MATURE: Biogeochemistry of the Maximum TURbidity zone in Estuaries. A summary report.

Peter. M.J. Herman and Carlo Heip

Netherlands Institute of Ecology, Centre for Estuarine and Coastal Ecology, Vierstraat 28, 4401 EA Yerseke, The Netherlands.

2 1 1 3 6 71. Introduction.

1 .a. Aims and objectives.

The EU-envlronment research programme MATURE (Biogeochemistry of the MAximum TURbidity zone in Estuaries) was a cooperation project between the following partners:- NIOO, Yerseke (coordinator)- University of Hamburg, Dept, of Marine Chemistry and Biogeochemistry- University of Hamburg, Center of Marine and Climate Research- NIOZ, Texel -TNO , Den Helder- University of Brussels- University of Gent- University of Bordeaux, station of Arcachon- Instituto Técnico Superior, Lisbon

The objectives of the programme are summarized as follows:

The role of biological processes in the formation and subsequent utilization of particles in the maximum turbidity zone will be studied in three European estuaries : Gironde, Schelde and Elbe. Special attention will be given to organic matter and biological processes acting on it. Nutrients and trace elements will be considered insofar as they are regulating biological processes in the estuary. Numerical modelling of water and suspended matter transport is an integrated part of the project.

Specific research questions addressed in the project are:• How is the concentration, stability and fate of aggregates influenced by the transforma­

tions of organic matter, advected from river and human sources?• How can formation, sedimentation and resuspension of particles be parameterized and

incorporated into numerical hydrodynamic models?• What is the importance of microbiological processes (primary production, bacterial min­

eralization, protozoan grazing) for the geochemistry of the system; what are the specific characteristics of the microbial loop on the aggregates; how do they affect the dynamics of the aggregates?

• How do the biological processes at higher trophic levels (selective grazing and manipula­tion of particles in the water column and upper layers of the sediment) affect the particle dynamics. Reversely, what is the influence of the geochemical environment on these processes?

1 .b . Organization of the project.

The field work for the project has been centred around six joint field campaigns of appr. 1 week. In Spring 1993 and in Spring 1994 a one-week campaign has been organized in Elbe, Schelde and Gironde. A great number of physical, chemical, microbiological and macrobi-

ological measurements have been performed on common stations during these campaigns. Each of the campaigns started with an aiong-estuary transect, followed by a 24h cycle with hourly measurements at a fixed station, and by a second transect along the estuary. Sam­pling on the common stations was usually concentrated on surface, middle water and bottom water levels, although not all variables have always been determined at all the levels. CTD casts provided a better vertical resolution for a number of physical and chemical variables. After analysis in the laboratory, all results have been gathered at NIOO and have been stored in a common database. This database (Herman & de Hoop, this report) with a manual and documentation, has been distributed among the partners o fth e project. In principle, it is freely available to the scientific community upon completion of the project.Alongside the field measurements, laboratory experiments have been performed on specific biological questions, e.g. relating to the feeding biology of the Zooplankton, and a mesocosm setup has been designed for experimental study of environmental stress on estuarine Zooplankton communities.Finally, 3D hydrodynamic models have been developed for the three estuaries. These mod­els have been calibrated on available field and monitoring data. Their application to the analysis o fthe biogeochemical field data is not completed, however.

This scientific summary report highlights the conclusions that can be drawn from the inter­disciplinary comparison of the results obtained. For a detailed description of the results of the different research groups, we refer to the reports of the partners given in appendix. A com­plete overview of these reports is given at the end of this summary.

2. Description of the MTZ during the field campaigns.

CTD profiles were obtained by the University of Hamburg (see Pfeiffer, this report) at many stations in and around the MTZ of the estuaries during the field campaigns. Two profiles along the estuary were sampled with a spatial resolution of around 10 km; during the 24h stations appr. every hour a CTD cast was made.Results of the CTD casts are delivered by the University of Hamburg in a number of ASCII, files. This partner provided a program to draw vertical profiles of the measured variables at every station (see Pfeiffer, this report). Results were combined in a Paradox data base con­taining all observations. From this database synoptic pictures were produced that summarize the data on salinity and light transmission (as a measure of turbidity) in this report. The CTD database is kept apart from the general biogeochemical database. It is included with the da­tabase as the files CTD.db (basic data) and CTDSTAT.DB (station data, linked to CTD.DB through the field “Station”).

Theflight transmissiörTvalues(gave clear pictures of the vertical and horizontal distribution of suspended solids along the estuarine salinity gradients. Figs. 1-11 summarize salinity (upper picture, a) and light transmission (lower picture, b) in the Elbe in 1993 and 1994, Gironde 1993 and 1994 and Schelde 1993 and 1994. With one exception (Schelde 1993, where gra­dients of light transmission were hardly visible) there are clear gradients in suspended solids along the estuarine axis. Vertical gradients in light transmission were clearly visible and most pronounced where the average transmission was lower than appr. 60 %.Fig. 12 shows the dependence of depth-averaged light transmission on salinity in the three estuaries and the two years. The relation of light transmission with salinity is closer than with position along the axis of the estuary (compare, e.g., the two transects in the Schelde in 1994: figs 10 and 11), indicating that the zone of maximum turbidity moves up and down the estuary with the tide.

In Gironde and Elbe, peak turbidity values are found at salinities between 0.1 and 4 psu. Turbidity in the Gironde is so high that transmissometry lacks enough resolution at the high­est turbidity values. The situation in the Schelde seems to differ from the other two estuaries. In 1993 no clear maximum turbidity zone could be detected; in 1994 a peak value was found at a slightly higher salinity (peak at 4 psu, distinctly lower turbidity in the range 0.1-3). Despite the difference in river discharge rates between 1993 and 1994, causing the salinity gradient to be moved further downstream in all three estuaries (figs 1-11), there were no large differ­ences in the relation of transmission with salinity between the two years in Elbe and Gironde. For the Schelde, again, the picture is less clear due to the lack of a distinct MTZ in 1993.From these results it can be concluded that the MTZ in. at least. Elbe and Gironde and pos­sibly also in the Schelde, is positioned in the very low salinity range of the estuary. It moves up- and downstream with the tide, and shifts along the estuarine axis, together w ith the salin- ity gradient, at differing discharge rates. In the field samples the concentration of suspended material in a maximum turbidity zone is closely linked to the transition zone from zero to very_ low salinities. Even if salinity per se is not the direct cause of the concentration of the sus­pended matter, it is closely correlated with the factors causing the peak turbidity.A full description of the biogeochemical variables in the MTZ during the field campaigns is given by Brockmann et al. (this report) and will not be repeated here. These authors discuss the differences between the estuaries with respect to nutrients, oxygen and organic loadings. They conclude that anthropogenic stress in the Gironde is far less pronounced than in thej other two estuaries. Nitrification in the maximum turbidity zone is well expressed in Elbe and Schelde. Denitrification is important in the Schelde.

relat

ive

heig

ht

relat

ive

heig

htElbe 1993 transect 1

680 700 720 740 760

km scale

32.025.6 12.8 6.4 3.21.6 0.8 0.4 0.2 0.1 0.0

740 760™

km scale

Fig. 1. (a) salinity and (b) light transmission profiles in the first transect, Elbe 1993. The horizontal scale is the longitudinal estuarine axis, the vertical scale is relative height above the bottom (1 =surface, 0=bottom).

trans

mis

sion

(%

) sa

linity

(p

su)

relative height

O - i M C 0 ^ U l O ) S 0 0 C Do o o o o o o o o

transmission (%)

10

0

640 660

680 700

720 740

760

km scale

relative height

° o o p p ^Ö Ko 4 ̂ CD 00 O

O O O O O —L C 0 ( D —‘■F' OCO

05 O) osalinity (psu)

Elbe 1993

transect 2

relative heighttû Tl

CD C/> •

w 3 'CD P) < w (D C/>7X »Lô ‘ 5 ‘a « --o »

CD

OCû 3 ;< ^ (D p)z r dçp w

cû' 3 £ w&!0 ^< T3

3- ~<d d

I *1 m

il aa , »

s «S gEiE CD o3 CDc i CD ' c»

05CDO

CD00O

■vlOo

-xjK>O

CDD"O

CD

7C3cn o £p_CD

-sj4̂O

CDO

3D

o —‘• r o G o - t ^ o i o ' ï ' v i c o c o —*■ 0 0 0 0 0 0 0 0 0 0 o

transmission (%)

640 660

680 700

720 740

760

km scale

relative height

o o o p o -*•Ö N> 4 ̂ CD 00 Ö

____________I..........................I............ 1_____________ I______

......

O O O O O —‘■ C O C D —‘■ l ' OCDô ^ - r o ^ b o o N o l ^ l ^ P 1 ! ^

00 05 O

salinity (psu)

Elbe 1994

transect 1

Fig. 4. (a) salinity and

(b) light transm

ission profiles

in the

second transect, Elbe

1994. The horizontal scale

is the

longitudinal estuarine

axis, the vertical scale

is relative

height above the

bottom

(1=surface, 0=bottom).

relative height

CDCDO

CDCOO

-siOO

*sjroo

-si

TC O3CD

8 SCD °

o - j - r o G o - t ^ c n c D - s j o o c o - * -o o o o o o o o o o o

transmission (%)

640 660

680 700

720 740

760km

scale

relative height

O O O O O —* - C 0 O 5 —l FOCOb ^ - k > 4 ^ b o C T > r o j ^ ! V ) P 1 .N> oo a o

salinity (psu)

Elbe 1994

transect 2

relative height

/ „ - ‘ l O W ^ U l O ' s l M C Dcno o o o o o o o o

transmission (%)

10

0

km scale

relative height

o o o o o CO CDIO 00 G) M 4^

salinity (psu)

—*■ N> COIO Ul IO(Do b) o

Gironde 1993

transect 1

relative height

I P> o o o o o

U10 0 0 0 0 0 0 0 0

transmission (%)

10

0

km scale

relative height

o o o o o

O O O O O - ‘- W ( T > - - ‘- K > C 0

C» O) osalinity (psu)

Gironde 1993

transect 2

relative height

s. ̂ o o o o o

O - ^ W W ^ Ü I O S O O C Do o o o o o o o o

transmission (%)

10

0

km scale

relative height

o o o o o 00 05

salinity (psu)

—*■ r\o ooM Ol M 00 05 Ö

Gironde 1994

transect 1

relative height

o - ^ r o c o - t ^ o i c D ^ i O o c o - *o o o o o o o o o ootransmission (%)

relative height

CDO

773cr>O Q9L o 0

_____________ —

-

I ..................r ...... .......... I ............ ' r ............ f

o o o o o CO 05l\5 03 G) M !

ro co ro cn r\o

salinity (psu)00 O) o

Schelde 1993

transect 1

Fig. 9. (a) salinity and

(b) light transm

ission profiles

in the

second transect, Schelde

1993. The horizontal scale

is the

longitudinal estuarine

axis, the vertical scale

is relative

height above the

bottom

(1=surface, 0=bottom).

relative height

o o o o

O - ^ M U ^ O i a i N J O O C D - io o o o o o o o o ootransmission (%)

0.8 -i

relative height

o o o o o

O O O O O - * - G û a > - > - N > G 0

oo a> osalinity (psu)

Schelde 1993

transect 2

relat

ive

heig

ht

relat

ive

heig

htSchelde 1994 transect 1

I______ L_

:■

f T I

90km scale

32.025.612.86.4

r 3.21.60.8

— - 0.4;---------- 0.2---------- 0.1

0.0

1 0 0

9080706050403020100

50 60 70 80 90km scale

Fig. 10. (a) salinity and (b) light transmission profiles in the first transect, Schelde 1994. The horizontal scale is the longitudinal estuarine axis, the vertical scale is relative height above the bottom (1=surface, 0=bottom).

trans

mis

sion

(%

) sa

linity

(p

su)

Fig. 11. (a)

salinity and

(b) light transm

ission profiles

in the

second transect, Elbe

1994. The horizontal scale

is the

longitudinal estuarine

axis, the vertical scale

is relative

height above the

bottom

(1=surface, 0=bottom).

relative height

o o o o o

o ^ r o u ^ o i O N i C ö c o - 1o o o o o o o o o ootransmission (%)

1.0 “i

relative height

o o o o o —■‘• w o ) —*-roco00 03 o

salinity (psu)

Schelde 1994

transect 2

=3 S ■n

3 CD t o

T3 Sto —f

CD N>

3 CD JO3 CDCD C/)

C &a 0) O

13. 3

O 8 O

d ’ 8mCOCO

SCD

c3

CO ? 3

3 Ooo

* <SP

<CD

E fCD

»c/5o

3t o

£CD CO3CO c

Q- COo ’3

*<3

o4^ S'S.

to §

CO CD*<CO

~0CD

r ¥3 "

s Ö) oCD 8

-A 3 cCOCO o 3

c rcoCD

3'7 * ' ÇD

<CD

CD3

o ' (O3CO~*

aCO

5 'S

ÖL3

çd ■ ?CD CDCO O3 3

t oooS

?CD

a S< Cçd CD

§ 3 .3CD

cÇ0 33' 3

(O CO

Sz r s .CD co

=j‘

Transmission (%)

no -t*. o> oo oo o o o o o

Transmission (%)

oo o o o

Gironde

1993 G

ironde 1994

Transmission (%) Transmission (%)

o

no

o

Transmission (%) Transmission (%)

M O ) CD Oo o o o o o

O r

mO 'CD

COCO4̂

NO CD 00 OO O O O O O

O r

Schelde 1993

Schelde 1994

3. The dynamics of suspended matter in the MTZ - flocculation proc­esses.

3.a. The in situ determination of floe size.

The dynamics of floes in the MTZ were studied primarily by in situ camera systems (see Eisma, this report, for lull details). Two different camera systems were deployed: a camera with a 1:1 image, and one with a 1:10 magnification. The volume of water photographed by the latter camera is appr. 100 times smaller than that covered by the 1:1 camera.Both camera systems have a different detection window for particle sizes. The mean particle diameter determined with both systems differs by approximately one order of magnitude, although both cameras seem to give consistent results. An example is shown in Fig. 13, where measurements of floe numbers in a size class vs. geometric mean diameter of the class by the two camera systems in one body of water were overlaid, after proper rescaling for the volume of water scanned. In the overlap zone between the two detection windows, averages between both readings were calculated. The consistency shown between the two camera systems was found in most samples where both cameras were deployed quasi- simultaneously.

1E31:1 camera

<T

1E01:10 camera

1E-1

1E0 1E1 1E2 1E3diameter (um)

Fig. 13. Comparison of floe size frequencies observed by the two camera systems in the same body of water. Number of particles per size class is plotted vs. geometric mean diameter of the size class. Observed numbers of particles in the 1:1 camera were divided by 100 since this camera scans a 100 times larger volume of water. In the overlap zone, the number of particles was calculated as the average of the rescaled numbers In the two camera readings.

Nevertheless, the (volume-based) mean particle sizes calculated from both cameras have a relatively low correlation coefficient of 0.32 (see Fig.14). This may be caused by scatter in the data, especially in the larger diameter classes which have a large weight in the calcula­tion of the volume-base mean size.

To filter out as much of the variability as possible, we calculated the geometric mean of the volume-based particle size of both cameras as a relative measure of particle size to relate with other measured variables.

In Fig. 15 this geometric mean particle size is plotted against the best predictor we could cal­culate from the environmental data available in the joint field campaigns. This predictor is expressed as:

1= 120

S 100

CDN

'c/1CCDCD

•o CD cn CD jQÖ>

80

60

40

20 V .

0 100 200 300 400 500 600

Vol-based Mean size 1:1 (um)700

Fig. 14. Relation between the volume-based mean floe size determined by two camera systems in the same body of water. All available data where both camera systems have been used are shown.

log-to (Particle size) = 1.714 + 0.105 * logio (SPM) + 0.145 * log« (POC) - 0.103 * log10 (SAL)

(eq.1)where SPM = suspended particulate matter concentration (mg.I'1)

POC = particulate organic carbon concentration (mg C .f1)SAL = salinity of the water (psu)

This regression had an r2 of 0.44 (n=97) and was highly significant. In view of the rather low correlation between particle sized determined by the two camera systems, its value may ac­tually indicate that most of the non-random variation in the particle size is effectively ex­plained by the three environmental variables.Among the single environmental factors, POC explains most of the variation in particle sizes. The (log-log) regression of particle size on POC has an r2 of 0.31. This value is 0.21 for SAL, and 0.15 for SPM.The result of this regression analysis using the average of the 1:1 and 1:10 camera systems is qualitatively in accordance with regressions of the particle sizes obtained by the 1:1 cam­era or the 1:10 camera systems on environmental variables separately. We consistently find’ that POC is the single most important variable in the environment explaining the size of the floes as observed in the field. =

Two main factors are thought to be important for the formation of a floe upon collision be­tween two particles (van Leussen, 1994): the organic coating of the particles and the salinity of the surrounding water influencing the double layer dynamics. The regression equation qualitatively indicates the importance of this coating with the positive exponents for POC and SPM. The influence of salinity is more complex. In the data set particle size decreases with increasing salinity, contrary to what is expected from theory. However, a closer inspection of the data reveals that the influence of salinity is non-linear. Fig. 16 shows the residuals of par­ticle size on POC and SPM (i.e. the difference between observations and the part of the pre­diction equation containing POC and SPM) as a function of salinity. It can clearly be seen that particle sizes are larger than expected at salinities up to 1 psu, but decrease again at higher salinities. Salinity change (from freshwater to slightly brackish) has a more profound effect than salinity per se, possibly because a new equilibrium between organic coating and salinity establishes at higher salinities than appr. 1 psu. It should be noted that, although most observations in this salinity range come from the Schelde estuary, the Elbe observations are fully in line with the Schelde data. Moreover, this zone of low salinities corresponds very well with the observed zones of increased turbulence, as observed in the hydrographic descrip­tions of the estuarine maximum turbidity zones.

m o h $ 3 2 ? 3 -

0 CD

■S g-0 oET w zzmÇ m œ 2.“ ° O — *< ^ 2, =>

°- Zt“33 M 3! M Q -" § - ® " § £ ® r o g<-+■ —. f(-\ t-2.— • O ^CT D) CD

§ ^ & ‘

s s ®3 " 03 03 CD • =3 _

3 ÿ a 5.

wCQ<'CD3

5 '

3o ~M 0 ®

>< 0 g

2 . 0 §° “■ r. w < i0 g

0

MW3Q .Q .

O - CQ »O C w —

(D (D ®

B B M 3g- w c -p

CD O

*• ® 3o

3 "0

O C

' o . COcr O

£

8 ¡ z

3 r - Is. ® z. 0 e -cn5 - ® “ 2 0 „• 3 ?T I — . O o —1 E.

CD Q . "O <£ r rc ^ 2 - , 0CD — . _N

3C S ' ® CL CD T '

O(Q

03 ’5 :3 ’

(Q0

• o(DCDQ .

Ili

N )enO)

coCD

oCÛ

C/)3000

9 -0’30r - t -0

0jQKd

0 3 0p

S- Z) h0 3’ 03 0 "i0IT Q. 0'

• o ’ 0' XJ—• =5 0

3 w2. 5 CD 2

- ®2 2 0 g-= t g -

8 E ' 5° 0

= Í 0 0 & O O

0CQ

8 ° 91 0

0 cr ^ 0 os 2! § ■

^ 0 -vg IL 0P M do gw 2■ 03 Q .JL 0 g 0

03

0 I 8 a-n ET.

| SCQ O

0 0

c m3 M& 3 o 0 3 . .

° 2 Z 0'

0 . O N 3 - 0 ^3 g 1— 00 "O02 o ̂3

f ®0-, 02 , 3 0 c- 3o crS * w o0 3

E ®

§■§

1| 0 «r+ _03 O

« -

> 0 Eî 02 5 '

, 2 *■ c o -CQ

O 0e 'S 0 2

003 0g- 0 0 0

CDÇT

Ç0d ’7sd '

CÛen

"OoCDQ .o

OOen

3 'en

§■'Q .CDr—+-

CD—t3d ’03

o '3

O 3- J ] C (D ( Q 13 — • Q- 13

= S ™ö ^ D

P Q. ST1 n ' Q.'v S ®

I (D 8S i a0 } CO = £

^ TJ -•3 o S(Q O Oco » 3 r- o. S S 0 5' 8 2 *

E | m ShS(D ? Q;¿ ( D 3

Ï & 8S g <9.

| i ' | I <8 Içd s* a! q . û> c* 3 Û3_

£2, w g o

Ä 8S ®

I Iï sCQ S

CO8 ! s gs roS o « 3 m 3

residuals from POC and SPMO CD CD O O O O

Ioen

oCQ^ ow2.3'r+><“ Ow u1C

ai

coC0 d

0°C5 co

COQ

<Sf>

J tfo coco

w awfàTÏUi

Ui Ci

•W71

co coUi

Uicoei

<WP c/ [0

IP

°8

m§ <pÇ0 _L0 en

1 33S P. °- 5Tio(D O O 33 o- çd œ

s- ?¡ ig X)s &a a8T&

S Q ^ 01 sr3 5CO myg CD CO Q.

g 3I S« ■p

H S 5 CD(D ÇO = : N*£ PCD •

« ^ ■o PI Ia »5 a”* O C CDcoCD ÇO.Q . N CO ÍS (D ET

I Ico ®=• tl)3zrCD

&

CD

OBSERVED MEAN GR. SIZE-*■ - i IV ) [OG ) C O DO ¡V ) 43.

O)

m

00

CD

N)CO

fiocs in the laboratory, the exponent is usually in the range of 0.4-1.0 (Gibbs 1985; Hawley, 1982; Kajihara, 1971). The present observations are intermediate between the two extremes.

1

0m u

Q .C/30 3C

'£ - 1

0 3

■232 .520 .5 1 1 .5

log (smallest diameter) um

Fig. 17. Relation between observed sinking speed of floes with smallest diameter of the floe. The regression line is a Geometric Mean regression.

The predicted difference in sinking speed of particles is two orders of magnitude over the size range observed. Combined with the predicted influence of POC on particle size, one can conclude, at least qualitatively, that the organic matter in estuaries is partly responsible for the dynamics of the suspended material. The importance of this factor can be estimated from the exponents of the regressions. Mean particle size variation attributable to POC and SPM was a factor two in our data base, leading to a predicted variation in sinking speeds of slightly over a factor two.

Due to technical difficulties (see: Eisma, this report) the number of determinations of particle setlling velocities was too small to re-appraise the relationship between settling velocity and particle concentration in the water. This relationship is important for the hydrodynamica! and sediment transport modelling (see: Cancino & Neves, this report). Literature data for the co­efficients of the relationship had to be used for the modelling. However, the form of the rela­tionship between settling velocity and suspended matter concentration as used in the model­ling (based on Dyer, 1986) is quite similar to the one predicted on the basis of the regression equations presented here.

3.C. Modelling suspended sediment concentration, deposition and erosion.

3-D hydrodynamica! and transport models have been developed for the three estuaries (Pfeiffer, this report, Cancino & Neves, this report). Numerical experiments, mainly with the Schelde model, point to several important conclusions.Local geographical information on sediment composition is important for a correct prediction (at short term) of the suspended sediment dynamics in the water column. The availability of erodable sediment and the critical erosion velocity parameters are dependent on the com­position of the sediment. It was shown for the Schelde that correct predictions of the along- estuary gradients in suspended matter concentrations were not possible without allowing for variable coefficients (based on observed sediment composition) along the estuarine gradient. Although obviously the composition of the sediments itself is a function of hydrodynamic factors, modelling at a much longer timescale would be needed to predict it, and thus make the sediment transport modelling self-contained. Models developed within the project make

predictions essentially at the time scale of a tidal cycle; this can only be succesfully per­formed with the inclusion of field data on sediment composition.The three-dimensional topography of the estuary determines essential features such as verti­cal mixing, salt stratification, upward bottom flows of suspended sediments etc..., leading to the observed turbidity in the MTZ. This is shown very clearly by the lateral transect in Han- sweert (Schelde) taken from the numerical simulations. Vertical gradients result from the specific structure of the northern and southern flow channels. (Fig. 18). In the Gironde, es­sentially the same model with similar parameters produced much more vertical stratification and a considerably higher turbidity in the MTZ.The transport models were able to produce hindcasts of the suspended matter concentration and salinity profiles well in accordance with observations during the field work. As such, they provide excellent tools to investigate further the relative importance of mechanisms of sus­pended sediment dynamics described by the field studies. Inclusion of these mechanisms into the models, however, was not possible due to time limitations of the project.

ÉI«P P Í|■I, Û

t P P * g i

Fig. 18. Model results showing vertical profile of salinity (psu) during low slack water period in a transversal section near Hansweert (simulated for the day 28/04/94). Y-axis givel local depth in meters, X-axis transversal coordinates. (From: Cancino & Neves, this report)

4. Th

e dy

nam

ics

of or

gani

c m

atte

r in

the

MTZ

.• C/D

0 5 jg

o £ 0.2

'S 5

8 -§ © eE ®

S %5.8© >, > _c 0 o. .cO f O 5Q ~o? s03 -O o

2 1• 8 0 Q.

• i?■A—I I—ra cok_C 00® 8 O g c E o o « 2© 2 E 2E f2 Ioí c o _ Q. co

C 0 CO X3E LÜm ®co -a

® .2-5 ® w "co O)

S I5 ®6 c O) —i—O co

_ © ° sc 3•j¡ toCO <» N r-

r OCDJDEZ3C03

CCDEE3CO03CDECDCOCD

SZ

Eo

03 =3

_ x S 'co

■Q 0 -m—— -*—* ̂'ZZ

o OT

»■H O Q-"2 E co co CO co CQ h-

. od 5 'S -g¿r a c c0 S co ®

too.co

co o ii= 03co -°£ £0)1co co3 -d Û CO

° - « CO ®

'E ° ° co co >,P> 2 = t co O t i C 3 CO

co "o§ co 2 $ r O)¿ o r )s i sS ° “0 . 9 - 1 o Q Ek_

2 <d

¿ -O

£ 2 o i à pS o >- ° - E 2'S c S ® .2 o— O - O j D C X I CO co co co —F CL~ -tr; Ol —P i- 0 Ç/5 »_ O0 o > © o z..2 co co * - c cc "o _ o — Fo co co 5 "o 3o) <_ o co — ° 2 'a) ÇO "O go £ o ® § _

m O S CL CON i O “ Ico —

co ®-a -oco <

E g © ÇO CL co

c 3 o o+3 O £ CO'5 o coCOCO 0co co o

co

coS*= . ° 73■o gL3 *v_cû a> ^■oCD CD

“D -Q

2 5a «73 .9 -

CO T0 CO CO

- - <0° 0 ^ co -C-C co—•2 i It j ü E S-E «S 05 § CO O ®g -o -° 5 c W * - ® ™ 03 7S£ .2 co

ï j j1 E ros p °-f o 'if |§>O w.l

c.2 r O ^ c)£ £ co .9 o _c® O c co r i :4= Q. CD (0 OC >. a i.© .2 .2o £ o o 2 «C "O C Q - -*—> ‘v_0 5 íO (Dü ° o ) ^ g t 5a) ü t í r i£ m 2 f o E

C _ I— O os ° .2 co; ® 5.2 CO fo .® O)^ o "O cô c o.-i. CO 3 ^

"O

03O

'c l>c03

OaECD

CCDEÎDCDcnCDccôE

CDcpto 0 0

0 "O H-2 c o -s s

CO~ c: co

.. £ N ® to E co t & g

0 ^ 73 oC CD

1.2 cc -g 2 -2 co

« ® E T ü 2<0 £ S -2 O c.2 >- -o o Q o*-3 0 C/D0 _C c/) “m ̂ 0E . o í d ™ 0 £ r £ O 0 l l ?

= -5 0 s -B -g-» Q. 0 0 ^

C »- 0 0

M i 130 E O

0

000 "O 13 -♦—»

2^ 0 0E ocoo >1- c:® F'S 5 0 ÖCD O

CO CO O O X3E ” 2 , c "°a P g ’S a ir ^ r V r ~

e ® o £co çaco x:

0 SO) 0 0 0 "o 'C0 E

O '0 .S2 E

0o00

EZ3

3ex0O)cEZ300•S000OO

coo0-O00

CDC

c000k-Q_0

00c r0

oII

OO*

es

IoOQ*

?%*

IIüOa,ro

III

-oc:coOOQ+OOQ.Iloo

ro

aic

cococcoçg

*k-50k_0XDC0Oszo!c

cr0

CJoft.

+

Ic<5O

g

® ÎH S" * - ÇO CMCO B ^-o 2 o 1— 0

« 0 S a0 .t=S ■oTD 00 9 £ -0

coci OTE 0 o 0

® ^ -

! ! ! 0 O i= d) O C

0 « C0 to

0 ̂0 0 CMa 2 £■■o ai uc

■« 05 o 0 00 •ï= £ Il® O)i- i E'0 ■§ pOD

5 . 2g -0

05 00 ^O 05 CO

^ S .2 5 .0 c

£ 5 c5 § 2> j= 0>§ .52 T3 £ 0 - 9 -2 E I2 w *“ c c 0 0 —'f— *■*-« 0 [10 0

.± 0 o a© c -C —- c c 0

_c 0 ■*- U 1- c © w 5

. o 0 —-°0 £00 0E 22 g .0 8CO co

i = S

il 0)0 0 o £ CLc .20 t50 »- § 0 2 nQ) -»-^ -o1 § fig0 o 0t . 03 0 ^CL c 0 0 0 Q

CO o c o- 8 ° .

o 0 g D.g' 0 ^ 8—5 O -*-1 43 .O 0 0 * ^ 0 0 0 ^ 0 ^ ®H 2 o

0 0 0 —- 0 T3r r c 2 „ 0■>- ■>- t - n . .

° ” > . 05

a ^-0 S- > & 0 0a o© aO co E 0 "co o

0 <?? 2 cô OTi5 0 ® ® O o i" £ O o - >*

n 0 i= -n ® ®O kD k_ r—

2 ® i ©0 0 > . §s § 5 i0 U * ' 5-i .5 -- i

cow00

w w = 0 o .o > io ° o —■= © 0 0 nT m+-> r"

0

. c >> 0 0 0 ç £ £ £ £

*0 .>. ® O O — W »■ r-

05= '8 g §»■§ 1 i TO

w OT 0 0 ^ C

0 a 8 " x “ d ® w ®, © ®

0 O £3co ® 3 r a o F d, ÇO co 0

n

00c

0

è ^ l l•+= 0 _

k- m 0 > •—0 2 J=> CL -*->® _Ç c

0 Q_

"Oo o05

° 'S C 0- ®

-0 -O W 3 - O c o ” w.2 g ^ 'So" p 0 gO OT-d 0 0 0

Ü0 7=1

— ® o 0 *3 > 0 0 C/3 £10 CO C .o co­co 0 0 ■£-■2 0_ g S i— E

5 ® - -oS > 0 oJ2 > n ® § 0 - 0 ° » 8 02 >,03 0 -C -5 -E

l » | 2^ o —

l a i i8 l t iCL -o E.0 - -2 c/3

S ï f ts l ü

O W S-*

X3 ^ 0 0

0 -O ra o c o c

0

. l |0 £î

O

0

c0O 0 O 0 ? - O ®O 0 O ®

. o 0 00 70

0 0 k_ k_

C0O0n0‘k-0>O)c‘c

0 û. 0)CO c c(D —

^ O) C

8 0 - •■§ . Í O 0 0 2 ° - - 8 §c -o S. 0 ® E — ~® 0

o E

-oc00■oos—*‘cC3)0E

0oo O o O E û 0 (- 2 0I I ™ §■! 20 -O Ö

0“O

05 °

2 0 O£ 0 0 o p 0 = 50 0 0 £

. CL

© 0 0 —0 0 — 0

8 1 S - o 0 ® s > c F£C 0

^ O Wg_£) 0 o "o >>

co 0S F

0 ® o ¿

■2 o c 0 00 aS 0 o

.2 ©0 0 —

0 h-p 0 o 0 o -p 0

X ® 0

Ü 3 > -O o ¿ E

0 c/38 0

1 w “~ o

c «2 § 0 — 0

Q 0 ®-g o - ° S c° -O - 2§ 1 >™S I 0 +■£ c o c0 Q E S ou: 0c £ -o c o o c o 0 0 0 0

o o

o00

0 ®CL O 0 ®© a

70 £ —o 5 ® rC k_ Q)

•§ S ° £ g3 8 § i §>. 0 — ° £T n i: m /n m

(D>Oen.2L—03>

Cc «Ï SC/D

Q)k_OE 2

9j W o o SO

CL

05p =L2 O .ti - ■ « C 0 0 O J c ô t œ O C “ m - i f c ;o 9 > . o E -5 3 ®" oc 2 0 -o o.

O o 0 S — o 0 .2 £ CD 2 o.®

0

0 0

0

^ 0 0 *1 c c — c 0 O N 0 o — c»-2 x £ 05 -

0 c C- 9

s | ï ï2l f l

0E0

ratio

DO

C/TO

C is

gene

rally

be

twee

n 0.1

an

d 0.

9. T

his

patte

rn

is on

ly ex

plai

nabl

e if

DOC

co-

varie

s wi

th the

ra

tio

PO

C/S

PM

. De

cisive

ev

iden

ce

on thi

s pr

oble

m,

how

ever

, sh

ould

co

me

S' 8 3 1 3

n 8? S æ! ^8-S I f°

S -n

l » J l l8lS!a8 <B w 3 3 Í j û 3 ®5 <9 o s ET9- Q. Ö> 3 CQ g 5> M ® gS ö | I i a. s f g S

CQ CD £ I »P ° ’ °S ~ 33 » =

l i |? |B n® ° SÎ S È Ï sa g. s P o g & | g > |

^ t l p- " « o=r cr =?> cd p 3 H

ci g ä Og g ?g

3 S ! îsj. S- » ®(D § 3 . ml&S »'-, ‘g S?® CD 3 9 - 3S e i ® < a s Q ® 5 ’ * 3 o®m!I l » -g. ® g 83 > n S S < G) œIls I i I R !D O (D O r« - i. MÄ S 5*CD ® 3

oor\ooo

Co co 13 o^ O

(Q313 o o

Ulo oO)o o"4o o

- 52. ö Er 0® CD 3 id o i o s >° Q < ° 3 g > ¥ 3 8 § < S § 5 2 3 §

N)

O O V ® f t C-. ^ 5 rô 5 M CD ^

P O C /T O C | S | 3 | S 3 | 8 Ü a § S Io o o o g-îgûffl 'J

CD 00 -*■ =®5a^ os ® “ g-= I ! i s. ? g - ^ t i s s g . j »? g â | & c, ¡ Í S 1 * ; 8 | i I I 5 ®S 2 I s s I s ! s I s | l » | g 8 ® 1 § a I ! g. I s® 3 â m m O Ç? 2 „■ _ ° =t 3.gsgro O 8| t r f f " 3 as

rofS^cp- + œ g ® 3 ® 3 ®

§. g a s. s p s i m a S®* gsa&gS S|I ® s _r -£ sâ^ggE-. S§I I g-5 o E g ' & ! | 8 ^ g

3 S s ® 5 < ?û . r \ O ju O i-* - — CL m -r~j

S § 3 “ c ^ § ex & I( d L c S o o Z . =;• o S -Q- lo 3 § £• "* °7 ® ® ®D 1 0 | | 8 ? « â S § - < °Z> “ CD 0 j < w Û5Œ "—— S S Z> O 5.o | ô ' S î 5 i = ° § °®" 3 3 O Q* •-»■ (Û o 3W 3 i ; 3 O < “ : J [ ) ¡ i 3® O < T 3 < 3 (D = j g CL

< ? CD8 - Dq T®5 ' « 0»g g ç o | o 2 , o S t3 ( ü cr t (l) 3 ® O í ^ » 3 ^ -o

® S- â 9 : œ : | s I

®®o

ü m • * c_c. —• — - i CD r - m3 — 5 3 | - I m 0 3 W < Q .-i —• or. c/) D w ÿ » ■ ® LJ-K-3 ® 5' O ^ S ® n < 3c 3 <Q(E 0 < 8 « £ - 3

R E L A T IV E D E G R A D A B IL IT Y ? g | 3 3 - 8. § 1 g0 < “ o 5 § rop ^ o r .3 S t 3 s ® â - 3

I l l a ° -g “ s § I5T rT — Ui i l - # -2.Q 3 . = t ~ “ C3

®C/J

i g l i c a 3 TlûC) (g I g I 0 cg

| “ | ! ¡51*56- S“ 0 3 0 T 7 5 . -1 ® T M

ero.<CQ

O01

QLU>OCLUCOCûo

100

10

1

0.1

0.011010.10.01

PREDICTED T-corr BAPR

Fig. 20. Comparison between observed (temperature-corrected) bacterial production rate (pg C I"1 h'1) and predictions based on a linear function of POC and DOC concentrations. Data are labeled by the first letter of the estuary sampled. For comparison the 1:1 line is also given.

With an r2 of 0.77 most of the variation in bacterial production rates is described as a simple function of substrates available. It was not possible to relate any of the remaining variance to a measured environmental variable, such as SPM or salinity. Strikingly, the largest deviation between observation and model occurred in a few samples from the Hoboken-Rupelmonde region in the Schelde, both in 1993 and in 1994. This is the region of maximal nitrification activity, and maybe the deviation bears some relationship to that. However, it was not possi­ble with the data obtained to go deeper into this problem. (Fig. 21)

cc -40.01 0.1 1 10

Salinity1 0 0

Fig. 21. Residuals of bacterial production rate (pg C f 1 h '1) from the regression on POC and DOC, plotted versus salinity. Data are labeled by the first letter of the estuary sampled. Large positive residuals are found in the low- salinity zone of the Schelde estuary only. Based on data by Goosen et al. (this report).

The plot of residuals vs. estimates indicates a slight curvilinearity in the relationship between bacterial production and substrate. In fact, a slightly better regression could be obtained with a power function regression: BAPR = A * DOC5 * POCc, but the increase in fit was not very

high, and this relationship has the disadvantage of being difficult to interpret in biological terms.

The bacterial production rates are not very high, considering the important amount of organic matter present in most samples. By assuming that organic matter decay rate = 2* bacterial production rate, and recalculating units, we conclude that the first order decay rates of DOC are in the order of 6 yr"1, and for POC in the order of 2.4 yr'1. These rates are low, and com­parable to first-order degradation rates of organic matter in sediments. This is further evi­dence that processes in the MTZ are in several respects more comparable to sediment proc­esses than to typical water column processes. The organic matter present is resistant to bacterial degradation; apart from adsorption onto particulate matter, this is probably also caused by the lack of local primary production as a consequence of light limitation (see Kromkamp et al., this report).

The difference in decay rates of POC and DOC is important for the evaluation of the conse­quences of accumulation of inorganic material in suspension in the MTZ. W hile the inorganic material is a factor retaining the organic matter in the MTZ, it is also responsible (through adsorption) for a slow-down of the degradation of this organic matter. The relative degrada­tion rate (taking the rate of POC as 1) is added to the figure showing the influence of SPM on organic matter partitioning between POC and DOC. The dependence of degradability on SPM is most pronounced in the range of SPM between 0 and 100 mg/l.

The results of this analysis do not lend support to the hypothesis that the high amounts of suspended matter provide the bacteria with ‘reaction centres’, through their attachment on the particulate matter. Enhancement of the degradability at high particulate concentrations is not found, on the contrary. However, this does not preclude that the availability of substrate may be of importance to nitrifying bacteria (that have a clear preference for attachment to particles) or for denitrifying bacteria (that are dependent on anoxic microsites). For bacteria in general, Hernandez et al. (this report) show that the percentage attached to particulate material increases with increasing suspended matter concentrations, with decreasing salinity and with decreasing oxygen concentrations. The cross-correlation between these factors clearly shows that bacteria in the MTZ are much more attached to particulates than else­where in the estuary.

Measurements performed to investigate the effects of attachment on processes did not yield conclusive evidence. Bonin et al. (this report) report a weak correlation between measured denitrification activity and suspended matter concentrations in the three estuaries. This corre­lation holds true in most cases for the location of the peak activity along the estuarine tran­sect only. These authors, however, also show that the highest denitrification activity in depth profiles of Elbe and Gironde is found in the surface waters. Considering the generally ob­served depth profiles of suspended matter, with distinctly higher concentrations near the bottom, this observation would rather point to a negative correlation between denitrification activity and presence of floes in the water column. The measurements of denitrifying activity by N20 production are in general very high. In fact, most are higher than the potential denitri­fying enzyme activity measured in the same samples, and the order of magnitude of meas­ured denitrification rates would, in combination with estuarine residence times in the order of tens of days, deplete almost all available N 0 3 from the water. Possible interference of nitrifi­cation with the N20 production (used to measure denitrification rate) would need more de­tailed research.

In general, variation in bacterial density or biovolume was limited (Hernandez et al., this re­port). Biovolume per cel varied over a range of appr. 2, while densities generally ranged be­tween 2.106 ml"1 and 107 m l'1. This variation is considerably less than the variability in bacte­rial production rates. Bacterial numbers or biomass are a bad predictor of bacterial activity rates.

5. Quality of organic matter - chemical analysis.

Burdloff et al. (this report) analyzed the quality of organic matter in 6 different estuaries by chemical analysis. (Fig. 22).They measured bulk organic material, and also the more de-

By and

larg

e,

ligh

t lim

itatio

n

is the

mo

st im

po

rtan

t fa

cto

r d

ete

rmin

ing

p

rima

ry p

rod

uc

tion

ra

te

in the

three e

stua

ries

(see K

rom

ka

mp

et

al.,

this re

po

rt). C

alcu

late

d

de

pth

-inte

gra

ted

p

rima

ry

30X3 'c3

m 3 < 3 © 0 CD —1 ffi O ffi

a o Q tc 5 3r \ r? O 3 ^ CL05 o ®

c o f o -a cd Q. = r

S - - « *© 3 O °

N O Z3 0

C “Oo S3CD CD

£D ®

3So3

° or r T 2 (D ? CL 0 O CD CO g

cu 03 5Q_ _ r-^ 5T CO5 " co 0 "o ' ° 0 ( 3 ,

T3 05 5— Q rt-3 o ' J— 3 _— 05 " *

2 3o

Ä - O - a i O - * Z o o c r C/5 O CDCD 05CT M CD O O «3 3 9 s a l

CD CD■a 8 a .O 05 5 Í5 ffi Äz n

< ( Q O

8 © • § 5 059 . ( 5 I S - S oCD CD =>

© o— 05 O 3

CQ CO

S 3Ü!. 05

CQ 3 o í 9>

o © CO ^

D) H a - NcoCD 3 3 ^

IT. ^3 CD5 m

c r

CO 0)

1 1CD —CO

0g ; c/5 ®

o 3? CD ÇL 3a .0 3

g - 3 © ®

O o

X 0<3 CD 0 •

- ICO

0 0

1 5 —* o c r 0 c r

00 r ♦ 0 Q.r -t-O3o■"ICD

© 5 o o

JSo -a w o S ©0 £ . 3 CDc r 5 '

1 -

P13 z r-K 00 o 0 = 3-CQ€ o05 ©Î I“ » Q 05 ©9 N Cl o^ 3 CD CD COc ©» 95 ‘ 2. CD o

-o05 CDco a . co2 "co3 CD g ; cd05 i

3 ¡ '© 58- 8-.S °

■o2 ® 2. 3-0 SP2 co C l O~ 3O CD

T lm

CO L< CD 3 -3 - 8© Ö T 05 3

© £CO 05 i3 3

9 ©O CD

3 f ' £ ®

i 3 '5 I

§ l i s -8 2 8 Sr , 3 J CO

3 ' 2 c t 3CD CD CD Q - Q ©C CD CD 05 CO CO CD C

, Q CL

© ® If « ® ® £ ®

2 2 ^ - 8

i I g iO g fsj ^û) (D n

CD nO 3 0 5 3 - w_ © a

CD

0

2 .'S o' o? m çd

Q.0o

o 2.3 -2- 0 ■ CL H

0 2

© os 0 ©

3 0 9 0© 3 ! c ^

■? 5 '

2 . CD Q- M 0 0

05 9Q 0

T l _ © 2.'S a o 0 © . _ w

2. s© o 0 3

P

■uZ T

SO

■O0)35o30)3CL

T J“ r

3 'a>

« 5

oaco

5 '

3

0 <2 -TlÍ cn <P a D , ,

a s Pa s °a | | t * e . '?

S •S-8 S’o »§ « a -o55 §5' a- ? s.0 T !CD B C-a 2 § aQ. 0W o o (3 2. &

ë 3s &3 0O =3--K Q) 0l ¡

&0 8)■o y i X 0 O Q.0 er

S - ° •< 3O- Oä ^0 5c/> o § pQ- < — 0■g. ¡?CL Cw wTl OO g. 3

m

^ m C /) N> T l

3CQ

mc o

P G L y P O G

rd p X o? s : m

ÿ s g S P S ' gCQ ^ ¡a 3 - 3 %C § S“ a = O.

o o CD

DcQ■M lr +

» <■

o >

mcto

T 5<D0)3(D0)

C0 3-%(D0)

3 o —, 0

0 Q 0

><2. -> « © , o Sc r 2-05 O

5 CD 32 8 s5 ® 2 j o ' o q ' cd' 3-z r z j O

coc13COcCD

S - 3 "(D r f3 . z r 0 0 — 3

c 5 3 92 . T O <

i 1 8 «sQ; 0

~oo 38 - 5o ©

— H ©5 = • ©- O 0

S.CQo 2

0

3-. ©

0 T l 0 O c o 9

-■ S- co ° 8 oQ ■ 0© o ilr t o CT 0 CD 0

0 0 Q 0 Q . ”

o 8 o z O 3 ' 0

m ' O. ; M ‘ o O ±LJ. V) V) O ^ O ÇO. ZJ8 3 - 2 2 Q . © 5? 00 © .v^r q_

8 « i .

0g

S00

3007T0Q.

0Q „ „Q- O 5 Q-

m â ' S î 3 2 - ®0 (Q 2 ©0 CD l<

- i . © E 0 3 0

CQ0=5o'30 / ♦CO

co —\ co

CQCD0Ç0o'15

O ~ ^

^ ® ° o0-8 3 z ^ ° a =?

0 c r q . 9"• = 0 CQ

0 § Q. 0 ' 0 O '

i 3T l 0S 31 0 0 -> 0 _ Q °0 e0 P-Q. _5 ‘ 2

0

0 ® © 0 0 =r.Q 3i “ 0 .0 «. o 2— co

Q ©o 0 oCD r ; Q. 0 05 0— 0

"O 0o TriClc c ro *■<

£ < ©

§-8 -< 0

0 0 Q05

=r O0 3- f c ©0 ag o o ’ 0 © 5 'T5-W

P f :Q.

3 0

5 80 Q.

O 0 “ Tc r oQ*

•<Q.

0 O'

O ©- © 2 . CO

dv O co © . ©

© 8 0

Q —

0c roc

toeno<0“ i00g0

0Q

CQ0

3 .

0 2r t 0 0 0—iCL T

8 £ —t i 0 ^ 0 , 0 H CD z r 0 CD

g 3 s I

0 c 0 "O 0 3 Q . =T CO CO CL -V

O3

production for a standard day (i.e. calculated with the observed P-l relationship, assuming the same illumination at the water surface and the same daylength, and corrected for tempera­ture differences with a Q10 of 2.5) is strongly dependent on light extinction in the water col­umn, as expressed by the dependence of production per unit biomass on the euphotic depth Zeu (Fig. 23).The relationship between depth-integrated primary production and euphotic depth is not the same for the three estuaries. There is a significant difference in slope between Gironde on the one hand, and Schelde and Elbe on the other hand. This difference in slope is surprising at first sight: the phytoplankton of the Gironde is the worst adapted to low light regimes, while this estuary has the lowest light intensities in the water column. Most probably the lack of a genuine estuarine phytoplankton assemblage in the Gironde may explain this pattern. Pri­mary production in the maximum turbidity zone of the Gironde is very low; only at the most seaward stations there is a substantial production. The low production in the maximum tur­bidity zone is performed by marine phytoplankton species that are badly adapted to the turbid estuarine environment, but not replaced by estuarine species because net productivity is negative even for these species. It should be kept in mind that these measurements are taken in 1994, when due to high river runoff the maximum turbidity zone was near to the es­tuarine mouth.

1E4

_ 1E3

0 1 E 2H

J 3 1 E 1

1 1E0CL

1 E -1

1E-20.1 1 10

Zeu (m)

Fig. 23. Depth-integrated, daily-integrated primary production per unit of chlorophyll-a for a standard day along transects in the three estuaries (mgC m'2cr.(m g Chia ffT3)'1), plotted versus euphotic depth. Each normalized production estimation is based on measured light extinction coefficient at the sampling point, and on measured P-l curves at that point. Slopes for Schelde and Elbe are not significantly different, slope for the Gironde Is significantly higher. Data from Kromkamp et al. (this report)

In general, production per unit of phytoplankton biomass is very low in these samples. As­suming as an approximation an average depth of 10 m at these stations, a carbon-to- chlorophyll ratio of 30 for the phytoplankton and the standard light and daylength values used in these calculations, gross primary production is on average only 7.5 % of phytoplankton standing stock per day. Taking into account the respiration of the phytoplankton, net primary production was most probably negative in most stations studied.The limited role of phytoplankton primary production for the organic matter dynamics in the MTZ is further emphasized by calculating the ratio of gross primary production on total or­ganic carbon. The average ratio for the samples taken is 0.39 % per day. Allowing for phyto­plankton respiration, this ratio approaches zero or negative values in most of the stations.

s -cSchelde

Gironde

Elbe

7. Higher trophic levels: occurrence

An interesting observation is made by Gasparini et al. (this report), on the shift of develop­mental stages of Eurytemora in Gironde and Elbe (Fig. 24), but not in the Schelde. Two al­ternative hypotheses are brought forward to explain this spatial distribution: (1) nauplii drift downstream and copepodites move upstream by active swimming, where swimming speed is dependent on body size or (2) specific gravity of animals increases as they grow older, and therefore nauplii would preferentially occur in surface waters with a net downstream current, whereas copepodites would occur in the bottom waters where the residual current is up­stream.

Although the available data are not sufficient to reach a conclusion on the mechanism, sev­eral factors are in favour of the first hypothesis: measured swimming speeds of Eurytemora are sufficient for the displacement made; the spatial distribution of Eurytemora with respect to suspended matter in the estuary is different for the three estuaries (coinciding with MTZ in the Gironde, upstream of MTZ in the Elbe, downstream of MTZ in the Schelde); in the Schelde the spatial distribution of the species suggests that it is actively avoiding the low oxygen zone, even if this has the highest potential food concentration; no clear gradients in the vertical distribution of the species have been found in long time series in the Gironde.The investigation of this problem with the aid of the numerical transport models developed within this project is an interesting topic for future research, since the retention of autochtonous estuarine populations within or around the MTZ is fundamental to the under­standing of the ecological processes in this area.

The occurrence of hyperbenthos along the transects in the three estuaries (Fockedey & Mees, this report) bears little correlation with the environmental factors measured, except for the strong avoidance of low oxygen levels in the Schelde. In the Gironde the hyperbenthos has its peak occurrence in the most upstream station, well ahead of the MTZ. In the Elbe it is slightly upstream of the MTZ, whereas in the Schelde it is downstream of the MTZ. Densities in the Schelde are lower than in the other estuaries, but considering the temporal variability known from this estuary it is unclear whether this difference is significant.

Higher trophic levels: food biology

Several studies have been directed towards the influence of abiotic and biotic factors on the food uptake by Zooplankton and hyperbenthos.

Fockedey & Mees (this report) describe the feeding behaviour of Neomysis integer as selec­tive towards Zooplankton. Besides Zooplankton, the animals have (sometimes large) amounts of non-specified detritus in the guts. Qualitative chemical analysis of the material is currently going on. In future descriptions of the dynamics of hyperbenthic animals In the estuaries, a closer connection with the Zooplankton stocks is certainly a point of interest. The problem is important since previous studies have shown that hyperbenthic animals may provide an im­portant link between primary production and mineralization processes in the estuary and fish productivity.

Tackx et al. (this report) performed experiments on selective feeding by Eurytemora affinis on phytoplankton. They showed that the copepod has a significant preference for phytoplank­ton, even in the presence of relatively high concentrations of (partly organic) suspended matter. The species can therefore not be described as a non-selective filter feeder, taking phytoplankton and detrital food in the proportions offered to them.Gut fluorescence measurements in many samples differing in chlorophyll content and sus­pended matter (Gasparini et al., this report) confirm these observations, but define more clearly the conditions under which selectivity may operate. As shown in Fig. 25, no correla­tion was found between gut fluorescence and chlorophyll concentration in the water. A clear negative correlation was observed between gut fluorescence and suspended matter concen­tration. Egg production rates are lower at higher suspended matter concentrations, but the decrease is less pronounced than that of gut fluorescence, showing that the animals com­

pensate for the lack of phytoplankton food by a higher intake of non-fluorescent food parti­cles.

A few simple model calculations, using observed data on POC, SPM and CHLa in the MA­TURE database (and, thus, taking into account the field correlations between these vari­ables), explore the possible impact of selectivity of the copepods. Assuming that the species feeds non-selectively, uptake of chlorophyll-a would be proportional to the fraction chl-a in the total POC. If total uptake rate of POC is described with a Monod function, gut fluores­cence would be given by:

GF = a * POC/(K+POC) * chla/POC

where GF is gut fluorescence recalculated as a dimensionless (gC/gC) ratio, a is a propor­tionality constant, POC is total particulate organic carbon (mg.I' ) and chia is chlorophyll a concentration recalculated to mg C.r1 with a C/chl-a ratio of 50.

Predicted gut fluorescence under these assumptions is given as a function of SPM and of chlorophyll-a in Fig. 26a,b (where a=0.04, K=4 mg.I“1). A clear relationship with Chl-a is shown, contrary to the observations.As an alternative, one could assume that the copepods’ clearance rate (volume swept clear per individual per unit time) is inversely proportional to chlorophyll-a concentration (animals look for phytoplankton food in larger volumes when this is rare) and inversely proportional to POC concentrations (high concentrations demanding more handling time for the particles). Further, one could assume that the animals take all phytoplankton food from the cleared vol­ume, but only a small proportion of the POC. Gut fluorescence would then be given by:

1 1G F ~ a * chia * -

K c + chia K p + nPOC

where Kc (mg C I"1) and Kp (mg C I"1) are saturation constants preventing the clearance rates from becoming infinite, chia is the phytoplankton concentration recalculated to mg C I'1, nPOC is the non-phytoplankton POC (mg C I'1), GF is a non-dimensionalised ratio and a (I (mg C)'1) is a proportionality constant.

Figs. 26c,d (where a=0.0275 I (mg C)‘1, Kc= 0.005 mg/l and Kp= 1 mg I'1) show that this model cannot be refuted by the data. Both the dependence of gut fluorescence on SPM, and the non-dependence on chlorophyll-a concentration, are reproduced. The constants of the model are chosen such that the order of magnitude of predicted gut fluorescence corre­sponds to the data in Fig. 25. Moreover, taking into account the gut passage tim e used by Irigoien et al.(), the clearance rate for phytoplankton was calculated without any other change to the model than recalculation of units. Its value is in the range measured by Tackx et al. (this report). The dependencies of phytoplankton clearance rate on phytoplankton biovolume (shown in Fig. 26e), total particulates volume and ratio phytoplankton/total particulates (not shown here) faithfully reproduce the figures given by Tackx et al. (this report). These features corroborate the simple model proposed here.

This model can be interpreted such that the animals take all phytoplankton food particles from the volume of water cleared per unit of time, and complete their feeding by uptake of a relatively small proportion of the non-phytoplankton POC in this volume. Assuming that the maximal gut content (calculated as a dimensionless ratio) is in the order of 0.02, one can calculate the non-phytoplankton gut content as 0.02-GF. Thus an apparent selectivity can be calculated as the ratio between relative concentration of phytoplankton POC in the gut over relative concentration of phytoplankton carbon in the water, (relative concentration being taken as concentration of phytoplankton divided by total POC, either in the gut or in the wa­ter). The apparent selectivity is almost independent of POC, but is a strong negative function of chlorophyll-a concentration. Under this model, we expect the highest apparent selectivity indices when both chlorophyll-a and POC are low, but an almost equally high selectivity at low chlorophyll-a and high POC.

Stag

e St

age

Sta

ge

CVIf

CVIm

CVf

CVm

CIVf

CIVm

CHI

cuCI

7 1 0

ELBEO

ox—*«o

ex ••— > o

C K *

cx-o

o Data• Model

oT T T T

7 0 0 6 9 0 6 8 0 6 7 0 6 6 0

Mean position (km from the upper river)

6 5 0

CVIf

CVIm

CVf

CVm

CIVf ■

CIVm

CHI

CU

Cl

3 8

ox

©

SCHELDEO X

oxox

ex—• o

©

— >o

t T T

o Data

• Model

T3 6 3 4 3 2 3 0 2 8

Mean position (km from Antweerp)

“ I2 6

CVIf

CVIm

CVf

CVm

CIVf ■

CIVm

CHI

CU

Cl

GIRONDE

*>oox*

ox •o*

ox*ox*©

o Data• Model

o*

75 7 0 6 5 6 0 5 5

Mean position (km from Bordeaux)

~i5 0

Fig. 24. Distribution of developmental stages of Eurytemora affinis along the axis of the estuaries Elbe, Schelde and Gironde. ‘Model’ values are calculated taking into account size-dependent swimming speeds of the animals. From: Gasparini et al., this report.

Rg. 25. R

elationship between

gut fluorescence of

Eurytem

ora affinis

and a) total suspended

particulate m

atter (SPM

) and b)

chlorophyll-a concentration

in the

water. From

: Gasparini et al., this

report.

Carbon specific gut pigment content (ngC/ngC)

00

0O

O0

OO O

OO

0 0 O O O —iro -F* CD 00 —* ro

+o

■ ■- O □ □ □

,0 0c9 ® ■■ G l o

□

ö l

n3;oo■O3"

ÍU/—. 3(3S3Ul

□ □ □

□□ □

en □

□

roo

0.014

Carbon specific gut pigment content (ngC/ngC)

0 0 O O‘0 0 O O0 0 O Or o -F«. (D 00

o'o

ob— iro

ob

+o '

oo

■■

□□□ ? > Q □

. 'V ' °

o

:/:/

:/

.7

I 1

//./ /

> ' /

/ // //

'/u

P

! to »(Q CQto to</) (/>W (fto o'D Df\) -*

Ozo

□ □ □■o

□ O

m CD 00 m2° Q È

m

5m

5 ' ö a ? 3!rt. s H g. (53* o •0 P 3 "9. roH 3 P^ 2. J 3 __H . =F D 0

S-= 8 Im J ' D 3 IOg o I » o 3 « ® » 3

u s n-8 I I “ 85 » 3. o g5 ? § & g §• 8 ® ÖT gq p o g 2

: C oÍ £ °3 S ® I” 2 . - 5 « 3 ? S i n ®

§■ I-1 . 8 IP CQ 3 0 <.

C o* -» ÇD i—3 g a a <§ i s# ?ro < §

1 1 1 Ï | S p i lIp g I I I “I i -o

o S_D “ 'S. o. ro ^ ro o ro 3 o o 'S.0 p Q. 0tu ro -6 ro roä f f s I-O o o 0 p P "O "S. C 03 ro*?“ 3.Ä ro = S- ro ro m ri, S io ro ® w o Û0 O 3 J r r3 ° d S s 5 ' o O 3 3 ? ® S l ro g- ro- S ro ro ‘ 0

00

=3'CQ

TIcq’roo>CD

Phyto-clearance rate (ml ind-1 h-1)

T)rr*§ oH 0

=3SO3

-o3

o

oo

Fig. 26c Fig. 26

d

0 1

2 3

4 5

0 100

200 300

400 500

Phytoplankton POC

(mg C

1-1) SPM

(mg 1-1)

Gut pigment content ratio

% D B 0

cpIOO)CD

Gut pigment content ratio

o o o o o oö ö b ö P ö ö

roo

? §ip—V. 00c o

CQ

O

I « I

Gut pigment content ratio

o oí -*• oí ro

. Gut pigment content ratioroo>

ob

ob

oo o

oo

poo03 o

COo ro 4 ^

o

§

COOo

oo

CDoo

0.016 -I

—

1 0.016

List of sub-reports of the partners.

Herman, P.M.J. and B.J. de Hoop. The MATURE database.

Eisma, D. In-situ measurements of suspended matter floe size and settling velocity in the estuaries of the Elbe, Schelde and Gironde.

Brockmann, U.H., F. Edelkraut, T. Raabe, K.H. Viehweger. MATURE. Project Organic Matter and Nutrients. Final report.

Cancino, L. & R. Neves. Mature. Final report (hydrodynamica! modelling).

Pfeiffer, K. Hydrographic measurements and numerical modelling contributions compiles and issued by ZMK.

Muylaert, K. & K. Sabbe. Structure and spatial distribution of spring phytoplankton assem­blages in and around the maximum turbidity zone of estuaries: a comparison between the estuaries of the Elbe (Germany), the Schelde (The Netherlands/Belgium) and the Gironde (France).

Kromkamp, J., J. Peene, P. van Rijswijk, P. van Breugel & N. Goosen. A comparison of phytoplankton primary production in three turbid, European estuaries: The Elbe, Wester- schelde and Gironde.

Hernandez, A., N. Raymond, J. Castel & P. Caumette. Spring distribution of bacterioplankton and production-predation relationships in the Gironde, Elbe and Schelde estuaries.

Bonin, P., N. Raymond, A. Hernandez & P. Caumette. Interaction between nitrate freshwater input, particulate organic matter and denitrification in the maximum turbidity zone in three european estuaries: Elbe, Gironde, Scheldt.

Raymond, N., A. Hermandez & P. Caumette. Sulfate reduction in estuarine sediments.

Tackx, M.L.M., R. Billones, L. Zhu, N. Daro & A. Flachier. Feeding of Zooplankton on mi­croplankton and detritus.

Irigoien, X. and J. Castel. A simple mathematical model to estimate net primary production in the Gironde estuary.

Irigoien, X., D. Burdloff, J. Castel & H. Etcheber. The role of the maximum turbidity zone controlling phytoplankton production and organic matter quality.

Irigoien, X., J. Castel & S. Gasparini. Gut clearance rate as predictor of food limitation situa­tions. Application to two estuarine copepods, Acartia bifilosa and Eurytemora affinis.

Gasparini, S., J. Castel & X. Irigoien. Egg production, growth rate and nutritional activity of the estuarine copepod Eurytemora affinis in the MTZ.

Gasparini, S., J. Castel & X. Irigoien. Longitudinal distribution of the copepod Eurytemora affinis (Poppe) in 3 european estuaries.

Fockedey, N. & J. Mees. Theme 4: Higher trophic levels, partim hyperbenthos.

Jak, R.G. & K.J.M. Kramer. Adjustment of low oxygen regimes and effects of hypoxia on an estuarine plankton community in enclosures.