higher root-induced ited the availability of P and consequently their lim A. germinans oxidation ine enhancing thR. mangleoderate root-oxidation in distribution; whereas the mation of extr.-P, and therefore can be considered as an adaptive response to mforinance in this plain. that could contribute to its domnutrient uptake in water-loggingThese findings indicated a species-specific differences in oxidise the rhizosphere that
ngrove zonation. acontribute to their differential flood tolerance and m

Differences of sulphur speciation between R. mangle and A. germinans demonstrate
ngroves. These athat plant activity greatly influence sulphur cycling in Bragança mchemical modifications are in situ processes resulting from root-sediment interactions.
At LP, as result of a higher A. germinans root-oxidation capacity in comparison to R.
mangle, a SO42- increase in A. germinans occurred. Conversely, at MP, as result of a
higherR. mangle root-oxidation activity related to A. germinans, sulphide
. When low root-oxidation R. mangleconcentrations were hardly detected below ation of iron sulphides can prevent the re-supply of iron occurred, the form ation of extr.-P available for the vegetation. The low(hydr)oxides, enhancing the formted evidence with respect to tidal hydrology, but do not entary OM at LP, supporsedim

ons Chapter 8 – Conclusi

910

- Rhizophoraposition. At the more reducedclearly behaves with respect to decomdominated plain (LP), the C:N ratio rising reflected the proportionally higher increase
in org.-C than in tot.-N values.

In Bragança the SO42- reduction is a less importance diagenetic pathway (Alongi et al.
., 1999) or the reoxidation of hydrogen sulphide due to the high et aler 1998b; Holmtrations low ide concenulating pore water sulphabundance of iron keeps accum(Moeslundet al., 1994). The SO42- was largely controlled by the degree of aeration in
ents due to the intense root-oxidation and probably the ventilation by benthic sedimfauna. The DOP values obtained here were also indicative for sulphide-limiting
suboxic conditions (Raiswell et al., 1988). The DOP was largely controlled by tidal
flushing and SO42- reduction, and to a lesser extent by root oxidation and bioturbation.
c to anoxic conditions (Eh = 0 mV) oxient shift fromAt the depressions, the sedimwas reflected in SO42- and react.-Fe reduction. In principle an oxic-anoxic interface is
., 1996; et alilkin ation via polysulphide pathway (Wfavourable for pyrite formRickard and Morse, 2005). On a whole-forest scale, the presence of two oxic-anoxic
interfaces apparently excluded the possibility of a species-specific effect on sediment
sulphur and iron pools. Therefore these pools patterns depend on physical settings in etersical paramthe first place. On the basis of pyrite observations and the physicochemwithin the mangrove sediment, it was confirmed that oxidative processes and reactant
orphologies. The wide variation availability, all control the observed resultant pyrite min pyrite texture and distribution of number of microcrystal’s in framboids suggest
that caution is needed in using pyrite as redox proxy to identify depositional ents. environm

ons Chapter 8 – Conclusi

011

Chapter 9 – Bibliography

BLIOGRAPHY CHAPTER 9– BI

111

., and Spagnol, S. 1998. Fluxes of nutrients and Ajukai, T., Miller, D., Wolanski, Edissolved and particulate organic matter in two mangrove creeks in Northeastern
2Mangroves Salt Marshes.Australia., 223-230.

Almeida, S. 1996. Estrutura e florística em areas de manguezais paraenses: evidências
Goeldi, Série ense EmílioBoletim Museu Paraazônico. da influência do estuário am, 93-100. 8Ciências da Terra.

mangrove sediments. Alongi, D.M. 1988. Bacterial productivity Microbial Ecology.15, 59-79. and microbial biomass in tropical

ngrove litter on a. Effect of exported mAlongi, D.M., Boto, K.G. and Tirendi, F. 1989bacterial productivity and dissolved organic carbon fluxes in adjacent tropical nearshore sediments. Marine Ecology Progress Series.56, 133-144.

Alongi, D.M., Boto, K.G. and Robertson, A.I. 1992. Nitrogen and phosphorus cycles. InGeophysical Union: Wa: Robertson, A.I. and Alongi, D.M. (Eds.) Tshington, DC. pp. 251-292. ropical marine ecosystems. American

microbial nutrient relationships Alongi, D.M., Christoffersen, P. and Tirendi, F. 1993. The iin tropical mangrove sedimnents. fluence of forest type on Journal of
, 201-223. 171.Experimental Marine Biology and Ecology

Alongi, D.M., Ayukai,Sources, sink and export of organic carbon through a tropical, sem T., Brunskill, G.J., Clough, B.F. and Wollanski, E. 199i-enclosed delta 8a.
, 237-242. 2Mangroves and Salt Marshes.(Hinchinbrook Channel, Australia).

ar, A., Tirendi, F. And Dixon, P. 1998b. The influence ofAlongi, D. M., Sasekummastand age on benthic decomngrove forests of Malaysia. position and recycling of organic mJournal of Experimental Marine Biology and Ecologyatter in managed .
, 197-218.225

organic mAlongi, D.M., Tirendatter in intertidal sedimi, F., Trott, L.A. and Brunskill,ents of a tropical semi-enclosed delta. G.J. 1999. Mineralisation of Estuarine,
, 451-467. 48Coastal and Shelf Science.

organic mAlongi, D.M., Tirendatter in forests of the mi, F. and Cloangroves Rhizophora stylosaugh, B.F. 2000. Below-ground decom and Avicennia marinaposition of
Aquatic Botanyalong the arid coast of Western Australia. , 97-122. 68.

unskill, G.J., rttayakorn, G., Pfitzner, J., Tirendi, F., Zagorski, I., BaAlongi, D.M., Wpathways in sediments of maDavidson, A. and Clough, B.F. 2001. Organic carbon accumngrove forests in southern Thailand. ulation and mMarine Geologyetabolic .
, 85-103. 179

Chapter 9 – Bibliography

211

ttayakorn, G., Boyle, S., Tirendi, F., Payn, C. and Dixon, P. 2004. aAlongi, D.M., Wtropical maInfluence of roots and climngrove soils. Biogeochemate on miistry.69neral and trace elem, 105-123. ents storage and flux in

. 2005. pp, D.WAlongi, D.M., Pfitzer, J., Trott, L.A., Tirendi, F., Dixon, P. and KlumRapid sediment accumulation and microbial mineralization in forests of the mangrove
Science.Kandelia candel63, 605-618. in the Jiulongjiang Estuary, China. Estuarine, Coastal and Shelf

(Ed.) Chemical analyAllen, S.E. 1989. Analysis of vegetation and other organic msis of ecological materials. Blackwell Scienaterials. Intific Publications: Allen, S.E.:
Australia. pp. 46-60.

AmQUASIMEME laboratory performinot, A. and Kérouel, R. 1997. Reference mance studies 1993-1996. aterial for nutrients for the Marine Pollution Bulletin.
(1-6), 78-83.35

crogradients in the rhizosphere of iAndersen, F.O. and Kristensen, E. 1988. Oxygen mthe mangrove Avicennia marina.Marine Ecology Progress Series.44, 201-204.

grassland soils. Anderson, D.W. and ColemJournal of Soil Water Conservationan, D.C. 1985. The dynam. 40, 211-216. ics of organic matter in

ed.) A.A. BalkemAppelo, C.A.J. and Postma: Leiden. pp. 649. a, D. 2005. Geochemistry, groundwater and pollution. (2nd

Arai, M., Truckenbrodt, W., Nogueira, A.C.R., Goés, A.M. and Rossetti, D.F. 1994.entos Barreiras, deposicional dos sedimNovos dados sobre estratigrafia e ambienteBoletimNE do Pará. IV Sim de Resumpósio de Geologia da Amos Expandidos. pp. 185-187. azônia, Belém. Anais, Belém: SBG/NO.

formation of iron sulfides in a mAragon, G.T., Ovalle, A.R. and Carmangrove ecosystem, Sepetiba Bay, Brazil. ouze, J-P. 1999. Porewater dynamics and tMangroveshe
and Salt Marshes, 85-93. 3.

ArmsOxygen distribution introng, W., Cousins, D. Ar wetlands pmlsant roots trong, J., Turner, D.Wand permeability. a barriers to gnd Beckett, P.M. 2000. as–exchange
australis.with the rhizosphere: a mAnnais of Botany. 86i, 687-703. croelectrode and modelling study with Phragmites

ArtemAcademic Publishers: Dyev, V.E. 1996. Geochemordrecht. pp. 190 . istry of organic matter in river-sea systems. Kluwer

Thesis. DepartamBastos, R.C. A. 2002. Piritização de mento de Geoquímieca. Universidade Federal Flumtais-traço na Baía de Sepetiba-RJ. PhD. inense. Niterói, RJ,
Brazil. pp. 187.

Chapter 9 – Bibliography

113

Begg, C.B.M, Kirk, G.J.D., Mackenzie, A.F. and Neue, H.-U. 1994. Root-induced ,128New Phytology.iron oxidation and pH changes in the lowland rice rhizosphere. 469-477.

ecosystem dynamBehling, H., Cohen, M.C.L. and Lara, R.J. 2001. Studies on Holocene mics of the Bragança Peninsula in Northeastern Pará, Brazil. angrove
, 225-242. 167Palaeogeography, Palaeoclimatology, Palaeoecology.

Behling, H. and Costa, M.L. 2004. Mineralogy, geochemistry and palynology of modern and late Tertiary mIsland, northeastern Pará state, eastern Amangrove deposits azonia. in the Barreiras FormJournal of South American Earth ation of Mosqueiro
.Sciences, 285-295. 17

mation of dibasic calcium phosphate Bell, L.C. and Black, C.A. 1970. TransforSociety of Amdehydrate and octacalciumerica Proceedings. phosphate in 34, 583-587. slightly acid and alkaline soils. Soil Science

Benitez-Nelson, C.R. and Buesseler, K.O. 1999. Variability of inorganic and organic phosphorus turnover rates in the coastal ocean. Nature.398, 502-505.

Berrêdo, J.F. 2006. Geoquímica dos sedimentos de manguezais do Nordeste do
. Ph.D. Thesis. Centro de Estado do Pará: o exemplo do estuário do Rio MarapanimGeociências, Universidade Federal do Pará. Belém do Pará, Brazil. pp. 185.

Berner, R.A. 1970. Sedimentary pyrite formation. American Journal of Science.268,
1-23.

Cosmochimica Acta. Berner, R.A. 1984. Sedim48, 605-615. entary pyrite formation: An update. Geochimica et

Academia Brasileira de Ciências.Bigarella, J.J. 1975. The Barreiras Group in the Northeastern Brazil. 47, 365-393. Anais de

Billen,aquatic continuum G., Lancelot, from land to ocean. C. and Meybeck, M. 199In1. N, P, an: Mantoura, R.F.C, Martin, J.M. and Wd Si retention alongo thllast, e
R. (Eds.) Ocean margin processes in global change. John Wiley and Sons: New York.
pp. 19-44.

.M., Rijnders, J.H., van de , C.W., Voesenek, L.A., Banga, M., Engelaar, WBlomadaptive responses of plants to submSteeg, H.M. and Visser, E.J. 1994. Physiological ecology of riverside speciesergence. Annais of Botany.74, 253-263. :

Baltic.Boesen, C. and PostmAmerican Journal of Sciencea, D. 1988. Pyrite form.288, 575-603. ation in anoxic environments of the

Boströmphosphorus across the sediment-water interface. , B., Andersen, J.M., Fleischer, S. and Jansson, M. 1988. Exchange of Hydrobiologia.170, 229-244.

Chapter 9 – Bibliography

Boudreau, B.P. and Jørgensen, B.B. 2001. The benthic boundary layer: transport rd Univ. Press: Oxford. pp. 404.istry. Oxfoprocesses and biogeochem

411

Bourclassificatioman of “laterite”. n, R.P. and Ollier, C.D. 2002. A critic of thCatena.47, 117-131. e Schelmann definition and

Northern Australian mBoto, K.G. and Bunt, J.S. 1981. Tidal exportangrove system. Estuarine, Coastal and Shelf Science. of particulate organic matter from13, 247- a
255.

Boto, K.G., Bunt, J.S. and Wproductivity in Northern Australia and Papua New Guinea. ellington, J.T. 1984. Variation in maEstuarine, Coastal angrove forest nd
321-329. 19,Shelf Science.

Boto, K.G. and WNorthern Australian mellingaton, J.T. 198ngrove forest. 3. Phosphorus and nMarine Ecology Progress Series.itrogen nutritio11nal statu, 63-69. s of a

Boto, K.G. and Wnorthern Australian meallingngrove forest. Estuaries.7ton, J.T. 1984. Soil characteristics and nutrient status in a , 61-69.

variation in concentrations and fluxes llington, J.T. 1988. Seasonal eBoto, K.G. and Wmangrove waterway. of dissolved organic and inorganic mMarine Ecology Progress Series. aterials in a tro50, 151-160. pical, tidally-dominated,

Boto, K.G., Alongi, D.M.M and Nott, A.L. 1989. Dissolved organic carbon and bacteria interactions at sediment-water interface in a tropical mangrove system.
, 243-251.51Marine Ecology Progress Series.

Nordbrasilien).Chemie der Erde.7Brandt, F. 1932. Ein neuer Typ von Eisentonerde-Phosphate-Vorkomme, 383-425. n (Maranhão,

Brown, J.L. 1981. Calciumions on hydroxyapatite crystal growth. Soil Science Society of America. phosphate precipitation: effects of common and foreign 45, 482-487.

: solubility and pH calculations. WButler, J.N. 1998. Ionic Equilibriumiley-Interscience Publication: Canada. pp. 543.

iron (II) monosulfide byButler, I.B. and Rickard, D. 2000. Fram hydrogen sulphide. boidal pyrite formaGeochimica et Cosmochimica Actation via the oxidation of.64,
2665-2672.

bianno, M.J. 1997. Tidal flushing of Caetano, M., Falcão, M., Vale, C. And Beammonium, iron and manganese from inter-tidal sediment pore waters. Marine
.Chemistry, 203-211. 58

anoxic maCanfield, D.E. and Bernrine sediments. Geochimica et Cosmochimica Actaer, R.A. 1987. Dissolution and pyritization of m.51, 645-659. agnetite in

Chapter 9 – Bibliography

Canfield, D.E. 1989. Reactive iron, 619-632. 53Cosmochimica Acta.

511

in marine sediments. Geochimica et

seawater solutions. Cappellen, P.van and Berner, R.A. 1991. Fluorapatite crystal growth fromGeochimica et Cosmochimica Acta.55, 1219-1234. modified

Caraco, N.F., Cole, J.J. and Likens, G.E. 1989. Evidence for sulphate-controlled phosphorus release from sediments of aquatic systems. Nature.341, 316-318.

Clark, M.Wand heavy m.,e McConchie, D., Lewis, D.Wtal partitioning in Avicennia-dominated ma. and Saenger, P. 1998. Redox stratification ngrove sediments: a
geochemical model. Chemical Geology.149, 147-171.

Chang, S.C. and Jackson, M.L. 1957. Fractionation of soil phosphorus. Soil Science.
, 133-144. 84

Cheng, K.L. and Da-Ming, Z. 2005. On calibration of pH meters. Sensors.5, 209-219.

Chesnin, L. and Yien, C.H. 1950. TurbidimSoil Science American Society Proceedings.15etric determ, 149-151.ination of available sulfates.

contribution to the understanding of the formChristoffersen, J., Christoffersen, M.R., Kibalczyc, Wation of calcium. and Andersen, F phosphates. .A. 1989. A Journal of
, 767-777.94Crystals Growth.

Christoffersen, J. and Christoffersen, M.R. 1982. Kinetics of dissolution of calciumJournal of Crystal Growth.57, 21-26.hydroxyapatite: V. The acidity constant for the hydrogen phosphate surface complex.

.W.Clark, M.W., Mc Conchie, D., Lewis, D and Saenger, P. 1998. Redox stratification and heavy metal partitioning in Avicennia-dominated mangrove
sediments: a geochemical model. Chemical Geology.149, 147-171.

waters.LimCline, J.D. 1969. Spectrophotomnology and Oceanography.etric determ14, 454-458.ination of hydrogen sulfide in natural

Hinchinbrook Channel, Australia. Clough, B. 1988. Mangrove forests productivity and biomMangrove Salt Marsh.2, 191-198. ass accumulation in

variability of mCohen, M.C.L, Lara, R.J, Ramagnesium, calciumos, J.F.F and Dittm and potassium in waters of a mar, T. 1999. Factors inangrove creek in fluencing the
Bragança, North Brazil. Mangroves and Salt Marshes. 3, 9-15.

model applied to maCohen, MC.L., Lara, R.J., Szlafsztein, C.F. and Dittmngrove coast analysis, Amazon region, Brazil. ar, T. 2001. Digital elevatioJournal of n
Interna, 1-8. 4l Environment Creation.tiona

Chapter 9 – Bibliography

611

and nutrient dynamics fromCohen, M.C.L., Lara, R.J., Szlafsztein, C. and Dittm a GIS perspective. ar, T. 2004. MangroWetlands Ecology and Managementve inundation .
, 81-86. 12

Colmradial oxygen loss fromer, T.D., Gibberd, M.R., W roots of rice (iOryza sativaengweera, A. and Tinh, T.K. 1998. The barrier to L.) is induced by growth in
, 1431-1436. 49.Journal of Experimental Botanystagnant solution.

Cordeiro, C.daC., Mendoza, U.N. and Lara, R.J. 2003. Mangrove zonation and ent along an inundation gradient in Bragança, North phosphorus distribution in sedim26 SetiemBrazil. X COLACMAR. Décibre. San José de Costa Rica. pp. 376. mo Congreso Latinoamericano de Ciencias del Mar. 22-

mineralização da mCordeiro, C. daC. and Maetéria orgânica no mandoza, U. 2007. Relações C/N, C/nguezal de Bragança, Norte do Brasil. XI P e N/P indicando fonte e
cano de Ciencias del Mar. 15-20, eriCOLACMAR. Onceavo Congreso LatinoamAbril. Florianópolis, SC, Brasil. pp 489.

de Geociências.21Costa, M.L. 1991. Aspectos geológic(2), 146-160. os dos lateritos da Amazônia. Revista Brasileira

bearingCosta, M.L. and Araújo lateritic crusts for iden, E.S. 1996. Applicationtification of their pa of murent rocks.lti-elem ents in Au-phosphate-Journal of Geochemical
, 257-272.57Exploration.

Costa, M.L., Fernández, O.C., Toledo, M.C.M., Passos, C.M. and Pereira, P.F. 2004a. A turquesa de Itacupim, Pará. Revista da Escola de Minas.57(4),261-266.

Costa, M.L., Behling, H., Berrêdo, J.F., do Carmo, M.S. and Siqueira, N.V.M. 2004b. sediments from NortheasteMineralogical, geochemical and palynologirn Pará State, Brazil. cal studies of late holocene mRevista Brasileira de Geociênciasangrove .
(4), 479-488.34

samCottampling. Ecology.37, G. and Curtis, J.T. 1956. The use of, 451-460. distance measures in phytosociological

Cundell, A.M., Brown, M.S., Standford, R. and Mitchell, R. 1979. Microbial degradation of Rhizophora mangle leaves immersed in the sea. Estuarine and Coastal
Marine Science, 281-286. 9.

dissolved and particulate maDagg, M., Benner, R., Lohrenz, S. and Lawrence, D. 2004. Transforterials on continental shelves influenced by large rivmation of ers:
plume processes. Continental Shelf Research.24, 833-858.

706.Davies, D.D. 1986. The fine control of cytosolic pH. Physiology of Plants.67, 702-

Chapter 9 – Bibliography

711

zur WDeutsche Inastitut für Normsser-, Abwasser- und Schlammung-DI-Untersuchung: PN 38405-D26. 1993. Deutsche Einhysikalische, chemheitsverfahrenische,
gelösten Sulfids. 2biologische und bakteriologische Verf, 2-10.ahren. Photometrische Bestimmung des

Deutsche InstituWasser-, Abwasser- und Schlammt für Normung-DIN 38414-S-Untersuchung: Physikalische, chem2. 1993. Deutsche Einheitsverfahren zuische, r
Trockenrückstandes bzw. der Trockensubstanz. biologische und bakteriologische Verfahren. Bestimmung des W4, 2-7.assergehaltes und des

Deutsche InstituWasser-, Abwasser- und Schlammt für Normung-DIN 38414-S-Untersuchung: Physikalische, chem3. 1993. Deutsche Einheitsverfahren zuische, r
des Glühverlustes des Trockenmbiologische und bakteriologische Verfahren. Bestimmasse eines Schlammes. 4, 2-5.ung des Glührückstandes und

ImDe Simpact of root morphology on meone, O., Haase, K., Müller, E., Junk, W.J., Gonsior, G. and Schmtabolism and oxygen distribution in roots and idt, W. 2002.
Biology.rhizosphere from29, 1025-1035. two Central Amazon floodplain tree species. Functional Plant

in stream water as influenced by pH and calciumDiaz, O.A., Reddy, K.R. and Moore Jr, P.A. 1994. Solubility of inorganic phosphorus concentration. Water Research.
(8), 1755-1763. 28

Dittmar, T. 1999. Outwelling of organic matter and nutrients from mangrove in North
Brazil: evidence from organic tracers and fluxes measurements. Ph.D. Thesis.
Zentrum für Marine Tropenökologie. Bremen Universität. Bremen, Germany.pp. 206.

Dittmar, T., Lara, R.J. and Kattner, G. 2001. River or mangrove? Tracing major
organic matter sources in tropical Brazilian coastal waters. Marine Chemistry.73,
253-271.

Dittmdynamics in a mar, T. and Lara, Ra.ngrove tidal creek in North Brazil. J. 2001a. Driving forces behind nutrienEstuarine, Coastal and Shelf t and organic matter
, 249-259. 52Science.

Dittmar, T. and Lara, R.J. 2001b. Do mangroves rather than rivers provide nutrients to
coastal environments South of the Amazonas River? Evidence from long-term flux
measurements. Marine Ecology Progress Series.213, 67-77.

Dittmreducing mangrove sedimar, T. and Lara, R. 2001c. Molecular events (Amazônia, Brazil). idence for lignin degradatioGeochimica et Cosmochimica n in sulfate-
, 1417-1428.65Acta.

Dorozhkin, S.V. 2002. A review on the dissolution mProgress in Crystal Growth and Characterization of Materials.44odels of calciu, 45-61. m apatites.

Chapter 9 – Bibliography

118

ss aical mond, J.M., Boyle, E.A., Grant, B. and Stallard, R.F. 1981. The chemEdmbalance in the Amazon plume I: The nutrients. Deep Sea Research.11(28A), 1339-
1374.

Ellison, A.physiology, growth, and reproduction of red mM. and Farnsworth, E.J. 1997. Simualated sea levngrove (Rhizophora mangleel change alters anatomy, L.)
, 435-446. 112Oecologia.

Ensminger, I. 1996. Hydrologische Veränderungen amBedeutung für die Regeneration salzgeschädigter Mangrove. Diplom Canal Clarín und ihre arbeit. Justus-
any. pp. 63. Liebig-Universität Gießen, Gießen. Germ

Environmorthophosphate in estuarine and coastal waters by autometal Protection Agency – EPA 365.5. 1997. Determated colorimetric analysis. ination of
d.) National Exposure Research Laboratory: Ohio. pp. 1-9.Arar, E.J. (E

neral das plantas: princípios e perspectiva. iEpstein, M. 1975. Nutrição mUniversidade de São Paulo: São Paulo. pp. 87.

Etcheber, H., Bourg, A.C.M. and Donard. O. 1983. Critical aspects of selective extractionand organic mas of trace metals fromtter interactions. Poc. Int. Conf. Heavy Me estuarine suspended matter. Fe and Mntals in the Environm hydroxides ent,
Heidelberg, CEP Consultants (Eds.): Edinburgh. pp. 1200-1203.

Hydrobiologia. 228, 37-42. Fabre, A.C. 1992. Inorganic phosphate in exposed sediments of the river Garonne.

sedimFabre, A., Froments of four representative mangrove stages (French Guiana). ard, Fr. and Trichon, V. 1999. Fractionation of phosphate in Hydrobiologia.
, 13-19. 392

phill Pub. Co.: Austin Tex. mentary rocks. HeFolk, R.L. 1974. Petrology of sedimpp.182.

Freemonto calcite. Journal of Soil Science.32, 75-84. an, J.S. and Rowell, D.L. 1981. The adsorption and precipitation of phosphate

soils of forest fallows and primFrizano, J., Vann, D.R., Johnson, A.H. and Johnson, C.M. 2003. Labile phosphorus iary forest in the Bragantina region, Brazil. Biotropican.
, 2-11. 35

gnésio, sódio, potássio e da aFurtado da Costa, M. 2000. Estudo dos cátions cálcio, msalinidade na água intersticial do sedimMaster Thesis. Centro de Geociências. Universidade Federal de Pará. Belémento do manguezal de Bragança - NE do Pará. do Pará.
Brazil. pp. 87.

salinidade na água intersticial do sedimFurtado da Costa, M. 2000. Estudo dos cátions cálcio, mento do maanguezal de Bragança - NE do Pará. gnésio, sódio, potássio e da

Chapter 9 – Bibliography

911

do ParMaster Thesis. Centro de Geociências. Universidade Federal de Pará. Belémá. Brazil. pp. 87.

removal from synthetic wastewater using alum and alumGeorgantas, D.A. and Grigoropoulou, H.P. 2006. Phosphorus and organic minum hydroxide. aGlobaltter
(10), 1-9.10.NEST Journal

control the salinity, comGibbs, R.J. 1967. The geochemistry of the Amposition and concentration of the suspended solids. azon River system: the factors that Bulletin of
78erica.mGeological Society of A, 1203-1232.

Ocean.Gibbs, R.J. 1970. Circulation inJournal of Marine Research. the Am28, 113-123. azon River estuary and adjacent Atlantic

GeomicrobiologyGiblin, A.E. 1988. Pyrite form.6, 77-97. ation in marshes during early diagenesis. Journal of

mangleGill, A.M. and Tom L.). The adult root system. linson, P.B. 1977. Studies on the red mBiotropica.9, 145-155. angrove (Rhizophora

Gleason, S.M., Ewel, K.C. and Hue, N. 2003. Soil redox conditions and plant-soil 56relationsh, 1065-1074. ips in a Micronesian mangrove forest. Estuarine Costal and Shelf Science.

o, P.M. 1990. ModeloGoés, A.M., Rosetti, D.F., Nogueira, A.C.R and Toleddo Museu Pdeposicional prelimaraense Emílio Goeldiinar da form.19, 3-15ação Pirabas no Nordeste do Estado do Pará. Boletim

an, C.E., Echlin, P., Lifshin, E., Goldstein, J.I., Newbury, D.E., Joy, D.C., Lymmicroanalysis. (Third EditionSawyer, L. and Michael, J.R. 2003 Scanning electron m) Kluwer Academiic/Plenum Publishers: New York. pp. croscopy and x-ray
689.

Golterman, H.L. and Meyer, M.L. 1985. The geochemistry of two hard water rivers,
hydroxy-apatite.the Rhine and the Rhone. Part 4: The determHydrobiologia.126, 21-24. ination of the solubility product of

GoltermHydrobiologia.159an, H.L. 1988. The calcium, 149-151. and iron bound phosphate phase diagram.

ez, E., Durillon, C., Rofes, G. and Picot, B. 1999. Phosphate adsorption and Gomrelease from sediments of brackish lagoons: pH, O2 and loading influence. Water
, 2437-2447. 33.Research

Goñi, M.A. and Hedges, J.I. 1995. Sources and reactivities of marine-derived organic
matter in coastal sediments as determined by alkaline CuO oxidation. Geochimica et
, 2956-2981.59Cosmochimica Acta.

Chapter 9 – Bibliography

012

Gotoh, S. and Patrick, Jinfluenced by redox potential and pH. R. W.H. 1974. TransforSoil Science American Proceedingsmation of iron in a waterlogged soil as .38, 66-70.

oxidation of organic mGoulden, P.D. and Antahony, D.H.Jterial in fresh water. . 1978. Kinetics of uncAnalytical Chemistry.50atalyzed peroxydisulfate, 953-958.

Grasshoff, K., Ehrhardt M. and Kremmling, K. 1983. Methods of seawater analysis. Verlag Chemie: Nürnberg. pp. 403.

GrimS.E. (Ed.) Chemshaw, H.M., Allen, S.E. and Pical analysis of ecologicala mrkinson, J.A. 1989. Nutrient elematerials. Blackwell Scientific ents. In: Allen,
Publications: Australia. pp. 81-159.

of apatite in seawater. Gulbrandsen, R.A., Roberson, C.E. and Neil, S.T. 1984. TimGeochimica et Cosmochimica Acta.48, 213-218. e and the crystallization

Gunnars, A., Blomation of apatite in qvist, S. and Martinsson, C. 2004. Inorganic formbrackish seawater from the Baltic Sea: an experimental approach. Marine Chemistry.
, 15-26. 91

Haase, K., De Simtransport in cuttings fromone, O., Junk, W.J. and Schm flood-adapted várzea tree species. idt, W.Tree Physiolog 2003. Internal oxygeny.23 ,
1069-1076.

fertilizers. 6th Edition. PrHavlin, J.L., Beaton, J.D., Tisdale, S.L. and Nelson, Wentice Hall : Upper Saddle River, NJ. pp. 499. .L. 1999. Soil fertility and

critical examHaynes, R.J. 1990. Active ion uptake and mination of their role in regulating rhizosphere pH. aintenance of cation-anion balance: A Plant and Soil.126,
247-264.

containings solids. Hedges, J.I. and Stern, J.H. 1984. Carbon and nitrogen determLimnology and Oceanography.29, 657-663.inations of carbonate-

A., Quay, P.D., Richey, J.E., Devol, A.H. and Santos, U.de M. .Hedges, J.I., Clark, W1986. Compositions and fluxes of particulate organic material in the Amazon River.
, 717-738. 31Limnology and Oceanography.

itation of barium tter on the precipaHesse, P.R. 1957. The effect of colloidal organic msulphate and a modified method for determining soluble sulphate in soils. Analyst.82,
710-712.

mangrove sHesse, P.R. 1961. Somwameps in Sierra Leone. differences between the soils of Plant and Soil.14, 335-346. Rhizophora and Avicennia

ical analysis. Murray: London. 513 pp. Hesse, P.R. 1971. A textbook of soil chem

Chapter 9 – Bibliography

calcareous sedimHieltjes, A.H.M. and Lijklements. Journal of Environmental Quality. a, L. 1980. Fractionation of inorganic phosphate in 9, 405-407.

112

sedimHinnes, M.E., Knollmentary biogeochemeyistry in a northern New England salt mer, S.L. and Tugel, J.B. 1989. Sulfate reduction and otherarsh. Limnology and
, 578-590. 34.Oceanography

ngroves. Oxford University Press Inc: New aHogarth, P.J. 1999. The biology of mYork. pp. 228.

in sedimHouse, W.ents. Biogeochemsitry.63A. 2003. Factors influencing the extent and developm, 317-333. ent of the oxic zone

, N. 1999. mboe, N., Kristensen, E. and Thongthamer, M., Andersen, F., HolmHolTransformbed systemation and exchange processes in, Thailand. Seasonal and spatial variations in benthic m the Brangrong maengrove forest-seagrasstabolism and
sulfur biogeochemistry. Aquatic Microbial Ecology.20, 203-212.

ecosystem metabolismHowarth, R.W. 1979. Pyrite: Its rapid form. Science.203, 49-51. ation in salt marsh and its importance in

submerged soils. Howeler, R.H. and Bouldin, D.R. 1971. The diffusion and consumSoil Science Society America Proceedings.35, 202-208. ption of oxygen in

pyrite from anoxic mHuerta-Diaz, M.A. 1989. Geochemarine environments.istry of trace mePh.D. Thesis. Texas University. Texas, tals associated with sedimentary
USA. pp. 299.

Huguenin-Elie, O., Kirk, G.J.D. and Frossard, E. 2003. Phosphorus uptake in rice fromEuropean Journal of Soil ned or flooded then drained. soil that is flooded, drai, 77-90. 54Science.

DF. Brazil. pp. 56.INMET, 1992. Normas climatológicas. Instituto Nacional de Meteorología, Brasília.

Jander, G. and Jahr, K.F. 1963. Mass Analyse: Theorie und Praxis der klasischen und ischen Tirierverfahren. Gruyter: Berlin. pp. 359. der elektrochem

Jeroschewski, P., Steuckart, C. and Kühl, M. 1996 An amperometric microsensor for
the determination of H2S in aquatic environments. Analytical Chemistry.68, 4351-
4357.

Denmark). Jorgensen, B.B. 1977. The sulfur cycle of coastal mLimnology and Oceanography.5, 814-832. arine sediments (Limfjorden,

plant response to soil flooding. Justin, S.H.F. and Armstrong, W.New Phytologist 1987. The anatom.106, 465-495. ical characteristics of roots and

Chapter 9 – Bibliography

212

nn, C.D. 1993. Total dissolved nitrogen and iKarl, D.M., Tien, G., Dore, J. and Wation ALOHA: Redfield reconciliation. phosphorus concentrations at US-JGOFS St, 203-208.41Marine Chemistry.

Karl, D.M. and Yanagi, K. 1997. Pphosphorus pool in the oligotrophic North Pacific. artial characterization ofLimnology and Oceanography. the dissolved organic 42,
1398-1405.

ith in seawater: Poisoning wKattner, G. 1999. Storage of dissolved inorganic nutrients mercuric chloride. Marine Chemistry.67, 61-66.

Khalid, R.A., Patrick, Wcharacteristics of flooded soils. .Soil Science SoH, and DeLaune, R. D. 1977. Phosphorus sorption ciety of America.41, 305-310.

ical Hydrogeology. Prentice Hall : Upper SaddleKehew, A.E. 2001. Applied ChemRiver, NJ. pp. 368.

from plant roots. European Journal of Soil Science.Kirk, G.J.H. 1999. A model of phosphate solubilization by organic anion excretion 50, 369-378.

Kirk, G.J.D. 2003. EffecEuropean Journal of Soil Science.ts of flooding and re54, 617-624.dox conditions on solute diffusion in soil.

istry of submerged soils. John WKirk, G. 2004. The biogeochem: iley and SonsEngland. pp. 291.

Mangrove EcosystemKjerfve, B. and Lacerda, L.D. 1993. Mangroves of Brazil. s Technical Reports. ITTO TS-13. pp. 245-272.In: Lacerda, L.D. (Ed.),

ecosystems. InKjerfve, B. and Macintosh, D. J. 1997. The impact of climatic change on m: Kjerfve B., Lacerda L.D. and Diop E. H.S. (Eds.) Mangrove angrove
erica and Africa. UNESCO: Paris. pp. 1-7. studies in Latin Amecosystem

Kjerfve, B., Lacerda, L.D. and Diop, E.H.S. 1997 Mangrove ecosystem studies inerica and Africa. UNESCO: Paris. pp. 349. Latin Am

oxygen exchange and growth in rice. Kludze, H.K. and DeLaune, R.D. 1995. Straw application effects on mJournal of Soil Science Society of Americaethane and .59,
824-830.

Kolowith L.organic phosphorus. C., Ingall E.Limnology and Oceanography. D. and Benner R. 2001. Com46positio, 309-320. n and cycling of marine

Komiyama, A., Ogino, S., Aksornkoae S. and Sabhasari, S. 1987. Root biomass of a
thod and the engrove forest in Southern Thailand. 1. Estimation by the trench mamzonal structure of root biomass. Journal of Tropical Ecology.3, 97-108.

Chapter 9 – Bibliography

ical adaptations to waterlogging in roots of wetland Koncalová, H. 1990. Anatomgraminoids: limitation and drawbacks. Aquatic Botany.38, 127-134.

312

Konsten, C., Van BreemeEffects of flooding on pH of rice-producing, acid sulfate soils in Indonesia. n, N., Suping, S., Aribawa, I. and Gronenberg, J. 1994. Journal of
erica.mSoil Science Society of A 871-883. 58,

KremKoroleff, F. 1983. Determling, K. (Eds.) 2nd Edition.ination of phophorus. Methods ofIn: seawater analysis.Grasshoff, K., Ehrhardt, M. and Verlag Chemie:
. pp. 125-139. Weinheim

. 1994. Partitioning and speciation of solid phase iron in Kostka, J. E. and Luther, G.Wsaltmarsh sediments. Geochimica et Cosmochimica Acta.58, 1701-1710.

Physiology . 1, 1-29. Kozlowski, T.T. 1997. Responses of woody plants to flooding and salinity. Tree

ecosystems: the Bragantinian mKrause, G., Schories, D. and Diele, K. 2001. Spatial patterns of mangrove of Northern Brazil (Bragaança, Pará).ngrove
, 93-107. 7Ecotropica.

dynamKrause, G. and Glaser, M. 2003. Co-evolics in a coastal fishing village of the Brving geomagança region (Pará, North Brazil). orphical and socio-economic
, 859-874. 46Ocean and Coastal Management.

tabolism and sulfate eer, M. and Bussarawit, N. 1991. Benthic mmKristensen, E., Holreduction in a southeast Asia mangrove swamp. Marine Ecology Progress Series.73,
93-103.

er, M., Banta, G.T., Jensen, M.H., Hansen, K. and Kristensen, E., King, G.M., Holmof Ao Nam Bor maBussarawit, N. 1994. Sulfate reduction, acetate and carbon ngrove, Phuket. Thailand. meMarine Ecology Progress Seriestabolism in sedim.109ents ,
245-255.

Kurmies, B. 1972. Zur Fraktionierung der Bodenphosphate. Phosphorsäure.29, 118-
149.

crosensor iS mKühl, M., Steuckart, C., Eickert, G. and Jeroschewski, P. 1998. A H2for profiling biofilms and sediments: application in an acidic lake sediment. Aquatic
, 201-209. 15ology.cMicrobial E

Küster, F.W. and Thiel, A. 1982. Rechentafeln fische Analitik. Berlin. pp. ür die chem102.

Chapter 9 – Bibliography

412

Lacerda,1993. The biogeochem L.Dl., Carvalhistryo, C.E.V., Tanizak and trace metal dii, K.F., Ovalle,stribution of mangrove rhizo A.R.C. and Rezende, C.Espheres. .
, 252-257. 25.Biotropica

soil organLacerda, L.ic maD., Ittekotter: a cot, V. and Patchineelmparison betweenam Rhizophora, S.R. 1995. Biogeochem and Avicenniaistry of ma soils in South-ngrove
, 713-720. 40.Estuarine, Coastal and Shelf Scienceeastern Brazil.

between environmLallier-Verges, E., Perrussel, B.P., Disnar, J-Rental conditions and the diagenetic evolution of organic m and Baltzer, F. 1998. Relationshipsatter
French West Indies). derived from higher plants in a modern mOrganic Geochemistry.29a, 1663-1686.ngrove swamp system (Guadaloupe,

particulate nitrogen and phosphorus. Langner, C.L. and Hendrix, P.F. 1982. EvaluatiDeep Sea Research.16on of a persulfate digestion m, 1451-1454. ethod for

Lara, R.J. and DittmBrazil) during the dry season. ar, T. 1999. NutrienMangrove and Salt Marshes.t dynamics in a m3a, 185-195. ngrove creek (North

ignition meLeeg, J.O. and Black, Cthod.. Soil Science Society ProceedingA. 1955. Determs.19ination of organic phosphor, 139-143. us in soils: II

sedimLeventhal, J.S. 1983. An interpents as indicators of deposition environments. retation of carbon and sulfur relationships in Black Sea Geochimica et Cosmochimica
, 133-138. 47Acta.

pyritization.Leventhal, J. and Taylor, C. 1990. ComGeochimica et Cosmochimica Actaparison of the m.54,e 2621-2625. thods to determine degree of

Inc.: Canada. pp. 715. Libes, S.M. 1992. An introduction to marine biogeochemistry. John Wiley and Sons,

Lindsay, W.L. 1979. Chemical equilibria in soils. John Wiley and Sons: New York,
USA. pp. 449.

J.B. and WLindsay, W.eed, S.B. (Eds.) 2nd editioL., Vlek, P.L.G. and Chien, S.H. 1989. Phosphate mn. Minerals in Soil Environminerals. ents. Soil Science In: Dixon,
s. pp. 1089-1130. ierica: Madison, WSociety of Am

sedimentary materials. Lord, C.J. 1982. A selective and precise mJournal of Sedimentary Peetrologythod for.52, 664-666. pyrite determination in

Love, L.G. and AmChattanooga Shale and Rastutz, G.C. 1966. Review of mmmelsberg Banderz. Foirtschrift Mineralogiecroscopic pyrite from.43 the Devonian , 273-309.

from the Wenlockian Denbigh Grit Love, L.G. 1971. Early diagenetic polyframboiGroup, Conway, North Wdal pyrite, primales, UK. ary and redepositedJournal of
.Sedimentary Petrology, 1038-1044. 41

Chapter 9 – Bibliography

512

Ecology and Systematic.5Lugo, A.E. and Snedaker, S.C. 1974. The ecology of m, 39-64. angroves. Annual Review of

warth, R.W. and Ryans, R.A. 1982. Pyrite and oxidized Luther, G.W., Giblin, A., Hoiron mineral phases formed from pyrite oxidation in salt marsh and estuarine
sediments. Geochimica et Cosmochimica Acta.46, 2665-2669.

Luther, G.W., Kotsky, J.E., Church, T.M., Sulzberger, B. and Stumm, W. 1992. ligand comSeasonal iroplexes withn cycling in the salt-m Fe(II) and Fe(III) in arsh sedimthentary enve dissolutioironmn of Fe(III) ment: the importance oinerals andf
pyrite, respectively. Marine Chemistry.40,81-103.

Madureira, M.J., Vale, C. and Gonçalves, S.M.L. 1997. Effect of plants on sulphur geochemistry in the Tagus salt-marshes sediments. Marine Chemistry.58, 27-37.

Marchand, C., Lallier-Vergès, E. and Baltzer, F. 2003. The composition of sedimentary organic matter in relation to dynamic features of a mangrove-fringed
coast in French Guiana. Estuarine, Coastal and Shelf Science.56, 119-130.

and Albéric, P. 2004. Pore-water Marchand, C., Baltzer, F., Lallier-Vergès, Echemdevelopmistry in mental stages (Fangrove sedimrench Guiana). ents: relationsMarine Geology.hip with s208pecies, 361-381. composition and

ic press: London. Marschner, H. 1995. Mineral nutrition of higher plants. Academpp.889.

Martenenvironment by magnesiums, C.S. and Harriss, R.C. 1970. Inhibition ions. Geochimica et Cosmochimica Acta of apatite precip.34itation in the m, 621-625. arine

promises and problemMartin, J. M., Nirel, P. and Thoms. Marine Chemistry.as, A.22, 313-341. J. 1987. Sequential extraction techniques:

zonation and soil redox statMatthijs, S., Tack, J., van Speybroeck, D.e , sulphide concentration and Koedam,and salinity in Gazi Bay N. 1999. Mangrove species
, 243-249. 3Mangroves and Salt Marshes.inary study. (Kenya), a prelim

erg, K.C. and Aller, R.C. bMayer, L.M., Keil, R.G. Macko, S.A., Joye, S.B., Ruttemnitrog1998. Imen to coastal zones. portance of suspended particulatesGlobal Biogeochemical Cycles.12 in riverine delivery of bioavailable , 573-579.

ination of pore m. 1988. Reexand Hester, M.WMcKee, K.L., Mendelssohn, I.A. amanglewater sulfid and e concentratiAvicennia germinans.ons and redox potentials near the aerial roots of American Journal of Botany.75, 1352-1359. Rhizophora

reciprocal effects? McKee, K.L. 1993. Soil physicochemJournal of Ecology. 81ical patterns and m, 477-487. angrove species distribution -

Chapter 9 – Bibliography

612

A., Peres, A., Giarrizzo, T., Menezes, M., Carvalho, LM., Carvalho, E.Medina, E.,Silva, B., Vilhena, R., Reise, A. and Braga, F.C. 2001. Mangal communities of the through soil and leaf analyses. “Salgado Paraense”: Ecological heterogeneAmazoniana.16(3/4), 397-416. ity along the Bragança Peninsula assessed

time-course investigation of soil waterlogging effects. Mendelssohn, I.A. and McKee, K.L. 1988. Spartina alterniflora die back in LousianaJournal of Ecology.76, 509-:
521.

growth patterns of the mMenezes, M., Berger, U. and Waongrove trees fromrbes, M. 2003. Annual gro the Bragança Peninsula, North Brazil.wth rings and long-term
Wetlands Ecology and Management, 233-242.11.

Brazil. Ph.D. Thesis. ZentrumMenezes, M. 2006. Investigations of ma für Marine Tropenökologie. Bremngrove forest dynamics in Amen Universität, azonia, North
any. pp.149. Germ

hydrogen sulfide. Meyer, B., Ward. K., Koshalp, K. and Peter,Inorganic Chemistry.22, 2345-2346. L. 1983. Second dissociation constant of

sedimMeyers, P.A. 1994. Prentary organic matter. eservation of elemChemical Geology.114, 289-302.ental and isotopic source identification of

biogeochemMiddelburg, J.J., Nieuwenhuize, J., Slimistry in an East African ma, F.J. and Ohowa, B. 1996. Sedimngrove forest (Gazi Bay, Kenya). ent
, 133-155. 34istry.Biogeochem

in seawater. Millero, F.GLimnology and Oceanography.33., Plese, T. and Fernandez, M. 1988. The dissociation of hydrogen sulfide , 269-274.

1975. Upper continental .aral, C.A. and Francisconi, Oan, J.D., Barreto, L., AmMillimmargin sedimentation of Brazil: Northern Brazil. Contribution of Sedimentology. 4,
11-43.

coastal sedimMoeslund, L., Thamdrup, B. and Jent: radiotracer studies and seasonal dynamorgensen, B.B. 1994. Sulfur and iron cycling in a ics. Biogeochemistry. 27,
129-172.

Morris, A.W., Bale, A.J., and Howland, R.J.M. 1981. Nutrient distribution in an estuary: evidence of chemical precipitation of dissolved silicate and phosphate.
, 205-216. 12oastal and Shelf Science.Estuarine, C

Morita, T. and Viegas, R.M.A. 1981. Manual de soluções, reagentes e solventes. dgard Blücher Ltda.: São Paulo. pp 592. Padronização, preparação e purificação. E

soils.MurrmSoann, R.P. and Peech, M. 196il Science Society of America9. Effect of Proceedings.33 pH on labile and so, 205-210.luble phosphate in

Chapter 9 – Bibliography

712

ie für Techniker und Ingenieure. Dt. Verl. mNäser, von K-H. 1976. Physikalische Chefür Grundstoffindustrie: Leipzig. pp. 480.

Nedwell, D.B., Blackburn, T.H. and Wturnover of organic carbon, nitrogen and suiebe, W.J. 1994. Dynalphur in the sedimmients of a Jamc nature of the aican
mangrove forest. Marine Ecology Progress Series.110, 223-231.

held, ann, G., Massonneau, A., Langlade, N., Dinkelaker, B., Hengeler, C., RomNeumphosphorus-deficient white lupins. AV. and Martinoia, E. 2000. Physiological aspects of cluster root and developmnnais of Botany.85, 909-919. ent in

sociation between pore water sulphide Nickerson, N.H. and Thibodeau, F.R. 1985. Asconcentrations and the distribution of mangroves. Biogeochemistry.1, 183-192.

Canadian Journal of Earth Sciences.Nriagu, J.O. 1976. Phosphate - clay m13i, 717-735. neral relations in soils and sediments.

intake of the mNordhaus, I., Waongrove crab lff, M. and Diele, K. 2006. LiUcides cordatus in a high intter processing and population food tertidal forest in Northern
67.Estuarine, Coastal and Shelf ScienceBrazil., 239-250.

Nye, P.H. 1979. Diffusion of ions and uncharged solutes in soils and soil clays. 31.Advanced in Agronomy, 225-272.

Influence of draining on soil Olila, O.G., Reddy, K.R. and Stites, D.L. 1997. phosphorus forms and distribution in a constructed wetland. Ecological Engineering.
, 157-169. 9

Oliveira, S. R. 1999. Ciclagem do fósforo em uma floresta de manguezal. Natal/RN.
Msc. Thesis. Universidade Federal do Rio Grande do Norte. Bioecologia Aquática. Brazil. pp. 43.

Orion Research. 2003. Model 94-16 Ionplus. Silver/Sulphide Electrode Instruction. ctron Corporation: Boston. pp. 67. o EleTherm

Orrett, K. and Karl, D.M. 1987. Dissolved organic phosphorus production in surface , 383-395. 32Limnology and Oceanography.seawater.

nde, C.E., Lacerda, L.D. and silva, C.A.R. 1990. Factors Ovalle, A.C.R., RezeEstuarine, Caffecting the hydrochemoastal and Shelf Scienceistry of a m.31a, 639-650. ngrove tidal creek, Sepetiba Bay, Brazil.

sedimPatrick, W.ents: effect of aerobic and anaerobic conditions. H. and R.A. Khalid 1974. Phosphate release and sorption by soils and Science.186, 53-55.

soil zinc. PavanasasivamCommunication of Soil Science and Plant Analysis, V. and Axley, J,H. 1980. Influence of flooding on the availability of .11, 163-174.

Chapter 9 – Bibliography

812

Pezeshki, S.toxicity: Inhibition of carbon assimilation in R., Pan, S.Z., De LauneS, R.D and Patrick, Wpartina a.lternifloraH. 1988. Sulfi.Photosynthetica.de-induced
437-442. 22(3),

Pizzaro,on bottom sedim M.J., Hammerley, J., Maine,ents of the Rio de la Plata. M.A. and SHydrobiologiaune, N. 199.228, 43-54. 2. Phosphate adsorption

ent of a sequential extraction Poulton, S.W. and Canfield, D.E. 2005. Developmparticulates.procedure for ironChemical Geology.214: implications for iron partition, 209-221. ing in continentally derived

PonnampAdvances in agronomerumy. Ama, F.N. 1972. The chemerican Society of Agronomistry of submy: Neerged soils. Inw York. pp. 29-88. : Brady, N.C. (Ed.)

PostmFe(III)/SO4-reduction ina, D. and Jakobsen, R. 1996. Redox zonaterface. Geochimica et Cosmochimica Actation: Equilibrium.60 constraints on the , 3169-3175.

aker, V. 1998. Ernährungsökologie, Habitatbeschreibung und Radem (Linneaus, 1763) im Ucides CordatusPopulationsstruktur der Mangrovenkrabbe Tropenökologie. Bremen Universität. BremCaeté-Mangrovenästuar, Nordostbrasilien. Msc. Thesis. Zentrum für Marineen, Germany.pp. 94.

sediments. Raiswell, R. and Berner, R.A. 1985. Pyrite formAmerican Journal of Science.285, 710-724. ation in euxinic and semi-euxinic

Raiswell, R., Buckley, F., Berner, R.A. and Anderson, T.F. 1988. Degree of Journal of Sedimentary Petrology.pyritization of iron as a paleoenvironm58, 812-819.ental indicator of bottom-water oxygenation.

Raiswell, R., Whaler, K., Dean, S., Coleman, M.L. and Briggs, D.E. 1993. A simple
formthree-dimation in fraensional mmodel of diffusiboids, fossils and detrital ironon-with-p minerals. recipitation applied to localised pyrite Marine Geology.113, 89-
100.

et A., Kazungu, J. and Dehairs, F. iF., Goeyens, L. Van R.Rao, R.G., Woitchik, Aleaves from an east African coastal lagoon (Kenya). 1994. Carbon, nitrogen contents and stable carbon isotope abundance in mAquatic Botany.47, 175-183. angrove

American Scientist.46Redfield, A.C. 1958. The biological control of, 205-222. chemical factors in the environment.

Reise, A. 2003. EstimBrazil. Ph.D. Thesis. Zentrumates of biomass and productivity in fringe m für Marine Tropenökologie. Contribution 16. angroves of North-
mUniversity of Breany. pp. 166. en, Germen. Brem

Lithos.boids. Rickard, D.T. 1970. The origin of fram, 269-293. 3

Chapter 9 – Bibliography

912

Rickard, D. and Luther, G.Wof iron (II) monosulfide in aqueous solution. 1997. Kinetics of pyrits between 25 and 125e formaotion by the HC: the m2echanismS oxidation .
, 135-147. 61.Geochimica et Cosmochimica Acta

97, 141-197. Rickard, D. and Morse, J.W. 2005. Acid volatile sulfide (AVS). Marine Chemistry.

Robertson, A.I. and Alongi, D.M (ERobertson, A.I., Alongi, D.M and Boto, ds.). Tropical Mangrove ecosystemK.G. 1992. Food chain and carbon fluxes. s - Coastal and In:
shington. pp. 293-326. aEstuarine Series 41. American Geophysical Union: W

Rossetti, D.estratigráfico dos sedim F., Truckenbrodt, W.entos Barreiras e Pós-Barreiras na Região Bragantina, and Goés, A.M. 1989. Estudo paleoambiental e
Nordeste do Pará. Boletim do Museu Paraense Emílio Goeldi.1, 25-74.

paleoenvironmental and redox proxy reevaluated. Roychoudhury, A.N., Kostky, J.E. and van Cappellen, P. 2003. Pyritization: aEstuarine Coastal and Shelf
, 1183-1193. 57Science.

thod for different eent of a sequential extraction mberg, K.C. 1993. DevelopmRuttemforms of phosphorus in marine sediments. Limnology and Oceanography.37, 1460-
1482.

:P ratios, and berg, K.C. and Goñi, M.A. 1997. Phosphorus distribution, C:NRuttemį13 Coc in artic, temperate, and tropical coastal sediments: tools or characterizing bulk
sedimentary organic matter. Marine Geology.139, 123-145.

Sahai, N. 2003. The effects of Mg2+ and H+on apatite nucleation at silica surfaces.
, 1017-1030. 67Geochimica et Cosmochimica Acta.

tropical terrestrial to . and Medeiros, C. 1995. Phosphorus transfer fromSalcedo, I.Henvironmaquatic systems - ment: transfer, cycles and mangroves. Inanagement.: Tiessen, H. (Ed.) Phosphorus in the global SCOPE 54. John Wiley and Sons
Ltd.: England. pp. 347-362.

Saleque, M.A. and Kirk, G.J.D. 1995. Root-induced solubilization of phosphate in the , 325-336. 129New Phytology.rhizosphere of lowland rice.

Sawáowicz, Z. 1987. Framof Stara Gora (Lower Silesia, Poland). boidal pyrite from the mMineralogy Polonicaetam.18, 57-67. orphic Radzimowice Schists

mSawáechanism. Geologie Rundschauowicz, Z. 1993. Pyrite framboids and their developm.82, 148-156. ent: a new conceptual

Sawáowicz, Z. 2005. Framboids: from their origin to application. Mineralog. Trans.
, 1-80. 88

Chapter 9 – Bibliography

130

argo, T.M.D. 1990. Schaeffer-Novelli, Y., Cintrón, M.G., Rothleder, A.R. and CarmVariability of mangrove ecosystems along the Brazilian coast. Estuaries.13(2), 204-
218.

Schellmann, W. 1983. A new definition of laterite. Natural Resources of Devonian.
, 7-21. 18

classSchellmificatioann, W.n of laterite” by R.P. Bourm 2003. Discussion of “a critiqan and C.D. Ollier (ue of the SchellmCatenaann definition an 47, 117-131). d
, 77-79.52Catena.

Schmitt, B.B. 2006. Characterization of organic nitrogen compounds in sediment and
in North Brazil. Ph.D. Thesis. Zentrum für Marine ngrove ecosystemaleaves of a mpp. 186. any.en, Germen Universität. BremTropenökologie. Brem

, L. and Scholander, S.I. 1955. Gas exchange in the roots Scholander, P.F., Van Damof mangroves. American Journal of Botany.42, 92-98.

Schulz, H.D., Fischer, WSchulz, H.D. 2000. Redox m.easurements in R., Bottcher, J. and Duijnisveld, Wmarine sedim.ents. In: Schuring, J., H.M. (eds.). Redox:
emFundantals, processes and applications. Springer: Berlin. pp. 81-93.

Schwendenmin a Brazilian mangrove ecosystemann, L. 1998. Tidal and seasonal variations of soil and water properties . Master Thesis, Ressources Engineering
any. pp. 101. e. University of Karlsruhe. Karlsruhe, GermProgramm

Science.85, 190-197. Shapiro, R.E. 1958. Effect of flooding on availability of phosphorus and nitrogen. Soil

Society America.Shepard, F.P. 1932. Sedim43, 1017-1040. ents of the continental shelves. Bulletin of Geology of

Shermneotropical mangrove forest: iron, an, R.E., Fahey, T.J. and Howarth, R.W.phosphorus and sulfur dynam (1998). Soil-plant interactions in a ics. Oecologia.114,
553-563.

de Itacurucá, Baía de Sepetiba, RSilva, C.A. 1992. Formas e taxas de ciclJ. PhD Theagem do fósforo no ecossistemsis. Ecologia e Recursos Naturais. a manguezal
Universidade Federal de São Carlos. São Paulo, Brazil. pp. 118.

Silva, C.A.R., Mozeto, A.A. and Ovalle, A.R.C. 1998. Distribution and fluxes of phosphorus in red mangroves, Sepetiba Bay, Brazil. Mangroves and Salt Marshes.2,
37-42.

forest in the Amazon estuarySilva, C.A.R. and Sampaio, L.S. 1998. Speciation of phosphorus in a tidal floodplain . Mangroves and Salt Marshes. 2, 51-57.

Chapter 9 – Bibliography

Smplants.ith, F.WPl.,ant and Soil. 248, 71-83. Mudge, S.R., Rae, A.L. and Glassop, D. 2003. Phosphate transport in

113

K.S. (Eds.)Snedaker, S. C. 1982. MangroveContribution to the ecology of halophytes. Kluwer Academ species zonation: why? In: Sen, D.N. and Rajpurohit, ic:
Netherlands. pp. 111-125.

limSombroek, W.G. 1984. Soils of the Amnology and landscape ecology of the miazon region. ghty tropical river and In: Sioli, H. (Ed.) Thits basin. Dr. Junk, e Amazon:
W. Publishers: The Hague. pp. 234-271.

.M., Martins, E. and El-Robrini, M. 1997. Influência das variaçõesSouza Filho, P.Wdo nível do mNordeste do Pará, Brasil. ar na sedimIn: Limaentação da planície costeira Bragan da Costa M. and Simtina duranões A.R. (Eds.) Contribuções te o holoceno -
azônia. FINEP/SBG-NO: Belém. pp. 307-337. à geologia da Am

r e a a.M. and El-Robrini, M. 1998. As variações do nível do mSouza Filho, P.WBoletim Muestratigrafia de sequências da Planície Costeiseu Paranaense Emílio Goeldi, Série Ciências dra Bragantina - Nordeste doa Terra. 10, 1-34. Pará, Brasil.

.M. 2000. Tectonic control on the coastal zone geomorphology of Souza Filho, P.Wthe Northeastern Pará State. Revista Brasileira de Geociências.30, 523-526.

Society for Mangrove EcosystemSpalding, M.D., Balsco, F. and Field, C.D. 1997. World ms: Okinawa. pp. 178.angrove atlas. International

Staupendahl, K. 2001. Das flächebezogene Wquantitativen Beschreibung der horizontalen Bauminkelmverteilung. aInss Wf-Ein Index zur : Akca, A., et al.,
(Ed.) WMethoden uand Verfahrenweisen. Zohab-Verl: Göttingldinventur, Waldwachstum und Fortplannung - Moderne Technologien, en. pp. 101-115.

Strickland, J.D.H. and Parsons, T.R. 1972. A practical handbook of seawater analysis. 2nd. Ed. Fisheries Research Board of Canada: Otawa. pp. 167.

chemical equilibStumm, W. and Morgan, J.J. 1981. Aquatic chemria in natural waters. John Wiley and Sons Inistry: An introduction emc.: New York. pp. 767. phasizing

ents: and Sulzberger, B. 1992. The cycling of iron in natural environm.Stumm, Wet Cosmochimica Acta.56, 3233-3257.Considerations based on laboratory studies of heterogeneous processes. Geochimica

of a meSuit, N.S. and Wromictic lake. Geology.ilkin, R.T. 1998. Pyrite form26, 1099-1102. ation in the water column and sediments

ngrove aThibodeau, F.R. and Nickerson, N.H. 1986. Differential oxidation of msubstrate by Avicennia germinans and Rhizophora mangle.American Journal of
, 512-516. 73Botany.

Chapter 9 – Bibliography

132

factors at patchy mThüllen, N. and Berger, U. 2000. A comaparative examngrove seedling stands on the peninsula of Bragança, Northern ination of environmental
, 1-12. 6Ecotropica.Brazil.

Tribovillard,paleoredox and paleoproductiv N., Algeo, T.J., Lyons, T. and Rity proxies: an update. iboulleau, A. 2006. Trace mChemical Geology.232e, 12-32.tals as

Twilley, R.Southwestern Florida estuary. R. 1985. The exchange of organic carbon inEstuarine, Coastal and Shelf Science. basin ma20, 543-557. ngrove forest in a

Verardo, D.J., Froelich, P.N. and McIntyre, A. 1990. Determination of organic carbon and nitrogen in marine sediments using the Carlo Erba NA 1500 (2100) Analyzer.
, 157-165. 37.Deep Sea Research

Vamos, R. aof hydrogen sulphide in the rice plant. nd Koves, E. 1972. Role of light on thJournal of Applied Ecologye prevention of the pois.9, 519-525. oning action

formsWang, H.D., Harris, in dairy-im W.pacted soils under simG., Reddy, K.R. and Flaig, E.G. 19ulated leaching. 95. StabilityEcological Engineering. of phosphorus 5,
209-227.

Wattayakorn, G., Win the Klong Ngao mangrove swamp, Thailand. olanski, E. and Kjerve, B. 1990. Mixing, trapping and outwellingEstuarine, Coastal and Shelf Science.
, 667-688. 31

Wijsma2001. Sulfur and iron spn, J.W.M., Middelburg, J.J. Hermeciation in surface sedan, P.iments along the noM.J., Böttcher, M.E. and Heip, C.H.R. rthwestern margin
of the Black Sea. Marine Chemistry.74, 261-278.

Williamphosphate by ms, J.D.H., Systers, J.K. and Walker, T.W. 1967. Frodification of Chang and Jackson´s procedure. actionation of soil inorgSoil Science Societyanic
, 736-739. 31American Proceedings.

y, S.L. 1996. The size distribution of ilkin, R.T., Barnes, H.L. and BrantleWframboidal pyrite in modern sediments: An indicator of redox conditions. Geochimica
(20), 3897-3912. 60et Cosmochimica Acta.

ation by reactions of iron ilkin, R.T. and Barnes, H.L. 1996. Pyrite formWCosmochimica Acta.60, 4167-4179. monosulfides with dissolved inorganic and organic sulfur species. Geochimica et

Wthe Black Sea indicated by pyrite. ilkin, R.T., Arthur, M.A. and Dean, W.Earth and Planetary Science LettersE. 1997. History of water-column anoxia in .148, 517-525.

: Robertson Inics. ., Mazda, Y. and Ridd, P. 1992. Mangrove HydrodynamWolanski, ESeries 41. American Geophysical Union: WA.I. and Alongi D.M (Eds.). Tropical Mangrove ecosystemashington. pp. 43-62. s - Coastal and Estuarine

Chapter 9 – Bibliography

313

arsenic system. Wolthers, M. 2003. GeochemGeologica Ultraiectinaistry and enviro.225nm, 185-190. ental mineralogy of the iron-sulfur-

Grand CaymWoodroffe, C.D. 1992. Geoman Island, West Indies. orphology and Bulletin Marine Sciencedevelopm.ent of m32a, 381-398. ngrove swamps,

WTW-Grundlagen und Tips zur Redox-Messung. Wissenschaftlich-Technische WeFirmerkstätten Gmnschrift: WeilheimbH, 1997. Redox-Fibel. . pp. 48.

Yadav, B.R., Paliwal, K.V. and Nimwaters on phosphate adsorption by calcite. Soil Scgade, N.M. 1984. Effect of ience.138, 153-157. magnesium-rich

rhizosphere oxidation in mYoussef, T. and Saenger, P. 1996. Anatomangrove seedlings. Australian Journal of Botanyical adaptive strategies to flooding and.44, 297-
313.

Ziebis, W.on oxygen fluxes in permeable seabeds. Huettel, M. and Forster, S. 1996. Impact of biogenic sedimMarine Ecology Progress Serie.ents topography 140, 227-
237.