www.elsevier.com/locate/marpolbul

Marine Pollution Bulletin 51 (2005) 119–127

Organic carbon deliveries and their flow related dynamicsin the Fitzroy estuary

Phillip Ford a,b,*, Pei Tillman a,b, Barbara Robson a,b, Ian T. Webster a,b

a CSIRO Land and Water, GPO Box 1666, Canberra, ACT 2601, Australiab Cooperative Research Centre for Coastal Zone, Estuary and Waterway Management, 80 Meiers Road, Indooroopilly, QLD 4068, Australia

Abstract

The Fitzroy estuary (Queensland, Australia) receives large, but highly episodic, river flows from a catchment (144,000km2) which

has undergone major land clearing. Large quantities of suspended sediments, and particulate and dissolved organic carbon are deliv-

ered. At peak flows, d13C (�21.7 ± 0.8&) and C/N (14.8 ± 1.3) of the suspended solids indicate that the particulate organic material

entering the estuary is principally soil organic carbon. At the lower beginning flows the particulate organic matter comes from in-

stream producers (d13C = �26&). The DOC load is about 10 times the POC load. Using the inverse method, budgets for POC and

DOC were constructed for high and low flows. Under high flows, only a small portion of the POC and DOC load is lost in the

estuary. Under dry season (low flow) conditions the estuary is a sink for DOC, but remains a source of POC to the coastal waters.

Crown Copyright � 2004 Published by Elsevier Ltd. All rights reserved.

Keywords: Dissolved; Particulate; Carbon; Soil; Budgets; Estuaries

1. Introduction

The Fitzroy river, Queensland, Australia (Fig. 1) is a

major source of sediments, and both particulate and dis-

solved nutrients to the southern lagoon of the Great

Barrier Reef. At Rockhampton, 60km upstream from

the mouth (and the last gauging point before the river

enters the estuary) the total sediment flux since 1950

has been estimated to be 260 million tonnes (Kelly and

Wong, 1996). The large catchment (144,000km2) wascovered with Brigalow scrub (Leguminosae: Acacia

harpophylla F. Muell) before European settlement in

the mid 19th century. Efforts to remove the extensive

vegetation cover were largely ineffectual until the

1960s. Since then there has been major land clearing

0025-326X/$ - see front matter Crown Copyright � 2004 Published by Else

doi:10.1016/j.marpolbul.2004.10.019

* Corresponding author. Address: CSIRO Land and Water, GPO

Box 1666, Canberra, ACT 2601, Australia. Tel.: +61 2 6246 5559; fax:

+61 2 6246 5560.

E-mail address: phillip.ford@csiro.au (P. Ford).

with much of the woodland replaced by wooded grass-

lands which are now mainly used for extensive cattlegrazing. Smaller areas are used for dry land agriculture

and irrigated horticulture. The sedimentary material

delivered to the head of the estuary consists mainly of

fine particles (�90% < 1lm) (Bormans et al., 2004) com-

posed of montmorillonite and other smectites. For the

adjacent Burdekin catchment, which has a similar

monsoonal climate, geology, and pattern of vegetation

clearing and landuse, sediment delivery post-Europeansettlement, is estimated to have increased 5- to 10-fold

(McCulloch et al., 2003). It is suggested that the sedi-

ment input from all Great Barrier Reef catchments has

increased, on average, by about a factor of 4 (Furnas,

2003) relative to pre-European times.

Rivers deliver both particulate and dissolved organic

matter to coastal regions via estuaries. An accurate

characterization of the nature and quantities of organicmaterial delivered, as well as the transformations of the

organic matter in the estuary, is central to understanding

vier Ltd. All rights reserved.

Fig. 1. Location map of the Fitzroy catchment, major contributing rivers, and Fitzroy estuary.

120 P. Ford et al. / Marine Pollution Bulletin 51 (2005) 119–127

the functioning of these systems in their present form.

Such knowledge is needed also to be able to make

well-founded predictions of the consequences for estua-

rine functioning, of changes to land use, or reductions inflows due to dam construction. The particulate organic

carbon (POC) arises from organic coatings on clay par-

ticles, and as well as fragments of vegetation, freshwater

phytoplankton and other forms of organic carbon. The

transported sediments will thus be a vector for the trans-

port of heavy metals and pesticides adsorbed to the or-

ganic layers (Karickhoff et al., 1979; Voice and Webber,

1983). When the terrigenous organic material enters theestuary it is subject to further heterotrophic consump-

tion by bacteria, as well as serving as food for some

higher organisms. While the sediment may decrease

the available light for photosynthetic organisms, the

associated organic carbon may be utilized by some sym-

biotic corals (Anthony and Fabricius, 2000) and com-

pensate for the energy losses due to the decreased light

and increased cleaning required. On a global scale thetransport of terrigenous organic carbon is the largest

flux of carbon from the land to the sea (Meybeck,

1982) and the rate of riverine discharge of POC is com-

parable with the global rate of accumulation of organic

carbon in all marine sediments (Berner, 1989). For all

estuaries, depending on the size of the riverine flows rel-

ative to the estuary volume and their tidal characteris-

tics, the residence time for organic matter in the

estuary can vary enormously. Even when the nominalestuarine transit time is short and microbially mediated

changes to terrigenous carbon are slight, physicochemi-

cal processes such as flocculation of fine particles at rel-

atively low salinities (S < 3&) (Millman et al., 1975) can

create more favourable conditions for breakdown and

greatly change the dynamics of the organic matter with-

in the estuary.

For sourcing and tracing offshore sediments it isimportant to be cognizant of the isotopic characteristics

of the organic material entering such coastal areas. On

occasion it has proved difficult to distinguish between

marine plankton remains and terrigenous organic mat-

ter delivered by nearby major rivers (Onstad et al.,

2000). The concept of dissolved organic carbon (DOC)

as encountered in the context of estuarine and marine

research needs to be carefully defined and related to cur-rent perspectives on the dissolved organic carbon gener-

ated in the upstream areas and delivered to the estuary

(where it becomes estuarine DOC). Riverine DOC is

produced primarily by leaching of leaf litter within the

stream, and by groundwater inflows which have infil-

Fitzroy River Discharge

1991 1993 1995 1997 1999 2001 20030

1000

2000

3000

4000

Dis

char

ge (m

3 s-1)

Fig. 2. Daily discharge showing highly episodic character of flows.

P. Ford et al. / Marine Pollution Bulletin 51 (2005) 119–127 121

trated through organic rich areas of the soil (reviewed in

Boulton et al., 1998). It is composed primarily of humicsubstances (Ertel et al., 1986) with lesser amounts of

polysaccharide carbohydrates and amino acids (Volk

et al., 1997). Part of this material can be taken up by

bacteria (O�Connell et al., 2000) and up to about 10%

of riverine DOC can be respired on passing through

the estuary (Moran et al., 1999). Riverine DOC has a

major impact on coastal DOC dynamics and forms part

of the microbial food web there (Zweifel et al., 1995).There are seemingly large, and unexplained differences

in the DOC degradation rates between nearby estuaries,

or within the same estuary.

In this study we report on the measurements made on

the delivery of particulate organic matter from the

catchment and make inferences as to how this has chan-

ged as a consequence of the human-generated changes in

vegetation and land use. In addition we quantify the re-moval and transformation processes for POC and DOC

in the estuary under high (i.e. flood conditions) and low

flow (very limited freshwater discharge).

The Fitzroy catchment lies at the boundary of the

tropical and temperate convergence zones. Both rainfall,

and runoff from the catchment, are highly seasonal

occurring mainly in the Antipodian summer leading to

large, but short-lived flows, which rapidly flush the estu-ary clean of salt water (Fig. 2). For the rest of the year

freshwater flows into the estuary are small. They arise

from limited discharges over the fish ladder at the Rock-

hampton barrage, and the release of treated waste water

(18Mld�1) at Rockhampton.

2. Experimental methods

2.1. Field site and sediment sample collection

One litre samples were collected daily during flow

events from the raw water inlet (1m beneath surface)

of Fitzroy River Water Corporation�s drinking water

treatment plant 6km above the barrage at Rockhamp-

ton, Queensland Australia. The polyethylene samplebottles had been previously acid washed and rinsed

twice with MilliQ water. The samples were preserved

by freezing at �20 �C. The samples were passed through

a <64lm sieve and the suspended sediment was recov-

ered by centrifugation (3000g, 15min) and dried to con-

stant weight at 105 �C. Carbon and nitrogen content and

the stable isotopic concentrations (d13C and d15N) were

determined simultaneously using an automatic C and Nanalyser interfaced to a Europa 20–20 mass spectro-

meter. The samples had been stood overnight over con-

centrated HCl to remove carbonate minerals. Stable

isotope ratios are expressed as parts per mil (&) devia-

tion from recognised world standards (PDB and air).

Measurement precision is ±0.2&.

2.2. Hydrology

Discharge was measured at the official gauging sta-

tion at the Gap (AWM 130005) about 80km upstream

of the barrage and then corrected for the transit time be-

tween the Gap and the downstream sampling station.

2.3. Estuarine water samples

Water samples for TOC (total organic carbon) and

DOC (dissolved organic carbon) were collected from a

depth of 0.25m, at 12 stations spaced along the estuary

downstream of the barrage. Sample collection was done

as closely as possible to the same time of the lunar cycle

each month. The DOC sample was filtered through a

0.45lm cellulose nitrate syringe filter (Sartorius), and

both the TOC and DOC samples were then preservedby freezing at �20 �C. On return to the laboratory the

samples were melted, thoroughly mixed, and, after acid-

ification, analysed using an Analytical 1010 TOC Ana-

lyser. Water column physical parameters (temperature,

salinity, pH, dissolved oxygen, and turbidity) were

measured at the surface, and then 1m intervals to the

bottom at each site, using a Hydrolab Minsonde 4a

water quality probe.

2.4. Calculation of POC and DOC budgets

Monthly budgets of TOC and DOC were calculated

using the box-model method of Webster et al. (2004).

Briefly, sample positions were corrected to position at

mid-tide based on a 1-D tidal model of the estuary cal-

ibrated against tidal observations at Port Alma near themouth, and 1/3 (Thompson�s Point) and 2/3 (Nerrim-

bera) of the length of the estuary. The changes in the ob-

served salinity distribution at the 12 stations was used to

estimate the spatially varying tidal dispersion coefficient.

The 1-D advection dispersion model was then used to

calculate the anticipated DOC concentration in the var-

ious boxes based on the concentrations of the preceding

month and modeled advection and dispersion. The dif-ferences between the actual concentration and the pre-

dicted represents the gain or loss of the material

POC vs TSS

5

6

122 P. Ford et al. / Marine Pollution Bulletin 51 (2005) 119–127

attributable to all causes such as sedimentation and

resuspension, microbial metabolism including sediment

diagenesis, and uptake by phytoplankton.

0

1

2

3

4

0 200 400 600 800 1000 1200TSS (mg l-1)

%C

Fig. 4. Variation in organic carbon content (%) of suspended solids as

function of suspended solids concentration (mgl�1) for 1998 event.

3. Results and discussion

3.1. Discharge and particulate organic carbon

concentration and sources

In the Fitzroy River, the concentration of suspended

solids (TSS) is highly dependent on the discharge (Fig.

3) rising from pre-flood levels of �20mgl�1 to more

than 1000mgl�1 at maximum discharge, and then slowlydeclines. The discharge increases by 4 orders of magni-

tude, and the suspended sediment concentration by 2 or-

ders of magnitude. At these TSS levels the river is highly

turbid (1500 NTU) and autochthonous production is

negligible during the flood flow. Thus the carbon con-

tent and isotopic signature of the sediment reflects the

characteristics of the sediment sources and the changes

(if any) which the material has undergone in passagefrom the catchment to the estuary. Suspended sediment

concentrations remain elevated (relative to pre-flood

values) well after the discharge has declined. This is seen

in the 1998 observations (Fig. 3), where one month after

the passage of the initial small flood peak (37Mld�1) the

TSS has only decreased from 360mgl�1 to 200mgl�1. A

further month later it sinks to 170mgl�1 when the major

flood flow arrives. This is due to the slow settling of thevery fine particles delivered by the flood and leads to

hysteretic effects between TSS and discharge which pre-

clude the construction of realistic empirically based dis-

charge-load relationships. For instance, in the major

event in April–May, 1998 the TSS concentration be-

tween the passage of the two major flood peaks (Fig.

3), is twice the maximum observed during the initial

maximum discharge, despite the discharge then beingabout 1/3 that in February.

1998 Flood TSS and

0

20000

40000

60000

80000

100000

120000

140000

26-Dec-97 23-Jan-98 20-Feb-98 20-Mardate

Dis

char

ge (M

l d-1

)

Fig. 3. Discharge (continuous line) and suspended sediment

At high TSS concentrations (Fig. 4) the organic car-

bon content in the 1998 event is remarkably uniform(1.54% (±0.2%; n = 23)). Similarly, the POC concentra-

tion inferred from measurement of TOC and DOC,

and TSS in the barrage (Fig. 3), shortly after the passage

of the flood peaks (S < 1&) gave an average organic

carbon content of the suspended solids of 1.87%. As

TSS declines below 200mgl�1 the carbon content in-

creases reaching an approximate asymptote of 5.5% C

at nominal 0mgl�1 TSS.The d13C signature of the suspended sediments shows

(Fig. 5) an analogous constancy at TSS concentrations

above 200mgl�1, staying in the range of �21 to

�22& (average 21.7 ± 0.8&; n = 23). While at lower

TSS concentrations, d13C becomes appreciably lighter

and asymptotic to approximately �27& at nominal

0mgl�1 TSS. Table 1 summarises these observations to-

gether with similar data for samples collected intermit-tently during flood events from 2000 to 2003. Samples

were collected approximately 500km upstream when

the water left the sub-catchment, and again when the

same water mass reached the barrage downstream.

Table 1 also shows the d13C and d15N values for para

grass (Poaceae: Urochloa mutica (Forsk) Stapf.), an

Discharge

-98 17-Apr-98 15-May-98 12-Jun-980

200

400

600

800

1000

1200

TSS

( mg

l-1)

concentration (filled squares) for flood event in 1998.

Table 1

Summary of characteristics of suspended sediments in Fitzroy River

during floods at concentrations greater than, and less than, 200mgl�1

and soil parameters

Suspended sediment property TSS > 200mgl�1 TSS < 200mgl�1

1998 Barrage suspended

sediment d13C

�21.7 ± 0.8& Decreasing towards

�27& as TSS

declines

1998 Barrage suspended

sediment d15N

5.1 ± 1.5& 3.0 ± 2.0&

% Organic C 1.57 ± 0.45% Increasing

towards 5.5%

C/N (atomic) 14.8 ± 1.3 10.8

d13C-upstream suspended

sediment 2001–2003

�22.5 ± 1.1& N/A

d13C-downstream suspended

sediment 2001–2003

�21.3 ± 1.3& N/A

Surface soils cropped areas

Comet catchment

�15.9& N/A

d13C Para grass (leaves) �12.4& N/A

(C/N = 40.2)

d13C Para grass (roots) �13.5& N/A

(C/N = 20.8)

d15N Para grass (leaves) 9.4 N/A

d15N Para grass (roots) 5.5 N/A

y = 1.064Ln(x) - 28.434R2 = 0.6033

-28

-26

-24

-22

-20

0 200 400 600 800 1000 1200

TSS ( mg l-1 )

del 13

C (0

/00 )

Fig. 5. Variation in carbon isotopic signature (d13C) of suspended

solids as function of suspended solids concentration (mgl�1) for 1998

event.

P. Ford et al. / Marine Pollution Bulletin 51 (2005) 119–127 123

introduced pasture plant which grows profusely in theriparian zone of the lower reaches of the Fitzroy and

is dislodged as floating rafts of vegetation during floods.

No mixture of para grass and sediments can account for

the observed d13C signal of the organic carbon at low

TSS.

We suggest that the organic carbon delivered to the

estuary at high suspended sediment concentrations is

soil organic carbon derived from a savanna environmentwith a mixture of C3 trees and C4 grasses. The observed

d13C (�21.7&) is close to the value calculated from %C

using the relationship of Bird and Pousai (1997). Fur-

thermore, the observed d13C is in the middle of the range

(�18 to �24&) found for fine particles collected during

a flood event in arid temperate Australia (Olley, 2002)

and for both the top soils and subsoils in the sediment

source region. It also matches the d13C value appropri-

ate for a wet (treed) or a dry (grassed) savanna predicted

from the particle size, based on an extensive survey of

tropical and subtropical biomes in northern Australia.

Further support for our identification of the suspendedsediment carbon as soil organic carbon comes from

the work of Masiello and Druffel (2001) on the Santa

Clara River—a highly episodic stream which, like the

Fitzroy, exports sediment in a few high flux events. They

showed that the exported sediment had an organic car-

bon d13C signature of �22.2 ± 0.8& and that the trans-

ported sediment was derived from old and deeply eroded

soils. The C/N was greater than 10 in both catchments(Fitzroy C/N: 14.8 ± 1.3; Santa Clara: 11.2 ± 3.3). C/N �10 is the global average (Meybeck, 1982) and values

greater than this are often indicative of soil organic

matter.

In contrast, the material, transported at low TSS con-

centrations at the beginning of the flow, has a much

lower d13C (�26&), lower C/N (10.8), and a much higher

%C, than the material at the flood peak. While the d13Cis close to the value characteristic of C3 trees and could

arise from vegetation growing immediately adjacent to

the river and mobilized in the first stages of the flood

(Bird et al., 1998), the low C/N precludes such an expla-

nation. The d13C and C/N sugget that the material trans-

ported in the initial stages of the flow is FSPOM

produced in situ by primary producers, including bio-

films (Burns and Walker, 2000), using terrestrial materi-als (Olley, 2002).

As has been noted previously (Onstad et al., 2000;

Masiello and Druffel, 2001) terrestrial material with a

d13C of �22& can be confused with marine detritus

which has a similar signature. This is a potentially con-

founding issue in the Fitzroy as we attempt to develop

more accurate estimates of changes in sediment delivery

due to human-induced land-use changes. If we assumethat the greatest impact of altered land use has been to

change the delivery of soil and associated soil carbon

disproportionately relative to in-stream debris, then we

can make three inferences from these results which

may help in unravelling the sedimentary history from

near-shore cores. Moving down core, the following ef-

fects should be seen crossing the boundary reflecting

pre- and post-European land-uses. They are:

1. d13C becomes more negative.

2. The C/N decreases.

3. The 14C age becomes younger (a counter intuitive

result!).

3.2. Dissolved organic carbon

Dissolved organic carbon (DOC) was measured in

samples collected at the regular sampling stations along

Table 2

DOC concentrations, flow volumes, and DOC loads (Fitzroy estuary

2000–2003)

Date of flow DOC concentration

(mgCl�1)

Flow

volume (Mm3)

DOC load

(Tonnes C)

November 2000 9.78 377 3687

December 2000 7.11 1616 11,490

January 2001 6.09 365 2223

February 2001 7.48 593 4435

April 2001* 4.27 115 491

February 2002* 2.88 101 291

March 2002* 4.87 29 141

Jun 2002* 3.71 95 352

February 2003 7.9 1809 14,291

March 2003 7.83 901 7054

* Total discharge less than two barrage volumes.

124 P. Ford et al. / Marine Pollution Bulletin 51 (2005) 119–127

the estuary, within 3weeks of the start of a flood event.

The data, when plotted against salinity (Mantoura and

Woodward, 1983) shows straight line behaviour with a

high correlation coefficient (R2 = 0.96). Table 2 brings

together the results from a total of 10 flood events ob-

served over 3years. This includes four events wherethe total discharge (integrated over the flow period)

was comparable to the volume of the barrage

(60Mm3). These small flows are not sufficiently large

to completely fill the estuary with water from the

upstream region, merely displacing the barrage contents

into the estuary. The larger flows (>2 barrage volumes)

have higher DOC concentrations than the smaller flows

(7.7mgl�1 and 3.7mgl�1 respectively). The linear rela-tionship between DOC and salinity is observed in many

estuaries (Laane and Koole, 1982; Mantoura and

Woodward, 1983; Berger et al., 1984) and is taken to

indicate that the riverine DOC is refractory on the time

scale of its passage through the estuary. A more recent

comparative study (Hopkinson et al., 1998) indicates

that estuarine bacteria could consume the DOC, but

the bacterial growth rate was low and of the order of1.5lgmg�1 d�1 DOC. Thus under short residence times

any DOC metabolism would not be detected. In the

Fitzroy, the waters are highly turbid post-flood and pho-

tochemical activation of DOC (Kieber et al., 1989)

Fig. 6. Fitzroy estuary showing sampling sites (filled circles

which can increase DOC removal rates several-fold

(Miller and Moran, 1997) is prevented. The budget

method, discussed later, quantifies the DOC behaviour

in the low-flow period when the water column clears

and photochemical effects are potentially possible.

The total load of DOC delivered by the river into theestuary in the 20months from April 2001 to January

2003 is 1275tonnes. This is very small compared to the

preceding 4months (November 2000 to February

2001) when 21,835 tonnes were delivered, or the subse-

quent episode covering only 2months (February to

March 2003) when a further 21,245 tonnes came down-

stream. These results exemplify the highly episodic char-

acter of tropical rivers, such as the Fitzroy, draining drysavannas, and the intermittent delivery of large loads of

dissolved nutrients and organic carbon, both dissolved

and particle-attached.

For comparison, we have calculated also the annual

DOC load from the City of Rockhampton�s three waste

water treatment plants using the weighted mean DOC

concentration and the annual discharge of the treated

effluent. This amounts to �50tonnes and is dwarfedby the inputs of even the smallest flow (indicated by *

in Table 2).

3.3. DOC and POC budgets by the inverse method

We applied the Inverse method of Webster et al.

(2004) to develop budgets of DOC and POC, and to esti-

mate fluxes of DOC and POC in the Fitzroy estuary.The estuary was divided into six sections (Fig. 6) with

each section containing at least one sampling station.

Budgets for DOC and TOC (the two carbon species ana-

lysed for) were constructed, and then the POC budget

was calculated by difference from the TOC and DOC.

Because of the highly episodic flows separate budgets

were calculated for the wet season (1 November 2000–

30 April 2001) and the dry season (1 May 2001–30November 2001). The total discharges were 2951Mm3

and 115Mm3 respectively—a factor of 25 different.

For ease of presentation the budgets of individual cells

have been consolidated into ‘‘up-stream’’ (cells 1–3)

) in relation to the sectors used for flux calculations.

P. Ford et al. / Marine Pollution Bulletin 51 (2005) 119–127 125

and ‘‘down-stream’’ (cells 4–6) units (Fig. 7a and b).

Each unit covers approximately the same length of the

estuary. The downstream unit includes the large ‘‘cut-

through’’ loop (Fig. 6), and ends at the station 2.5km

from the actual estuary mouth.

The wet season delivery of DOC calculated by thismethod (21,136tonnes) is in good agreement with the

same parameter inferred from the riverine DOC concen-

tration and the total discharge (21,835tonnes). The con-

cordance between the results of the two methods is not

as good for the dry season delivery (Inverse method:

49tonnes; concentration · discharge: 491tonnes). The

disparity probably arises from two factors. The small

flow delivers only a relatively small load of DOC whichproduces only small concentration differences at all sta-

tions pre- and post-flood. The budget depends on these

differences and thus has a proportionally larger error.

The assumption of a constant DOC concentration

(a)

(b)

Fig. 7. Annual budget (tonnes) for (a) DOC, (b) POC under high flow

(wet season) and low flow (dry season) conditions in upstream and

downstream sectors (Fig. 6). Arrows indicate direction of net flux.

through out the flood may be poorly founded. The re-

sults above show that there is discharge-dependent dif-

ferences in DOC concentration, with the smaller flows

having lower DOC concentrations. The initiation and

concluding phases of these smaller flows would have

even lower concentrations also and thus taking the max-imum concentration will overestimate the load.

Under wet season (high flow) conditions there is a

slight loss of DOC (�3%), but the bulk of the incoming

material passes straight through the estuary. The inverse

method deals with net changes in DOC, so this loss in-

cludes adsorption of DOC by particles, diffusive ex-

change between the sediments and the water column,

as well as microbial metabolism. The small loss explainsthe apparent contradiction between the view, discussed

in Section 1, of riverine DOC being refractory because

of its conservative character demonstrated by straight

line behavior in DOC vs salinity plots; and the in vitro

measurements showing measurable consumption of

DOC in estuarine waters. Our results show that the

amount of DOC removed is too small to produce a

noticeable deflection in the DOC vs salinity plots.In the dry season (low flow) conditions the Fitzroy

estuary is a net source of DOC from both the upstream

and downstream sectors. The total flux of 587tonnes of

DOC into the estuary waters is very nearly balanced by

the wet season loss of 679tonnes. Examination of the

fluxes (Fig. 8) shows that different parts of the estuary

have quite different characteristics for DOC production

and removal. The most downstream station (cell 6) is asource of DOC under both high and low flow conditions

(�20mmolm�2 d�1). Under the low flow conditions the

contribution from this cell is sufficient to offset the losses

of DOC (40mmolm�2 d�1) occurring in cells 4 and 5

immediately upstream, and make the contribution for

this sector positive i.e. flux into the water column. In

contrast, upstream cells 2 and 3 are net contributors of

DOC at approximately 8mmolm�2 d�1 under dry sea-son conditions. While under wet season conditions these

cells are sinks for DOC reflecting losses by microbial

metabolism and agglomeration to form particles.

DOC Fluxes

-50

-40

-30

-20

-10

0

10

20

30

0 10 20 30 40 50 60

Distance downstream from barrage (km)

Flux

( m

mol

m-2

d-1)

Wet SeasonDry Season

Fig. 8. Fluxes of DOC (mmolm�2d�1) measured along the estuary

under high flow (filled squares) and low flow (filled circles).

126 P. Ford et al. / Marine Pollution Bulletin 51 (2005) 119–127

Under both wet and dry season conditions our bud-

get calculations show (Fig. 7b) that the Fitzroy is a

major source of POC to the coastal seas. As would be

anticipated, the contribution is much greater in the

wet season with the amount of POC mobilized within

the estuary being comparable to the amount totalamount delivered by the flood. The downstream sector

is the dominant source of POC in the wet season (down-

stream:upstream 16:1). This dominance extends to the

dry season though the disparity is reduced (down-

stream:upstream 4:1). As the POC is associated with

the fine sediments the POC budget indicates that the

wet season flow moves a significant amount of fine sed-

iment out of the estuary. If we take the POC content ofthe incoming riverine suspended solids (1.57% see Table

1) as typical of the POC content of the mobilised sedi-

ments, then the estuary contributes an additional

151,000tonnes of sediment in the wet season and a fur-

ther 34,000 tonnes in the dry season over and above the

amount that is carried through the estuary by the flood

flow. This may come from bank erosion during the high

flow periods (suggestion of anonymous referee). Theestuary as a whole, is however prograding (R. Packett,

pers. commun.) and there must be a long term source

of this additional material. We suggest that it is material

deposited in previous floods in the near shore coastal

area. Our tidal data used in the model calibration shows

that tides in the Fitzroy estuary are quite asymmetric,

with the ‘‘ins’’ being of shorter duration, and conse-

quently of higher velocity than the ‘‘outs’’. Thus therewill be a net movement of sediment into the estuary

on each tidal cycle, but this is too small to be detected

by our monthly sampling.

4. Conclusion

In the Fitzroy POC is delivered in the episodic floodsand is primarily composed of soil organic carbon. It

makes up about 1.5% of the total suspended solids.

Under low flows the proportion of POC derived from

in-stream primary production rises, but the load of

POC decreases. Given that the sediment load has been

considerably elevated, relative to pre-European settle-

ment, the character of the POC entering Keppel Bay

has changed. Under pristine conditions, the vegetationcomponent would have been higher and less diluted with

relatively refractory SOC. The ecological consequences

are that land clearing has led to conditions in the receiv-

ing waters which are more favorable to organisms able

to metabolise relatively refractory organic matter. The

DOC concentrations in large flood events are remarka-

bly uniform (7mgl�1), but smaller (insufficient to flood

the estuary) events are approximately half the DOC con-centration of the larger events. Budgets for the estuary

constructed using the Inverse method show that an addi-

tional amount of POC, comparable to that transmitted

through the estuary, is removed from the estuary in

flood events. We hypothesize that that much of this

material is ‘‘pumped’’ back into the estuary by tidal

processes which are not fully captured by the budget

method. Under flood flows DOC is removed in the estu-ary, but in the dry season, there is a flux of DOC into the

water column from the estuary. The mouth of the estu-

ary, where macrotidal effects are largest is always a

source of DOC to the water column. The upstream por-

tion is a source of DOC under low flows and a sink

under high flows, while the lower portion of the estuary

is always a sink though it removes more DOC in flood

events than during the dry season.

Acknowledgments

We gratefully acknowledge the assistance of the

Queensland EPA, especially Andrew Moss and John

Ferris, in the collection of the samples, and Bob Noble

and Bob Packett of the CRC Coastal Zone, Estuary,and Waterway Management (CRCCZEWM) for logis-

tic support in Rockhampton and the collection of tidal

data used in calibrating the tidal model. Dr Lynda

Radke and Dr Jon Olley (and 2 anonymous referees)

provided penetrating and helpful comments on an ear-

lier draft. We thank the management and staff of Fitzroy

River Water Corporation for assistance with the flood

sampling program. This work was conducted underthe auspices of the CRCCZEWM and we gratefully

acknowledge their support.

References

Anthony, K.R.N., Fabricius, K.E., 2000. Shifting roles of heterotro-

phy and autotrophy in coral energy budgets at varying turbidity.

Journal of Experimental Marine Biology and Ecology 252, 221–

253.

Berger, P., Laane, R.W.P.M., Ilahude, M., Ewald, M., Courtot, P.,

1984. Comparative study of dissolved fluorescent matter in four

West-European estuaries. Oecologica Acta 7, 309–314.

Berner, R.A., 1989. Biogeochemical cycles of carbon and sulphur and

their effect on atmospheric oxygen over Phanerozoic time. Palaeo-

geography, Paleoclimatology and Palaeoecology 73, 97–122.

Bird, M.I., Pousai, P., 1997. Variations of d13C in the surface organic

carbon pool. Global Biogeochemical Cycles 11, 313–322.

Bird, M.I., Giresse, P., Ngos, S., 1998. A seasonal cycle in the carbon-

isotope composition in the Sanaga River, Cameroon. Limnology

and Oceanography 43, 143–146.

Bormans, M., Ford, P.W., Fabbro, L., Hancock, G., 2004. Onset and

persistence of cyanobacterial blooms in a large impounded tropical

river, Australia. Marine and Freshwater Research 55, 1–15.

Boulton, A.J., Findlay, S., Marmonier, P., Stanley, E.H., Valett, H.M.,

1998. The functional significance of the hyporheic zone in streams

and rivers. Annual Reviews of Ecology and Systematics 29, 59–82.

Burns, A., Walker, K.F., 2000. Biofilms as food for decapods (Aytidae,

Palaemonidae) in the River Murray, South Australia. Hydrobio-

logia 437, 83–90.

P. Ford et al. / Marine Pollution Bulletin 51 (2005) 119–127 127

Ertel, J.R., Hedges, J.I., Richey, J.E., Ribeiro, M.N.G., 1986.

Dissolved humic substances of the Amazon River system. Geochi-

mica et Cosmochimica Acta 13, 739–754.

Furnas, M., 2003. In: Catchments and corals: terrestrial runoff to the

Great Barrier Reef. AIMS, Townsville, p. 334.

Hopkinson, C.S., Buffam, I., Hobbie, J., Vallino, J., Perdue, M.,

Eversmeyer, B., Prahl, F., Covert, J., Hodson, R., Moran, M.A.,

Smith, E., Baross, J., Crump, B., Findlay, S., Foreman, K., 1998.

Terrestrial inputs of organic matter to coastal ecosystems: an

intercomparison of chemical characteristics and bioavailability.

Biogeochemistry 43, 211–234.

Karickhoff, S.W., Brown, D.S., Scott, T.A., 1979. Sorption of

hydrophobic organic pollutants on natural sediments. Water

Research 13, 241–248.

Kelly, J.N., Wong, W.T., 1996. Sediment transport in the Fitzroy

River during flood events. In: Rutherford, I., Walker, M. (Eds.),

Proceedings of the 1st National Conference of Stream Manage-

ment in Australia. IEA, Merrijig, pp. 19–21.

Kieber, D.J., McDaniel, J., Mopper, K., 1989. Photochemical source

of biological substrates in sea water: Implications for carbon

cycling. Nature 341, 637–639.

Laane, R.W.P.M., Koole, L., 1982. The relationships between

fluorescence and dissolved organic carbon in the Ems-Dollart

estuary and the Western Wadden Sea. Netherlands Journal of Sea

Research 15, 217–227.

Mantoura, R.F.C., Woodward, E.M.S., 1983. Conservative behaviour

of riverine dissolved organic matter in the severn estuary: chemical

and geochemical implications. Geochimica et Cosmochimica Acta

47, 1293–1309.

Masiello, C.A., Druffel, E.R.M., 2001. Carbon isotope geochemistry of

the Santa Clara River. Global Biogeochemical Cycles 15, 407–416.

Meybeck, M., 1982. Carbon, nitrogen, and phosphorus transport by

world rivers. American Journal of Science 282, 401–450.

McCulloch, M., Fallon, S., Wyndham, T., Hendy, E., Lough, J.,

Barnes, D., 2003. Coral record of increased sediment flux to the

inner Great Barrier Reef since European settlement. Nature 421,

727–730.

Miller, W.L., Moran, M.A., 1997. Interaction of photochemical and

microbial processes in the degradation of refractory dissolved

organic matter from a coastal marine environment. Limnology and

Oceanography 42, 1317–1324.

Millman, J.D., Summerhayes, C.P., Barreto, H.T., 1975. Oceanogra-

phy and suspended matter off the Amazon River February–March

1973. Journal of Sedimentary Petrology 60, 456–470.

Moran, M.A., Sheldon Jr., W.M., Sheldon, J.E., 1999. Biodegradation

of riverine dissolved organic carbon in five estuaries of the south

eastern United States. Estuaries 22, 55–64.

O�Connell, M., Baldwin, D.S., Robertson, A.I., Rees, G., 2000.

Release and bioavailability of dissolved organic matter from flood

plain litter: influence of oxygen levels. Freshwater Biology 45, 333–

342.

Olley, J.M., 2002. Organic carbon supply to a large lowland river and

implications for aquatic ecosystems. The structure, function and

management implications of fluvial sedimentary systems (Proceed-

ings of an International Symposium, Alice Springs, September

2002). IAHS Publication 276, 27–33.

Onstad, G.D., Canfield, D.E., Quay, P.D., Hedges, J.I., 2000. Sources

of particulate organic matter in rivers of the continental USA:

Lignin phenol and stable carbon isotope compositions. Geochimica

et Cosmochimica Acta 64, 3539–3546.

Volk, C.J., Volk, C.B., Kaplan, L.A., 1997. Chemical composition of

biodegradable dissolved organic matter in stream water. Limnol-

ogy and Oceanography 42, 39–44.

Voice, T.C., Webber Jr., W.J., 1983. Sorption of hydrophobic

compounds by sediments, soils, and suspended solids. I, Theory

and background. Water Research 17, 1433–1441.

Webster, I.T., Ford, P.W., Tillman, P., 2004. Estimating nutrient

budgets in tropical estuaries subject to episodic flows. In: Hutch-

ings, P.A., Haynes, D. (Eds.), Proceedings of Catchment to Reef:

Water Quality Issues in the Great Barrier Reef Conference. Marine

Pollution Bulletin, doi: 10.1016/j.marpolbul.2004.10.027.

Zweifel, U.L., Wikner, J., Hagstrom, A., Lundberg, E., Norrman, B.,

1995. Dynamics of dissolved organic carbon in a coastal ecosystem.

Limnology and Oceanography 40, 299–305.