SEARESEARCH 6 N and Ô C Dynamics of Suspended Organic

In Collaboration with ------the Netherlands Institute for Sea Research JOURNAL OF SEARESEARCH

ELSEVIER Journal of Sea Research 48 (2002) 1-15 www. elsevier. com/locate/seares

6 N and ô C dynamics of suspended organic matter in freshwater and brackish waters of the Scheldt estuary

L. De Brabandere a'*, F. Dehairs a, S. Van Dammeb, N. Briona, P. Meireb, N. Daroc

3Laboratory o f Analytical Chemistry, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussel, Belgium bDepartment o f Biology, Universitaire Instelling Antwerpen, Universiteitsplein 1C, B-2610 Wilrijk, Belgium cLaboratory o f Ecology and Systematics, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussel, Belgium

Received 4 October 2001; accepted 22 March 2002

A bstract

Suspended particulate organic matter was sampled monthly between June 1999 and April 2000 in the Scheldt river and estuary to investigate the seasonal and spatial patterns of ô13C and ô15N signatures. ô15N of suspended matter showed large seasonal variation. Minimum values ranged from — 0.5 %c in the freshwater zone (spring situation) to + 2.3%o in the mesohaline zone (winter situation). Maximum values (summer situation) ranged from + 8 . 8 %o in the freshwater zone to + 12.9 %c in the mesohaline zone. ô13C showed less seasonal variation and ranged overall from — 31.1 %c in the freshwater zone to —23.7 %c in the mesohaline zone. During the growth season, decrease of ô13C and increase of ô15N of suspended matter were due to local phytoplanktonic and bacterial biomass. There is strong evidence that the 15N enrichment of suspended matter during the growth season reflects the 15N enrichment of the ambient NH 4 pool induced by nitrification and NH 4 uptake.Zooplankton in the mesohaline section of the river was consistently enriched in 15N relative to suspended matter but followed its seasonal trend. During summer and autumn the isotopic offset between Zooplankton and the suspended particulate organic matter was consistent with a pattem of selective feeding on phytoplankton. During summer, ô15N of Zooplankton reached a value as high as +25.5 %o, the highest value observed during this study. During spring, present-day ô15N of suspended matter in the oligohaline and mesohaline section increased compared to the 1970s, probably because today nitrification, which enriches the NH4 pool in 15N, starts earlier in the season. For summer, the discrepancy between present-day suspended matter ô15N values and those observed in the 1970s was even larger, especially in the oligohaline and freshwater reaches, probably as a result of improved O 2 conditions now favouring nitrification. Likewise, the present decreased input of

Keywords: ô15N; ô13C; NHÎ; Zooplankton; Suspended organic matter; Seasonal variation

1. Introduction fate and seasonal processing of suspended particulate organic matter (SPOM) in riverine and estuarine envi Numerous studies have illustrated that natural sta ronments (e.g., Gearing et al., 1984; Mariotti et al., ble isotopes are a useful tool to investigate origin, 1984; Owens, 1985; Cifuentes et al., 1988, 1989; Montoya et al., 1991; Fichez et al., 1993; Canuel et Corresponding author. al., 1995; Qian et al., 1996; Ostrom et al., 1997; E-mail address: [email protected] (L. De Brabandere). Middelburg and Nieuwenhuize, 1998). ô15N com-

1385-1101/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 5 -1 1 0 1 (0 2 )0 0 1 3 2 -6 2 L. De Brabandere et ál. / Journal o f Sea Research 48 (2002) 1-15 monly shows larger differences between reservoirs reprocessing of particulate organic matter and a close than ô13C and could be a more sensitive indicator of coupling of production and consumption processes origins and biogeochemical processing (Ostrom et al., mediated by algae and bacteria (Middelburg and 1997). Stable nitrogen isotope ratios have been studied Nieuwenhuize, 2000). to track anthropogenic nitrogen in estuarine food Several earlier studies focused on the C and N webs, to detect causes of eutrophication (McClelland isotopic signature of suspended organic matter and et al., 1997; Riera et al., 2000) and to trace biogeo- phytoplankton in the Scheldt system (e.g., Laane et chemical processes that act on the dissolved inorganic al., 1990; Middelburg and Nieuwenhuize, 1998; Hel nitrogen pool in estuarine systems (McClelland and lings et al., 1999, 2001), but our knowledge about the Valida, 1998). different processes in control is still incomplete. The The Scheldt estuary is a temperate well-mixed tidal main objective of this study is to further document and estuary characterised by the occurrence of a maximum understand the seasonal variability of ô13C and 8 15N of turbidity zone (Middelburg and Nieuwenhuize, 1998; SPOM of the Scheldt estuary and to extend the inves Herman and Heip, 1999) and long water residence tigation into the freshwater reaches. Further objectives times of two to three months (Soetaert and Herman, are to understand the seasonal dependency of Z o o 1995a; Van Damme et al., 1999). Phytoplankton plankton b 13 N composition on the one of suspended blooms at different timings in different areas of the matter and to compare today’s seasonal trends of estuary. In the uppermost freshwater reaches (>km suspended matter isotopic composition with earlier

1 2 0 ), which receive phytoplankton advected from the observations. Our hypothesis is that ô1 5 N SPOm will tributaries, chlorophyll-a concentrations up to 70 |ig have increased over the years, following the improved dm “ 3 are found during spring (Muylaert et al., 1997, O2 conditions and related increase in nitrification. 2000). In the lower freshwater reaches (between km 97 and km 120) Chl-a concentrations exceed 100 |ig d m ~ 3 during the phytoplankton bloom in summer 2. Methods (Muylaert et al., 1997, 2001). The highest Chl-a contents occur in the oligohaline and mesohaline areas 2.1. Study area ( > 2 0 0 |ig dm “ 3) during the bloom period extending from spring to early summer (Soetaert and Herman, The Scheldt river (Fig. 1) is a lowland rain river 1994; Muylaert and Sabbe, 1999). Lowest Chl-a with a seasonally varying freshwater discharge (aver contents (up to 2 0 |ig dm “ 3 ) are found in the polyha age 100 m 3 s - x; Heip, 1988). Freshwater discharge to line and marine stations (Soetaert and Herman, 1994). the estuary is several orders of magnitude smaller than In case of long residence times of the water, nutrients tidal exchange (Soetaert and Herman, 1995a). This and plankton produced in situ, or imported, undergo results in long water residence times of two to three significant biogeochemical modification (Cifuentes et months (Soetaert and Herman, 1995a; Van Damme et al., 1988; Middelbrug and Nieuwenhuize, 1998) and al., 1999) and a salinity gradient intruding to about physical mixing (Cifuentes et al., 1988). Biogeochem 1 0 0 km upstream from the river mouth (km 0 ) ical transformations of nutrients and organic matter (Soetaert and Herman, 1995a). The estuary can be induce seasonal variability of isotope ratios and affect divided into three main zones: a marine (km 0 to km the isotopic composition to a greater extent than does 40), brackish (km 40 to km 97) and freshwater zone physical mixing (Cifuentes et al., 1988). The input of (km 97 to km 160 ) that represents one of the largest organic matter in the Scheldt estuary is high because freshwater tidal areas in Western Eiuope. The brackish the river drains one of the most densely populated and zone itself is divided into a mesohaline (km 40 to km industrialised areas of Eiuope (Frankignoulle et al., 57) and an oligohaline zone (km 57 to km 97). The 1996; Baeyens et al., 1998) and biogeochemical latter is characterised by a steep salinity gradient reprocessing of this material results in a net hetero- between km 57 and km 80 (Van Damme et al., trophic system sustaining significant CO 2 efflux 1999). The maximum turbidity zone extends roughly (Frankignoulle et al., 1998; Hellings et al., 2001). from km 90 to km 110 (Van Damme et al., 1999). The Also, NH 4 is efficiently recycled, implying extensive maximum tidal amplitude (5.3 m) occius at Schelle L. De Brabandere et al. / Journal of Sea Research 48 (2002) 1—15 3°00’E 3°20’ 3°40’ 4°00’ 4°20’ 4°40’

The Netherlands Vlissingen Hansweert North Sea km 20

- Belgium km 0 Zandvliet km 40 Terneuzen km 57 Westerschelde

/ ' Antwerp km 78 Zeeschelde Temse Schelle km 97 Rupel Gent

Germany Dendermonde km 121

'Scheldt, Dender Bovenschelde

Fig. 1. Map of the Scheldt estuary showing the location of the sampling stations for SPOM and Zooplankton. Numbers represent the distance in km from the mouth of the estuary.

(km 90) in the freshwater part of the estuary (Claes- Zandvliet station (km 57) is located in the mesoha- sens, 1988). line zone where a strong salinity gradient occurs Between June 1999 and April 2000 (December (Van Damme et al., 1999). Here, salinity ranges from not sampled), sixteen stations along the river and 1.8 to 13.8 PSU, while the yearly average is 8.9 estuary were sampled for physico-chemical parame PSU. ters (nitrate, ammonium, dissolved oxygen, temper Fig. 2 shows the freshwater discharge recorded at ature and salinity). At four of these stations km 90 (Schelle) for the period between June 1999 (Dendermonde, Temse, Antwerp and Zandvliet) and April 2000 (data from Tavemiers, 2001). River SPOM and Zooplankton were sampled. Dender discharge fluctuated between 50 and 125 m 3 s - 1 monde station (km 1 2 1 ) is located in the freshwater (summer-autumn), while higher values (up to 425 m 3 zone. Temse station (km 97) and Antwerp station s " 1, late December) were recorded in winter and (km 78) are located in the oligohaline zone, upstream spring. and downstream from the Rupel mouth, respectively. The area Temse - Antwerp is influenced by dis 2.2. Physico-chemical parameters charge from the river Rupel receiving untreated sewage from the Brussels sewage collectors. At Samples for nutrients (NH4 and NO3 ) were taken Temse station, salinity ranges from 0.4 to 1.1 PSU, just below the water surface with a clean PE bucket. with a yearly average of 0.7 PSU, while at the more Samples were stored in glass bottles, kept in cool downstream station of Antwerp salinity ranges from boxes and analysed within 24 h in the home laboratory 0.4 to 8.7 PSU, with a yearly average of 2.6 PSU. using a Skalar auto-analyser. Salinity, temperature and 4 L. De Brabandere et al. / Journal of Sea Research 48 (2002) 1—15

450 remove carbonates. This disc was packed in a tin cup 400 ready for combustion in the elemental analyser (Carlo 350 Erba NA1500). C0 2 gas produced during combustion 300 was led into a boro-silicate vacuum line and cryogeni- 250 cally trapped in glass tubes, which were subsequently ês 200 sealed with a hand torch (Hellings et al., 1999; S 150 Hellings, 2000). 100 For ô15N, more material was needed. This was 50 obtained by scraping the filtered matter from the filter

0 with a clean scalpel and transferring it into a tin cup. N 2 161 191 222 253 283 314 344 375 406 435 466 gas formed during combustion in the elemental analy Julian days 1999/2000 ser was led into in a stainless steel vacuum line and cryogenically trapped in stainless steel tubes fitted with Fig. 2. Discharge (m 3 s _ !) of the Scheldt estuary measured at a gas-tight valve and filled with molecular sieve (Mar- Schelle, km 90 (Data by Tavemiers, 2001). Arrows indicate the sampling events. guillier et al., 1997; Bouillon et al., 2002). Mass spectrometric measurements were per formed using a Delta E Finnigan Mat dual inlet dissolved 0 2 were measured in situ using a HYDRO isotope ratio mass spectrometer. Reference materials LAB 3® Data Probe. for C were graphite (USGS-24: 8 13C = - 16.1 sucrose (IAEA-C- 6 : — 10.4 %o) and polyethylene 2.3. SPOM collection foil (IAEA-CH-7: 8 13C = -3 1 .8 % c). Values are expressed relative to the VPDB (Vienna Peedee SPOM for isotope analyses was collected by sam Belemnite) standard. For nitrogen, high purity tank pling surface water with a clean PE bucket. Depend nitrogen gas was used as a working standard. This ing on suspended matter load, 80 to 300 cm 3 of water working standard was calibrated against ammonium were immediately filtered through Whatman GF/C sulphate (IAEA-N1: Ô15N = +0.4%o, IAEA-N2: glassfiber filters (0 = 47 mm). After filtration, ô15N = +20.4%o) and potassium nitrate (IAEA- samples were quickly frozen using liquid nitrogen. NO-3: ô15N = +4.7%c). ô15N values are expressed In the laboratory, samples were thawed and dried for relative to atmospheric N 2 reference. The precision several days at 50 °C. for 8 consecutive measurements was < 0 .1 %c for S15N and <0.04%c for 8 13 C. 2.4. Zooplankton

Copepods were sampled monthly at Zandvliet 3. Results station by towing a 300 pm Zooplankton net just below the water surface for 5 to 10 min. The copepods 3.1. Temporal and spatial variation of physico were kept in filtered Scheldt water for 2 h for gut chemical parameters content emptying. Then, samples were frozen in liquid nitrogen. In the laboratory samples were thawed by Temporal evolution of temperature and dissolved submerging them in distilled water. For each sample, 0 2 at the four stations is shown in Fig. 3. 0 2 600 to 800 calanoid copepods were handpicked for N concentrations during winter were markedly higher isotope analysis and dried at 50 °C to constant weight than during summer at all stations (Fig. 3). At Ant (between 0 .1 and 2 mg dry weight). werp, Temse and Dendermonde, the water was hypoxic ( < 2 mg dm " 3) during summer and autumn. 2.5. ô13C and ô15N analysis Lowest NH 4 concentrations (<100 pM) occurred in summer (July-September); (Fig. 4). Generally, NH4 For ô13C analysis, a disc (0 = 10 mm) was cut out at Zandvliet was lower than at the more upstream of the filter, and pre-treated with HC1 acid vapour to stations. Higher NH4 concentrations occurred in L. De Brabandere et al. / Journal of Sea Research 48 (2002) 1—15 5

26 26 10 24 Zandvliet 24 Antwerp 22 22 20 20 18 18 'O 16 16 OX 14 14 ft S 12 12 10 10 8 8 6 6 4 0 4 s > = > .Q s 5 o s o

1999/2000 1999/2000

26 10 26 10 24 Temse 24 Dendermonde 22 22 8 20 20 fo 18 18 'S 6 ^ (SX 16 16 S 14 14 ft 12 12 10 10 8 8 6 6 4 4 S3 ex a = 5 S 0) S •“ 3 s & 1999/2000 1999/2000

Fig. 3. Temporal variation of temperature (circles) and dissolved oxygen (mg dm ; triangles) at Zandvliet (km 57), Antwerp (km 78), Temse (km 97) and Dendermonde stations (km 121). autumn, winter (up to 406 pM, Dendermonde, and NH4 decreased slightly downstream, but NO3" December). The high discharge in December 1999 largely exceeded NH4 . In April also, NO3" exceeded (Fig. 2) is probably the cause of the decrease in NH4 and downstream of km 78, NH4 decreased NH 4 concentrations recorded in January 2000, par while NO3 increased slightly. ticularly at Temse and Dendermonde. Spatial patterns of NH4 and NO3 concentrations for a typical winter, 3.2. Temporal and spatial variability o f SPOM spring, summer and autumn month are shown in Fig. isotope ratios 5. In July, a strong decrease in NH4 between km

155 and km 133 coincided with a sharp increase of 3.2.1. Temporal and spatial variation ôin15N SPO m

NO3 . Downstream of km 133, NH4 and NO3 were Considerable temporal variation in ô1 5 N SPOm was relatively constant, but with NO3" largely in excess observed at the four sites sampled for SPOM (Fig. 4). of NH4 . In October, NH4 exceeded NO3 for the Generally, ô1 5 N SPOm was lower in winter, early spring section upstream of km 85 and a sharp decrease in and increased during spring, reaching a maximum in NH4 with simultaneous increase in NO3 occurred summer, followed by a decrease in late summer, between km 88 and km 72. In January, both NO3 autumn. However, the timing of extreme ô1 5 N SPOm 6 L. De Brabandere et al. / Journal of Sea Research 48 (2002) 1—15

14 14 Zandvliet Antwerp _ 12 12 10 10 8 8 6 6 4 4 2 2 0 500 0 500 450 450 400 400 350 350 300 300 250 250 ^ 200 200 150 E 150 100 100

o u a o a>o = u O a

14 14 12 Temse 12 Dendermonde _ 10 10 8 8 6 6 4 4 2 2 0 0 -2 500 -2 500 450 450 400 I 400 350 350 300 300 Ó" 250 250 200 200 150 150 X 100 100 £

= ÖJD a > « c = u s s Qi o a> a s 'S o a

1999/2000 1999/2000

Fig. 4. Temporal variation of ammonium (pM; closed squares), nitrate (pM; open squares) and ô 15N Sp o m (circles) at Zandvliet (km 57), Antwerp (km 78), Temse (km 97) and Dendermonde stations (km 121).

values (minimum and maximum) differed for the four minimum + 0.5 %o (March); Dendermonde maximum stations. + 11.5 %c (September), minimum — 0.5 %c (April and The most salient feature is the large discrepancy bet June). ween maximum and minimum values at all stations. The yearly averaged ô 15N Sp o m value was highest at Extreme values are: Zandvliet: maximum +12.9 %c the most downstream station Zandvliet ( + 8 . 6 %c ) and (June), minimum +2.3%o (January); Antwerp: max decreased upstream, with annual averages for Antwerp, imum + 10.8 %c (August and September), minimum Temse and Dendermonde of +5.6%o, +4.5 %c and + 1.3%o (March); Temse: maximum +8 .8 %o (July), + 4.0 %o, respectively. Zandvliet differed from the other L. De Brabandere et al. / Journal of Sea Research 48 (2002) 1—15 1

October 1999

500 - 450 - - 400 -

I 350 ' r í 300 - 0 * 250 -

” 200 -

1 150 - 100 - 50 -

0 -

Distance to the mouth of the estuary (km) Distaice to the mouth of the estuary (km)

January 2000 April 2000 A T D

O O Z Z

X zX z 150 -

100 -

40 60 80 100 120 140 40 60 80 100 120 140 160

Distance to the mouth of the estuary (km) Distance to the mouth of the estuary (km)

Fig. 5. Spatial variation of ammonium (pM; closed circles) and nitrate (pM; open circles) during a typical summer, autumn, winter and spring season. Z = Zandvliet (km 57), A = Antwerp (km 78), T = Temse (km 97) and D = Dendermonde stations (km 121). The upstream boundary of the area of intense nitrification can be recognised by the sharp decline of NH 4 + coinciding with a sharp increase of N 0 3 “ . The arrow indicates the position of the zone of intense nitrification during the seventies according to Mariotti et al. (1984).

stations by showing enriched ô1 5 N SPOm values + 1.5 %c to +6.0 %o for the oligohaline and mesoha throughout spring and summer (March to November). line part of the estuary. However, they are in good

At the other stations ô1 5 N SPOm started to decrease agreement with more recent values reported by Mid earlier (from September on). delburg and Nieuwenhuize (1998) and averaging

Our ô1 5 N SPom values are generally much higher + 12.0 %o for the same river section. The very high than those observed some twenty years ago by Mari values for SPOM (up to + 24 %c) reported by Mariotti otti et al. (1984), who reported values ranging from et al. (1984) for downstream Scheldt (unspecified 8 L. De Brabandere et al. / Journal of Sea Research 48 (2002) 1—15 area) in early summer were not observed here, nor in the freshwater, oligohaline and mesohaline sections of Middelburg and Nieuwenhuize (1998). the estuary closely overlaps with the ones reported in previous studies ( — 25.0 %o to — 32.2 %o; Laane et 3.2.2. Temporal and spatial variationô in 13C SPOm al., 1990; Middelburg and Nieuwenhuize, 1998; Hel ô13C values of SPOM ranged from — 31.1 %o to lings, 2000; Hellings et al., 1999; 2001). — 23.7 %c and also showed seasonal patterns (Fig. 6 ), but these were less pronounced than for ô 15N. 3.3. Temporal variation of ô15N of calanoid copepods ô 13 C Sp o m increased downstream with annual averages at Zandvliet of — 29.1 %c at Dendermonde and Temse, — 28.1 %c at Antwerp and — 26.6 %c at Zandvliet. Maximum ô15N values of calanoid copepods

Generally, ô1 3 C SPOm values were lowest in spring ( + 25.5%o) were observed in July and values stayed and summer. For the stations of Zandvliet and Ant high until November (Fig. 7). In January ô15N values werp maximum values were reached in January, while had decreased to + 13.5 %c. A further slight decrease at Temse and Dendermonde maxima were reached was observed from February to March. Average dif only in March. The range of our ô1 3 C SPOm values in ference between ô1 5 N SPOm and ô 15NCaianoids was +

Zandvliet Antwerp

-24 -24

¿ -26 -26 OS mcu P -28 -28 ~cO

-30 -30

-32 -32 5 ÖD & o s o> o O fc Q 1999/2000 1999/2000 -22 -22 Temse Dendermonde -24 -24

¿ -26 -26 s o Ph P -28 -28 cO

-30 -30

-32 -32 m a tt k y a U 0 -o u s « X o « On u 01 « « O n <¡ cc O £ Q < 1999/2000 1999/2000

Fig. 6. Temporal pattem of ô13C Sp o m in the Scheldt estuary at Zandvliet (km 57), Antwerp (km 78), Temse (km 97) and Dendermonde stations

(km 1 2 1 ). L. De Brabandere et al. / Journal of Sea Research 48 (2002) 1—15

30 (Wada, 1980; Owens, 1985; Ostrom et al., 1997) 25 and bacterial biomass (Caraco et al., 1998). Outside the bloom period, isotopic signatures of SPOM are 20 Calanoidea likely to shift towards those for terrigenous detritus & and domestic sewage end-members. 15 Z To 10 4.1. Can microbial biomass account fo r the 15N SPOM enrichment of SPOM during bloom? 5 Since, at present, there are no ô15N data for pure 0 a w> & s- s s

November (except August: 19 pM; Fig. 4), we Lower ô 15N S p o m values will be found during periods assume that ô15Nnh| values in the range +23 to of abundant N H 4 and low productivity. As a result, the

+ 29%o, as reported by Mariotti et al. (1984) apply discrepancy between ô 15N SPO m and ô15N of the food also at present. substrate effectively consumed by copepods will vary During microbial uptake of NH 4 significant dis seasonally. crimination against 15N occurs. For natural marine bacterial assemblages growing in a system with high 4.2. Spatial and seasonal patterns o f S13Cspom and

NH 4 regeneration, Hoch et al. (1994) report a dis ô 15N S p o m crimination of 10 %c during bacterial NH 4 uptake.

An average discrimination of 9.1 %c was reported for The different temporal patterns of ô1 5 N SPOm and algae during a bloom period in the Delaware estuary d 13CSpoM observed at the four study sites reflect (Cifuentes et al., 1989) while values between 6.5 %c differences in the timing of the phytoplankton bloom and 8 %c were reported for Chesapeake Bay (Mon (Fig. 9). toya et al., 1991). Both these estuaries have a NH4 - based productivity, as is the case for the Scheldt estuary (Mariotti et al., 1984; Middelburg and Nieu 14 y = -2.3151*ln(x)+14.944 wenhuize, 2000). For the purpose of the present 12 r2= 0.81 discussion we will assume that bacteria and phyto 10 plankton discriminate by 8 %c against the 15N isotope 8 during NH 4 uptake (i.e. an average of the values ZT 6 reported by the other authors). Given that ô 15NN H 4 values vary between + 23 and + 29 %c during periods 4 of low NH 4 and intense nitrification, ô15N of the 2 microbial community should vary between +15 %c 0 and +21 %c. For Zandvliet, this range overlaps with D d -2 the value we calculated above for the food substrate 0 50 100 150 200 250 300 350 during June-November (+18.9%o) as based on the [NH4+] ( u M) isotopic composition of copepods. Since at the other stations, the NH 4 concentration is low from July to Fig. 8 . Relationship between ammonium (pM) and ô 15N Sp o m i n the September (For Antwerp from June to September) the Scheldt estuary. Z = Zandvliet (km 57), A = Antwerp (km 78), T = ô15N signal of the remnant NH 4 pool will probably Temse (km 97) and D = Dendermonde stations (km 121). L. De Brabandere et al. / Journal of Sea Research 48 (2002) 1—15 11

14 14 Zandvliet Antwerp 12 12

10 10

8 8

o o PLh ZA '• »IO JF 4 4

2 2

0 0

-2 •2 -32 -30 -28 -26 -24 -22 -32 -30 -28 -26 -24 -22

,(%o)

14 Temse Dendermonde 12

0 • 6 • 4 •2 -32 -30 -28 -26 -24 -22

SPOM ""'SPOM ( % ° )

Fig. 9. ö13C Spo m versus ö 15N Sp o m in the Scheldt estuary at Zandvliet (k m 57), Antwerp (km 78 ), Temse (k m 97) and Dendermonde stations (k(km m 121).1 2 1 ). ArrowsA point to the direction to which ô13C Spo m and ô 15N Spo m shift along the annual cycle (from June 1999 (month 6) to April 2 0 0 0 (month 4).

During winter (months 1 to 2 ) ô1 5 N SPOm is low and the preferential uptake of 14NH 4 .Similarly, the ô13Cspom high for all stations. Low microbial bio decrease in ô13C of SPOM likely reflects the effect mass and higher discharge are likely reasons for this. of autotrophic fixation of carbon from a DIC pool Peak discharges such as the one recorded during enriched in 12C 0 2 during winter (Hellings et al., 1999, December (Fig. 2) can advect large amounts of 2001; Hellings, 2000). terrestrial organic detritus having lower ô15N and From late spring to late summer (months 6 to 9), higher ô13C signatures (Hellings et al., 1999; Hellings, Antwerp, Temse and Dendermonde show a strong

2 0 0 0 ) and wash out local microbial populations increase in ô1 5 N SPOm- This coincides with lowered (Brion et al., 2000; Muylaert et al., 2001). NH 4 due to uptake during bloom events (e.g., For Antwerp, Temse and Dendermonde the early Cifuentes et al., 1989) and enhanced nitrification season (months 1 to 4) shows a slight decrease in ô15N during spring-summer (e.g., Iriarte et al., 1998; Brion, and ô13C (the latter not for Temse). This early season 1997). These processes induce a progressive 15N decrease in ô1 5 N SPOm could reflect microbial activity enrichment of the NH4 P0 0 I. During months 6 to 9, under conditions of high ambient NH 4 (Fig. 4), due to ô1 3 C SPom remains relatively constant and low at Ant- 12 L. De Brabandere et al. / Journal o f Sea Research 48 (2002) 1-15 werp and Temse, while at Dendermonde we observe a sections of the estuary are higher than the ones slight increase in ô 13 C s p o m - This could reflect an reported for the 1970s by Mariotti et al. (1984). For increased demand for CO 2 by blooming phytoplankton the Temse to Zandvliet section (km 78 to 57) during depending on dissolved inorganic carbon (DIC) that April, values ranged from + 1.5%o to + 5%o in the became progressively enriched in 13C over the growth 1970s, while today values range between +2.7 and season as a result of previous phytoplankton activity + 8 . 6 %o (compare their Fig. 8 with oiu Fig. 4). During (Farquhar et al., 1982; Hinga et al., 1994; Rau et al., summer, the oligohaline section (Temse to Antwerp) 1996; Hellings et al., 1999, 2001). also shows an increased ô 15N S P O m signal today During autumn (months 10 to 11), ô15N SPOm at ( + 8.7%o for Temse in July and +10.8%o for Ant Antwerp, Temse and Dendermonde is lower. ô13C for werp in August; Fig. 4), while for the same section in the former two stations shows little change, while at the 1970s values did not exceed +5 %o. However, in Dendermonde ô13C increases after an initial decrease. the downstream area diuing summer the situation This situation reflects a decreased microbial and an might be reversed. Indeed values as high as increased terrestrial detritus contribution to SPOM. + 24%o were reported for the 1970s by Mariotti et Zandvliet station is peculiar in that ô 15Nspom is al. (1984), and these were attributed mainly to high from month 4 to 11 ( + 8 .6 %o; Fig. 9). Because autochtonous phytoplankton. Oiu present-day highest NH4 content is low diuing this whole period of high summer values at Zandvliet are + 12.9 %c (June), but ô15NSpojvp it would appear at first sight that local since the downstream summer sampling area is not phytoplankton and bacteria thriving on this reduced detailed by Mariotti et al. (1984), it is difficult to nutrient pool are responsible for this situation. It is compare their values with oius. In any case, Middel unlikely that these lasting high ô15N values are burg and Nieuwenhuize (1998 ), who investigated the sustained by local phytoplankton only, since phyto Scheldt over its polyhaline to oligohaline sections in plankton contribution of SPOM is relatively small in 1994 (August), did not observe the very high down the Zandvliet area compared to upstream stations stream ô15NSPom values of the 1970s, indicating that (Muylaert and Sabbe, 1999) and since the growth for that section, too, significant changes have occurred season in this mesohaline zone extends only from over time. spring to early summer (Soetaert and Herman, 1994; For the mesohaline - oligohaline section, the Muylaert and Sabbe, 1999). For Zandvliet we spec increase in ô15NSPOm probably reflects improved ulate that the lasting 15N enrichment of SPOM is conditions for nitrification, as NH 4 concentrations caused by bacteria processing phytoplankton detritus are generally lower at present than in the 1970s imported from upstream regions. Indeed, high salinity (Van Damme et al., 1999). Today 15N enrichment of stress induces phytoplankton mortality (Muylaert and SPOM occius earlier in the season and is observed in Sabbe, 1999; Van Damme et al., 1999; Goosen et al., more upstream areas, probably because nitrification 1999) and it is thus possible that phytoplankton starts earlier in the season (there is even evidence for washed out from the freshwater reaches dies off in winter nitrification, N. Brion, unpublished results) and the strong salinity gradient of the mesohaline zone occius more upstream in the estuary. Oiu nutrient data close to Zandvliet. Furthermore, degrading phyto indicate that the area of intense nitrification during plankton becomes enriched in 15N as a result of summer is now situated upstream of km 130 (Fig. 5), bacterial processing (Wada, 1980; Owens, 1985; a situation already documented in the 1990s (Soetaert Ostrom et al., 1997) and bacteria colonising phyto and Herman, 1995b), while in the 1970s nitrification plankton detritus will be enriched in 15N since they was insignificant upstream of km 70 (Mariotti et al., experience low ambient NH 4 in the Zandvliet area 1984; Billen et al., 1985). This shift occurred despite (Fig. 4). occasional low O 2 contents ( < 1 mg dm “ 3) in these freshwater reaches (Fig. 3), but Van Damme et al.

4.3. Long-term variation oô I5 f N s p o m (1999) suggest that nitrification can probably proceed in hypoxic conditions when coupled with oxygen Oiu spring (April) and summer (June to August) production by phytoplankton. As an alternative explan ô 15N S P o m values for the mesohaline and oligohaline ation for the present generally increased ô 15N SPOm L. De Brabandere et al. / Journal o f Sea Research 48 (2002) 1-15 13 signals for the mesohaline - oligohaline sections in anaerobic in the 1970s and would have experienced spring and summer, we can invoke a decreased input of low nitrification most of the year. domestic sewage. Domestic sewage has a very light During most of the year, zooplankton- 8 15N in the ô15N signal ( + 2 %o\ Fisseha, 2000) and a reduction of mesohaline section followed the one of SPOM but sewage load relative to other less light N components with an offset exceeding by far the normal increment would increase the present ô15N signal of SPOM. associated with trophic level, probably as a result of Indeed, at present the input of untreated sewage comes selective grazing on phytoplankton. The large 15N mainly from the city of Brussels ( 1. IO 6 inhabitants), enrichment of Zooplankton (up to +25.5 %o) during whereas in the 1970s, mostly untreated sewage was summer is among the highest observed in estuarine released also by Antwerp and Ghent ( 6 .IO6 inhabi systems and reflects the intensity of nitrification tants). today. It is likely that the seasonal 15N enrichment will be transferred also to the higher trophic levels. Future studies of trophic relationships in the Scheldt 5. Conclusions system will have to consider carefully these strong fluctuations of isotopic composition at the lower During this study, we observed considerable spa trophic levels. tio-temporal variability of the 8 15N and 8 13C com position of suspended particulate organic matter in the Scheldt estuary. In general, the ô 13C s p o m signal Acknowledgements followed a quite predictable seasonal trend, set mainly by phytoplankton activity, with least negative This research was carried out in the framework of values in winter and most negative values in spring, research project “Biotic Interactions in Turbid Estuar summer. The spatio-temporal variation of ô 15N SPOm! ine Systems” (G.0104.99) supported by the Fund for however, was more complex. At Antwerp, Temse Scientific Research (Flanders). The fieldwork was and Dendermonde highest a 13N values observed performed in the framework of the OMES initiative during the bloom period (spring, summer) were (Onderzoek Milieueffecten Sigmaplan) sponsored by attributed to the uptake of NH| enriched in 15N as the Flemish Government. We thank the offices of a result of ongoing uptake and nitrification. The “Maritieme Schelde” and “Zeeschelde” for use of the pattem observed for Zandvliet, more downstream in ships ‘Veremans’ and ‘Scaldis’ and for their assistance the mesohaline part of the river, was different with during fieldwork. We are grateful to E. Keppens for high ô 15N S P o m values from spring to autumn. The access to mass-spectrometry facilities and assistance persistent high 8 15N signal probably resulted from during IRMS analysis as well as to E. Boschker for the advection of 15N-enriched phytoplankton detritus discussions and useful criticisms on the manuscript. from upstream regions and from further 15N enrich ment during bacterial processing. Mariotti et al. (1984) observed an increase in ô 15N S P O m dining References the growth season for the mesohaline section. This trend is confirmed by our results, but there is Baeyens, W., Van Eck, B., Lambert, C., Wollast, R., Goeyens, L., evidence that the zone of intense nitrification in 1998. General description of the Scheldt Estuaiy. Hydrobiologia summer has shifted upstream relative to the situation 366, 1-14. Billen, G., Somville, M., Debecker, E., Servais, R, 1985. A nitrogen in the 1970s. Also, the nitrification period now budget of the Scheldt hydrographical basin. Neth. J. Sea Res. appears to start earlier (in winter-spring) resulting 19, 223-230. in higher ô 15N S P O m values during spring than in the Bouillon, S., Koedam, N., Raman, A.V., Dehairs, R, 2002. Primary 1970s. Oin ô 15N S P o m data for Dendermonde (range: producers sustaining macro-invertebrate communities in interti — 0.5 %o to +11.5%o) are the first reported for the dal mangrove forests. Oecologia 130, 441-448. Brion, N., 1997. Etude du processus de nitrification à 1’ échelle de freshwater part of the estuary. The largest change in grands réseaux hydrographiques anthropisés. Ph.D. Thesis, Uni f) 13 N composition of SPOM has probably occurred in versité Libre de Bruxelles, Belgium (In French). the freshwater part of the Scheldt that used to be Brion, N., Billen, G., Guezennec, L., Ficht, A., 2000. Distribution of 14 L. De Brabandere et al. / Journal o f Sea Research 48 (2002) 1-15

nitrifying activity in the Seine River (France) from Paris to the of the Scheldt estuary as traced by stable carbon isotope com estuaiy. Estuaries 23, 669-682. position. Biogeochemistry 47, 167-186. Canuel, E.A., Cloem, J.E., Ringelberg, D.B., Guckert, J.B., Rau, Hellings, L., Dehairs, F., Van Damme, S., Baeyens, W., 2001. Dis G.H., 1995. Molecular and isotopic tracers used to examine solved inorganic carbon in a highly polluted estuary (the sources of organic matter and its incorporation into the food Scheldt). Limnol. Oceanogr. 46, 1406-1414. webs of San Francisco Bay. Limnol. Oceanogr. 40, 67-81. Herman, P.M.J., Heip, C.H.R., 1999. Biogeochemistry of the MAx- Caraco, N.F., Lampman, G., Cole, J.J., Limburg, K.E., Pace, M.L., imurn TURbidity Zone of Estuaries (MATURE): some conclu Fischer, D., 1998. Microbial assimilation of DIN in a nitrogen sions. J. Mar. Syst. 22, 89-104. rich estuary: implication for food quality and isotope studies. Hinga, K.R., Arthur, M.A., Pilson, M.E.Q., Whitaker, D., 1994. Mar. Ecol. Prog. Ser. 167, 59-71. Carbon isotope fractionation by marine phytoplankton in cul

Checkley, D.M., Entzeroth, L.C., 1985. Elemental and isotopic frac ture: the effects of CO2 concentration, pH, temperature, and

tionation of carbon and nitrogen by marine, planktonic cope species. Global Biogeochem. Cycles 8 , 91-102. pods and implications to the marine nitrogen cycle. J. Plankton Hoch, M.P., Fogel, M.L., Kirchman, D.L., 1994. Isotope fractiona Res. 7, 553-568. tion during ammonium uptake by marine microbial assemb Ciftientes, L.A., Sharp, J.H., Fogel, M.L., 1988. Stable carbon and lages. Geomicrobiol. J. 12, 113-127. nitrogen isotope biogeochemistry in the Delaware estuary. Lim Iriarte, A., De la Sota, A., Orive, E., 1998. Seasonal variation of nol. Oceanogr. 33, 1102-1115. nitrification along a salinity gradient in an urban estuary. Hydro- Cifuentes, L.A., Fogel, M.L., Pennock, J.R., Sharp, J.H., 1989. biologia 362, 115-126. Biogeochemical factors that influence the stable nitrogen iso Keough, J.R., Sierzen, M.E., Hagley, C.A., 1996. Analysis of a lake tope ratio of dissolved ammonium in the Delaware Estuary. superior coastal food web with stable isotope techniques. Lim Geochim. Cosmochim. Acta 53, 2713-2721. nol. Oceanogr. 41, 136-146. Claessens, J., 1988. Het hydraulisch regime van de Schelde. Water Laane, R.W.P.M., Turkstra, E., Mook, W.G., 1990. Stable carbon 43, 163-169 (In Dutch). isotopic composition of pelagic and benthic organic matter in Farquhar, G.D., O’Leaiy, M.H., Berry, J.A., 1982. On the relation the North Sea and adjacent estuaries. In: Ittekot, V., Kempe, S., ship between carbon isotope discrimination and the intercellular Michaelis, W., Spitzy, A. (Eds.), Facets of Modem Biogeochem carbon dioxide concentration in leaves. Aust. J. Plant Physiol. 9, istry. Fetschrift for the 60th Anniversary of E.T. Degens. Spring 121-137. er Verlag, Berlin, pp. 214-224. Fichez, R., Dennis, P., Fontaine, M.F., Jickells, T.D., 1993. Isotopic Marguillier, S., Van der Velde, G., Dehairs, F., Hemminga, M.A., and biochemical composition of particulate organic matter in a Rajagopal, S., 1997. Trophic relationships in an interlinked

shallow water estuary (Great Ouse, North Sea, England). Mar. mangrove-seagrass ecosystem as traced by ô15N and ô 13C. Chem. 43, 263-276. Mar. Ecol. Prog. Ser. 151, 115-121. Fisseha, R., 2000. Use of stable isotopes as a tracer of domestic Mariotti, A., Germon, J.C., Hubert, P., Kaiser, P., Letolle, R., Tar sewage pollution in the Woluwe Brook-Woluwe sewer collector dieux, A., Tardieux, P., 1981. Experimental determination of system. M.Sc. Thesis Environmental Science and Technology, nitrogen kinetic isotope fractionation: some principles; illustra Vrije Universiteit Brussel, Belgium. tion for the denitrification and nitrification processes. Plant Soil

Frankignoulle, M., Bourge, I., Wollast, R., 1996. Atmospheric CO 2 62, 413-430. fluxes in a highly polluted estuary (the Scheldt). Limnol. Ocean Mariotti, A., Lancelot, C., Billen, G., 1984. Natural isotopic com ogr. 41, 365-369. position of nitrogen as a tracer of origin for suspended organic Frankignoulle, M., Abril, G., Borges, A., Bourge, I., Canon, C., matter in the Scheldt Estuary. Geochim. Cosmochim. Acta 48, DeLille, B., Libert, E., Theate, J.M., 1998. Carbon dioxide 549-555. emission from European estuaries. Science 282, 434-436. McClelland, J.W., Valiela, I., 1998. Changes in food web structure Gearing, J.N., Gearing, J.P., Rudnick, D.T., Requejo, A.G., Hutch under the influence of increased anthropogenic nitrogen inputs ins, M.J., 1984. Isotopic variability of organic carbon in a phy- to estuaries. Mar. Ecol. Prog. Ser. 168, 259-271. toplankton-based, temperate estuary. Geochim. Cosmochim. McClelland, J.W., Valiela, I., Michener, R.H., 1997. Nitrogen-stable Acta 48, 1089-1098. isotope signatures in estuarine food webs: a record of increasing Goosen, N.K., Kromkamp, J., Peene, J., van Rijswijk, P., van Breu- urbanisation in coastal watersheds. Limnol. Oceanogr. 42, gel, P., 1999. Bacterial and phytoplankton production in the 930-937. maximum turbidity zone of three European estuaries: the Elbe, Middelburg, J.J., Nieuwenhuize, J., 1998. Carbon and nitrogen sta Westerschelde and Gironde. J. Mar. Syst. 22, 151-171. ble isotopes in suspended matter and sediments from the Scheldt Heip, C., 1988. Biota and abiotic environment in the Westerschelde estuary. Mar. Chem. 60, 217-225. estuaiy. Hydrobiol. Bull. 22, 31-34. Middelburg, J., Nieuwenhuize, J., 2000. Uptake of dissolved inor Hellings, L., 2000. Origin and fate of dissolved inorganic and par ganic nitrogen in turbid, tidal estuaries. Mar. Ecol. Prog. Ser. ticulate organic carbon in a highly polluted estuary (The 192, 79-88. Scheldt) as traced by stable isotopes. Ph.D. Thesis, Vrije Uni Montoya, J.P., Horrigan, S.G., McCarthy, J.J., 1991. Rapid, storm- versiteit Brussel, Belgium. induced changes in the natural abundance of 15N in a planktonic Hellings, L., Dehairs, F., Tackx, M., Keppens, E., Baeyens, W., ecosystem, Chesapeake Bay, USA. Geochim. Cosmochim. Acta 1999. Origin and fate of organic carbon in the freshwater part 55, 3627-3638. L. De Brabandere et al. / Journal o f Sea Research 48 (2002) 1-15 15

Montoya, J.P., Wiebe, P.H., McCarthy, J.J., 1992. Natural abun tidal benthic algae and invertebrates in the Scheldt Estuary (The dance of 15N in particulate nitrogen and Zooplankton in the Gulf Netherlands): effect of river nitrogen inputs. Est. Coast. Shelf Stream region and warm-core ring 86A. Deep Sea Res. 39, Sei. 51, 365-372. 363-392. Soetaert, K., Herman, P.M.J., 1994. One foot in the grave: Zoo Muylaert, K., Sabbe, K., 1999. Spring phytoplankton assemblages plankton drift into the Westerschelde estuary (The Netherlands). in and around the maximum turbidity zone of the estuaries of the Mar. Ecol. Prog. Ser. 105, 19-29. Elbe (Germany), the Schelde (Belgium/The Netherlands) and Soetaert, K., Herman, P.M.J., 1995 a. Estimating estuarine resi the Gironde (France). J. Mar. Syst. 22, 133-149. dence times in the Westerschelde (The Netherlands) using a Muylaert, K., Van Kerckvoorde, A., Vyverman, W., Sabbe, K., box model with fixed dispersion coefficients. Hydrobiologia 1997. Structural characteristics of phytoplankton assemblages 311, 215-224. in tidal and non-tidal freshwater systems: a case study from Soetaert, K., Herman, P.M.J., 1995b. Nitrogen dynamics in the the Schelde basin, Belgium. Freshw. Biol. 38, 263-276. Westerschelde estuary (SW Netherlands) estimated by means Muylaert, K., Sabbe, K., Vyverman, W., 2000. Spatial and temporal of the ecosystem model MOSES. Hydrobiologia 311, 225-246. dynamics of phytoplankton communities in a freshwater tidal Soetaert, K., Van Rijswijk, P., 1993. Spatial and temporal patterns of estuary (Schelde, Belgium). Est. Coast. Shelf Sei. 50, 673-687. the Zooplankton in the Westerschelde estuary. Mar. Ecol. Prog. Muylaert, K., Van Wichelen, J., Sabbe, K., Vyverman, W., 2001. Ser. 97, 47-59. Effects of freshets on phytoplankton dynamics in a fresh Tackx, M., Irigoien, X., Daro, N., Castel, J., Zhu, L., Zhang, X., water tidal estuary (Schelde, Belgium). Arch. Hydrobiol. 150, Nijs, J., 1995. Copepod feeding in the Westerschelde and the 269-288. Gironde. Hydrobiologia 311, 71-83. Ostrom, N.E., Macko, S.A., Deibel, D., Thompson, R.J., 1997. Tavemiers, E., 2001. Zeescheldebekken: de afvoer van de Schelde Seasonal variation in the stable carbon and nitrogen isotope in 2001. Report AMS-01/02. Ministerie van de Vlaamse Ge biogeochemistry of a coastal cold ocean environment. Geochim. meenschap, Departement Leefmilieu en Infrastmctuur, Afdeling Cosmochim. Acta 61, 2929-2942. Maritieme Schelde, Antwerp, Belgium (In Dutch). Owens, N.J.P., 1985. Variations in the natural abundance of 15N in Van Damme, S., Ysebaert, T., Meire, P., Van den Bergh, E., 1999. estuarine suspended particulate matter: a specific indicator of Habitatstructuren, waterkwaliteit en leefgemeenschappen in het biological processing. Est. Coast. Shelf Sei. 20, 505-510. Schelde Estuarium. Rapport Instituut voor Natuurbehoud, Brus Qian, Y., Kennicutt, M.C., Svalberg, J., Macko, S.A., Bidigare, sel, IN99/24 (In Dutch). R.R., Walker, J., 1996. Suspended particulate organic matter Wada, E., 1980. Nitrogen isotope fractionation and its significance (SPOM) in Gulf of Mexico estuaries: Compound-specific iso in biogeochemical processes occurring in marine environments. tope analysis and plant pigment compositions. Org. Geochem. In: Goldberg, E.D., Horibe, Y., Saruhashi, K. (Eds.), Isotope 24, 875-888. Marine Chemistry. Uchida Rokakuho Publishing Co. Ltd., To Rau, G.H., Riebesell, U., Wolf-Gladrow, D., 1996. A model of kyo, pp. 375-398. photosynthetic 13C fractionation by marine phytoplankton based Wu, J.P., Calvert, S.E., Wong, C.S., 1997. Nitrogen isotope varia

on diffusive molecular CO 2 uptake. Mar. Ecol. Prog. Ser. 133, tions in the subarctic Northeast Pacific: Relationships to nitrate 275-285. utilisation and trophic structure. Deep Sea Res. I — Oceanogr. Riera, P., Stal, L.J., Nieuwenhuize, J., 2000. Heavy ô15N in inter Res. Papers 44, 287-314.