development of indicators for assessing and …development of indicators for assessing and...

Development of Indicators for Assessing and

Monitoring Nutrient Influences in Coastal Waters

Simon D. Costanzo

Development of Indicators for Assessing and Monitoring Nutrient Influences in Coastal Waters

A Thesis submitted by

Simon D. Costanzo BSc. (Hons.)

The University of Queensland, Australia

to the

Department of Botany

The University of Queensland

AUSTRALIA

in fulfilment of the requirements for

the degree of Doctor of Philosophy within


August 2001

V

STATEMENT

The work presented in this thesis is, to the best of my knowledge and belief, original, except as acknowledged in the text, and the material has not been submitted, either in whole or in part, for a degree at this or any

other University.

Signed................................................

Acknowledgements What an amazing journey the last four years has been for me. It has encompassed

some of the best and not so best times of my life, though I’ll only remember the good

times because they were aplenty. The key to me getting to this stage was the

tremendous help and support that I have received from so many people throughout

that time. Fortunately, at the beginning of all this I met Kath, who has been by my

side day and night throughout the entirety of the PhD. Kath without your love,

support, comfort, cheering up ability, eternal optimism, great smile and lack of

geographical direction, this would have been a much more cobbled road to travel.

Your qualities will surely shape the world. Thankyou so much! Bill Dennison, you

taught me to talk, think and write. These are important features for a young man in

the real world and I have you to thank for finding these qualities within me. Thanks

for providing the opportunities that have helped shape me into a good scientist.

Anthony White is another amazing person I owe a debt of gratitude. I’ve never spent

so much time mapping sewage with anyone else. You could have volunteered for

coral reef research, but no you continuously were at my side singing ‘70’s songs in

prime-time sewage territory. I’ll never forget those times at sea mapping sewage.

Ben Longstaff, you still owe me for helping you tag seagrass 6 years ago! Perhaps I’ll

let that debt go, but I will not let that happen to our friendship that has grown over the

last 6 years. We’ve been through a lot together over those years and have both come

out the better for it. I’m proud of what you have achieved and wish you, Andy and

Junior all the best for the future. Tim Carruthers, what a scientist and what a pool

partner. Your help over the last stage of this PhD has been invaluable and I can offer

no greater gratitude for your time and energy spent on listening to me blabber on

about my ideas and thoughts....cheers buddy. Ian Hewson, it’s bioassay time! You

may be 12, 000 km away but I seem to have been talking with you more than anyone

else in Australia. I’m sure you share the great times that we have had in the many

unusual, yet groovy places that we have done fieldwork. You also have been a great

help with the crux of this PhD and I owe you an expensive beer when I get to Seattle.

Andrew Watkinson….partner in crime. Probably not crime, but the next best thing.

You made me laugh a lot and realise what being a bloke is all about. Thanks for your

constant good mood and help in the field. Glad I could teach you something about

fishing. Mark O’Donohue….the man with the plan. I guess it is where it all started

VIII Acknowledgements

for me and the big blue sea and the big bad respirometer! Those days seem like a

dream….a Californian dream? It would take a whole page for me to thank you for the

all the help and guidance you have given me from the beginning….but that would

stuff up my formatting, so just thankyou for your enthusiasm and the good times spent

in awesome places. John Bertram, my fairy god-father. Thanks for keeping an eye

on me since day one and making things happen. John Gehrke, the man on a mission.

Who would have thought someone would be so happy to see a section of the research

station torn down by a boat trailer. Thankyou for your belief in me and my abilities

since that fateful day and hope that we always keep in touch. I can now reverse a

trailer with my eyes shut! Paul Bird, thanks for your effo rts in the field and those late

night cups of tea that kept us writing. The list continues, including Alan Goldizen,

Neil Loneragan, Michele Burford, Adrian Jones, Eva Abal, Tracey Saxby, the

Roelfsema Dutchies, Norm Duke, Joelle Prange, Caroline Gaus, Liz Duffy, Cindy

Heil, Catherine Collier, Dieter Tracey, Courtney Henderson, Gordon Moss, Dan

Wruck, Theresa Leijten, Judy O’Neil, Ros Murrell, Tom Toranto, Peter Toscas,

Mervyn Thomas…I know you’ve assisted me somewhere along the line and that help

I have not forgotten. Like they say at the Grammy Awards, I’ve probably forgotten

someone, but if you ever were a Marine Botanist (MARBOT), you probably helped

me out, so thanks. My family has been the cornerstone of my existence, literally. I

know I’ve been a busy bee lately and not been around as much as either you or I

would like, but knowing I have a loving family to always turn to is something that is

irreplaceable. Thanks Mum for always being willing to listen, dad for striving for my

success, Adam for looking up to me, Nonna for making the best gnocchi ever, Nanny

for knowing life is there to grab and Nonno for nudging me to do this PhD so that i)

I’d be the first Dr. in the family and ii) my kids would get into good schools! I hope

I’ve done you all proud and continue to do so.

Cheers Everybody! Simon This thesis research was initially funded by the Brisbane River & Moreton Bay Wastewater Management Study. Additional funding was provided through grants from the Tweed Shire Council and CRC for Aquaculture. Scholarship and professional education was provided by the Marine Botany Group and the CRC for Coastal Zone, Estuary and Waterway Management.

Development of Indicators for Assessing and

Monitoring Nutrient Influences in Coastal Waters

Abstract

With increasing human pressures on coastal ecosystems, there is a need to develop

better approaches to assess and monitor anthropogenic influences in these systems.

The aims of this thesis were to a) develop indicators that describe and predict nutrient

input effects, b) synthesise and interpret these indicators in assessment programs, and

c) provide cost-effective methods for use in regular monitoring programs. Three

study regions in tropical/sub-tropical coastal environments (Queensland, Australia)

were chosen to represent a variety of nutrient inputs: aquaculture (land-based shrimp

farming), sewage effluent (urban centre) and agriculture (sugar cane farming). Field

sampling programs were conducted to assess the spatial and temporal variabilities

associated with these nutrient inputs (i.e. discharge periodicity, run-off events and

seasonality).

In addition to standard water quality parameters, a variety of novel biological

indicators were developed and implemented to characterise zones of biotic responses

to the various nutrient inputs. Passive biological indicators employed in this study

included nitrogen stable isotope ratios (δ15N) of natural marine flora communities

(macroalgae, seagrasses and mangroves), which were used to infer nitrogen source

and fate. In regions devoid of macrophytes, an active biological indicator was

employed using nutrient deplete macroalgae incubated in the water column, enabling

intensive spatial resolution. A phytoplankton response index was developed in order

to ascertain a measure of the susceptibility of a system to nutrient additions. This

technique was based on responses of 7-d phytoplankton bioassays to increased light

and nutrient availability.

δ15N signatures of marine flora proved successful in distinguishing nitrogen sources in

each study region, often revealing trends not discernible using physical/chemical

parameters. δ15N signatures were characteristic of the nitrogen source – sewage

effluent (~10 ‰); shrimp effluent (~6 ‰); and cane farming (2-4 ‰). Incubated

X Abstract

macroalgae rapidly changed their δ15N signature within 4-d when exposed to 15N

enriched nitrogen sources (e.g. sewage). Analysis of a variety of marine flora

provided a more comprehensive understanding of nitrogen sources and fates.

Mangroves integrated longer temporal variations and were indicative of sediment

processes, whereas incubated macroalgae reflected short term conditions in the water

column. Phytoplankton bioassays highlighted regions susceptible to nutrient

enrichment, not predictable from water quality parameters alone.

The nature of the different anthropogenic nutrient inputs strongly influenced water

quality and biotic responses in receiving waters. Variable release rates of shrimp

pond effluent and runoff variability from agriculture were detected with the methods

described. The continuous input of sewage effluent resulted in large detectable

plumes, which were influenced by environmental factors such as temperature and

water currents.

Statistical and spatial analyses were used to enhance data interpretation and

presentation, which facilitated communication with resource managers and the wider

community. Communication of these results led to successful integration and

application into environmental management and monitoring programs.

Contents

Development of Indicators for Assessing and Monitoring Nutrient Influences in Coastal Waters

Abstract IX

Chapter 1 21

Introduction 21

1.1 Assessment of Australian Coastal waters 21

1.2 Nitrogen and Coastal Eutrophication 21

1.2.1 Urban Sewage 24

1.2.2 Crop Agriculture 25

1.2.3 Shrimp Aquaculture 25

1.3 Understanding the Problem 26

1.4 Thesis Outline 27

Chapter 2 29

A new approach for detecting and mapping sewage nitrogen 29

Abstract 29

2.1 Introduction 30

2.1Nitrogen Stable Isotopes 30

2.2 Sewage Nitrogen and Marine Plants 32

2.3 Study Region 33

2.4 δ15N of ambient marine plants 33

2.5 δ15N of deployed marine plants 35

2.6 Spatial Analysis 36

2.7 Statistical verification 40

2.8 Applications 41

2.9 Summary 42

12 Contents

Chapter 3 43

Nutrient Bioassays: a tool for determining phytoplankton bloom potential in estuarine and coastal waters 43

Abstract 43

3.1Introduction 44

3.2 Materials & Methods 46

3.2.1 Site Location 46

3.2.2 Sampling Strategy and Analysis 47 3.2.2.1Spatial Comparison 47 3.2.2.2 Temporal Comparison 47

3.2.3 Physical/Chemical parameters 48

3.2.4 Phytoplankton Bioassays 49 3.2.4.1 Phytoplankton Response Index 51

3.2.5 Data Analyses 52

3.3 Results 53

3.3.1 Physical / Chemical Properties 53

3.3.2 Phytoplankton Bioassays 54 3.3.2.1 Transect Data 54 3.3.2.2 Monthly Data 56

3.3.3 Bioassay Responses and Physical/Chemical variables 57 3.3.3.1 Transects 57 3.3.3.2 Monthly 57

3.3.4 Bioassay Predictions 58

3.4 Discussion 59

3.4.1 Physical/Chemical Water Quality 60

3.4.2 Phytoplankton Bioassays 60 3.4.2.1 Nutrient Responses 61 3.4.2.2 Light Responses 63

3.4.3 Bioassay Response Predictability 63

3.5 Conclusion 65

Chapter 4 67

Assessing the influence and distribution of shrimp pond effluent in a tidal mangrove creek 67

Abstract 67

4.1 Introduction 68

4.2 Materials and Methods 70


4.2.2 Sampling regime 70

Contents 13

4.2.3 Physical/chemical parameters 72

4.2.4 Biological Indicators 73 4.2.4.1Nitrogen Stable Isotope Mapping 73 4.2.4.2 Phytoplankton Bioassays 73

4.2.5 Statistical Analyses 75

4.3 Results 75

4.3.1 Physical / Chemical Properties of Receiving Waters 76

4.3.2 Biological Indicators 78 4.3.2.1 Vegetation δ15N Mapping 78 4.3.2.1 Phytoplankton Bioassays 81

4.4 Discussion 84

4.4.1 Water Column Physical/Chemical Parameters 84

4.4.2 Biological Indicators 85 4.4.2.1 Vegetation δ15N Mapping 85 4.4.2.2 Phytoplankton Bioassays 87

4.5 Conclusions 88

Chapter 5 91

Assessing the Seasonal Influence of Sewage and Agricultural Nutrient Inputs in a Sub-Tropical River-Estuary 91

Abstract 91

5.1 Introduction 92

5.2 Materials & Methods 94


5.2.2 Sampling Strategy 96

5.2.3 Physical/chemical parameters 97

5.2.4 Biological Indicators 98 5.2.4.1 Nitrogen Stable Isotope and %N Mapping 98 5.2.4.2 Phytoplankton Bioassays 99

5.2.5 Data Analyses 100

5.3 Results 101

5.3.1 Physical / Chemical Indicators of Water Quality 101

5.3.2 Biological Indicators of Water Quality 104 5.3.2.1 Macroalgae and Mangrove Indicators 104 5.3.2.2 Phytoplankton Indicators 107

5.3.3 Analysis of Bioindicator Data 108

5.4 Discussion 110

5.5 Conclusion 115

14 Contents

Chapter 6 117

Discussion 117

6.1 Conceptual Overview of Thesis Findings 117

6.1.1 Moreton Bay – sewage effluent dominated 118

6.1.2 Cardwell – Shrimp Effluent 120

6.1.3 Tweed River – Agriculture Dominated 122

6.2 Applicability of Physical/Chemical and Biological Indicators for Assessing Nutrients in Estuarine and Coastal Waters 124

6.3 Application of Findings 128

6.5 Conclusion 130

References 131

Appendix I 149

Appendix II 167

List of Tables

Table 3.1 Mean (±? SE) water column physical/chemical measurements between transects, surveys and regions. 54

Table 3.2 Significance values for nutrient and light PRI comparisons between transects

(Caboolture, Pine, Brisbane and Logan), surveys (September 1997 and February

1998), regions (river/bay) and interactions between these factors. 56

Table 3.3 Significant variables that explain variability experienced in nutrient and light PRI’s in

the transect and monthly data sets. 58

Table 4.1 Mean water column physical/chemical parameters of the influent and effluent creeks

associated with the shrimp farm. Data represent surveys conducted over three stages

of shrimp -pond maturity – empty, full and harvest. (TSS – Total Suspended Solids;

FRP – Filterable Reactive Phosphorus; TN/TP – Total Nitrogen / Total Phosphorus).

76

Table 4.2 Significance values for comparisons between surveys (empty, full and harvest),

locations (influent and effluent creeks) and interactions between surveys and

locations for δ15N of mangroves and macroalgae. 81

Table 4.3 Mean Phytoplankton Response Indices (PRI) following nutrient additions for the

influent and effluent creeks associated with a shrimp farm. Sampling was conducted

during 3 stages of shrimp -pond maturity– empty, full and harvest.

82

List of Figures

Figure 1.1 Conceptual diagram of the nitrogen cycle indicating dominant nitrogen inputs and

processes in atmospheric, terrestrial and aquatic environments. 23

Figure 2.1 Location and nitrogen loading (tonnes/yr) of sewage effluent outfalls into Moreton

Bay and tidal estuaries. 32

Figure 2.2 δ15N values of naturally occurring marine plants (mangroves, seagrass and

macroalgae) throughout Moreton Bay and tidal estuaries. 34

Figure 2.3 Macroalgal deployment chambers with embedded image of macroalgae incubating in

situ 36

Figure 2.4 Locations of deployed macroalgae in Moreton Bay (solid red circles). 37

Figure 2.5 Spatial distribution of deployed macroalgal δ15N values in September 1997 (a) and

February 1998 (b). Macroalgae (Catenella nipae, Rhodophyte) was deployed at ~100

sites (yellow solid circles) in Moreton Bay, Australia. 39

Figure 2.6 Co-efficient of variation map of interpolated δ15N values for February. 41

Figure 3.1 Study location showing the four major rivers entering Moreton Bay and the study

transects with sampling sites along each (solid circles) and monthly sampling sites

(open squares). 48

Figure 3.2 Graphic description of phytoplankton bioassay responses over 7days to added

nutrients and no added nutrients. FO – initial fluorescence; FN – maximum nutrient

fluorescence; FC – maximum light fluorescence; FCN – control fluorescence at

maximum nutrient fluorescence; TN – time to maximum nutrient fluorescence; TC –

time to maximum light fluorescence. 52

Figure 3.3. Comparison of mean nutrient and light phytoplankton response indices (PRI) (± SE)

between a) transects, b) surveys, and c) regions (river or bay) 55

Figure 3.4 Mean monthly nutrient and light bioassay responses expressed as phytoplankton

response indices (PRI) and mean monthly water column nutrients. 57

List of Figures 17

Figure 3.5 Correlations between monthly predicted and monthly observed a) nutrient bioassays

and b) light bioassays. 59

Figure 4.1 Site map showing locations of sites within the influent and effluent creeks and sites

located in Hinchinbrook Channel. 71

Figure 4.2 Dissolved (a) and total nitrogen (b) concentrations along the influent creek (sites 1-3)

and the effluent creek (sites 4-7) for each ponds stage – empty, full and harvest. 78

Figure 4.3 Mean δ15N values for macroalgae sampled in the influent and effluent creeks and

Hinchinbrook Channel for each: a) survey, and b) site. Bars indicate standard error.

79

Figure 4.4 Mean δ15N values for macroalgae sampled in the influent and effluent creeks and

Hinchinbrook Channel for each: a) survey, and b) site. Bars indicate standard error.

80

Figure 4.5 Phytoplankton bioassay data along the influent creek (sites 1-3) and the effluent creek

(sites 4-7) for each pond stage – empty, full and harvest. a) represents ambient

fluorescence values of creek water at the time of bioassay collection (proxy for

chlorophyll a concentrations); b) represents nutreint (N+P) PRI’s along each creek for

each pond stage; and c) represents light PRI’s along each creek for each pond stage.

83

Figure 5.1 Average Tweed Heads and Murwillumbah daily rainfall data from June 1997 –

January 1999 and from January 2000 – March 2001. 96

Figure 5.2 Map of Tweed River system with sampling sites (solid circles) and distance up-river,

sewage effluent outfalls (open squares) and major land uses (urban/agricultural). 97

Figure 5.3 Physical/chemical variables measured along the Tweed River between wet and dry

seasons. Parameters include a) temperature, b) salinity, c) total suspended solids

(TSS), d) chlorophyll a, e) pH, f) ammonium (NH4+), g) nitrate (NO3

-), h) phosphate

(PO43-), i) total nitrogen (TN) and j) total phosphorus (TP). 103

Figure 5.4 Two dimensional non-metric MDS on physical/chemical data from all sites and all

surveys. Broken lines group data from wet and dry seasons. Significant principal

axis correlation vectors (with correlation coefficients in brackets) are shown, each

point represents one site at the indicated sampling time. The stress level of 0.064 was

acceptable (Clarke & Warwick 1994) 104

18 List of Figures

Figure 5.5 Mean macroalgae δ15N signatures and tissue nitrogen contents: a) along the Tweed

River and b) between wet and dry seasons. Error bars represent standard error.

Initial δ15N signatures and %N values of macroalgae prior to incubation are

represented and the transition between urban and agricultural regions is highlighted in

(a). STP represents location of sewage treatment plant outfall in the Tweed River.

105

Figure 5.6 Mean mangrove δ15N signatures and tissue nitrogen contents: a) along the Tweed

River and b) between wet and dry seasons. Error bars represent standard error.

Transition between urban and agricultural regions is highlighted in (a). STP

represents location of sewage treatment plant outfall in the Tweed River. 106

Figure 5.7 Correlations of (a) macroalgal tissue nitrogen content (%N) and (b) δ15N signatures

with dissolved inorganic nitrogen (DIN). 107

Figure 5.8 Mean phytoplankton light and nutrient responses (PRI) represented a) spatially and b)

temporally in the Tweed River. 108

Figure 5.9 Two dimensional non-metric MDS on bioindicator data from all sites and all surveys.

Broken lines group data from wet and dry seasons. Significant principal axis

correlation vectors (with correlation coefficients in brackets) are shown, each point

represents one site at the indicated sampling time. The stress level of 0.31 was above

the acceptable limit [Clarke, 1994 #289]. 109

Figure 6.1 Diagrammatic representation of findings in Moreton Bay in a) early autumn and b)

summer. 119

Figure 6.2 Diagrammatic representation of findings in Cardwell shrimp farm with a) full and b)

empty ponds. 121

Figure 6.3 Diagrammatic representation of findings in the Tweed River during the a) wet and b)

dry seasons. 123

List of Plates

Plate 3.1 Satellite image of Moreton Bay and catchment. Four major rivers entering Moreton

bay are highlighted in blue. 46

Plate 3.2 Bioassay containers used for phytoplankton bioassays. Each container holds 4 L of

site water. The container on the right at the front was a nutrient treatment and the

subsequent phytoplankton response is evident by the green coloured water. 49

Plate 3.3. Outdoor incubation tank with bioassay containers inside. 51

Plate 4.1 Aerial photo of the studied shrimp farm with the influent creek highlighted in blue

and the effluent creek in red. The dotted lines show the continuation of the natural

creeks. 71

Plate 5.1 NOAA true colour satellite image of the Tweed River Estuary and Catchment,

surrounded by the Border ranges to the north and Mt Warning to the south.

Approximately 49% of catchment area is rural as evident in this image. Residential

development (pale white) can be seen towards the mouth and in the central region

(Murwillumbah). 95

Plate 5.2 Aerial images of a) agricultural dominated mid-upper region of the Tweed River

showing cane drainage channel joining the river; b) weir dividing river at ~40km

upstream from the mouth; and c) urban dominated lower region of the Tweed River

with stabilised river mouth. 95

CChhaapptteerr 11 Introduction

1.1 Assessment of Australian Coastal waters

The combination of oligotrophic coastal waters (Zann 1995), variable rainfall (Lake et

al 1985), and anthropogenic settlement along the coastline (Cullen 1982) makes

Australian coastal waters particularly susceptible to event driven nutrient fluctuations

(Eyre 1993). Traditional water quality monitoring strategies were largely developed

in North America and Europe where ambient nutrient and chlorophyll concentrations

are generally higher than in tropical regions (e.g. Funderbunk et al 1991). Australian

coastal waters differ considerably from northern hemisphere counterparts, in that

generally low terrigenous soil nutrients, coupled with rapid chemical and biological

uptake of available nutrients, result in low ambient nutrient concentrations and high

sensitivity to elevated nutrient enrichment (Norris & Norris 1995). Subsequently,

static sampling of physical/chemical parameters of Australian coastal waters has not

provided resource managers with powerful enough diagnostic tools to make informed

decisions. This has stimulated interest in biological indicators for environmental

assessment and monitoring (Hart et al draft).

1.2 Nitrogen and Coastal Eutrophication

Assessment and monitoring attention has largely been directed towards nitrogen

loading to coastal waters, as rates of primary production in many estuarine and marine

waters are principally limited by nitrogen supply (Downing 1997, Hecky & Kilham

1988, Howarth 1988, Lapointe & Clark 1992, Vitousek & Howarth 1991). There are

22 Introduction

three forms of nitrogen in the marine environment: i) dissolved inorganic nitrogen

(DIN), ii) dissolved organic nitrogen (DON), and iii) particulate organic nitrogen

(PON) (Lobban & Harrison 1994). The abundance and distribution of each of these

forms of N can vary seasonally (Sharp 1983), geographically (Fisher et al 1999) and

with water depth (Redfield et al 1963). DIN has been shown to be the primary

limiting factor in marine phytoplankton productivity (Ryther 1971). DIN occurs in

the marine environment in four significant forms, in order of abundance: gaseous

nitrogen (N2), nitrate (NO3-), ammonium (NH4

+) and nitrite (NO22-).

Despite the abundance of gaseous N in the marine environment, it must be converted

if it is to be utilised by most biological forms (Sharp 1983). There are two primary

processes by which this occurs – lightening and biological N fixation (Vitousek et al

1997) (Figure 1.1). Once in a biologically available form, nitrogen can undergo a

variety of further transformations including ammonification (organic N to NH4+),

nitrification (NH4+ to NO3

-) and denitrification (NO3- to N2) (Valiela 1995). The

micro-organisms responsible for these transformations are influenced by external

environmental conditions such as oxygen (Reddy & Patrick 1976), temperature (Focht

& Verstraete 1977, Kaplan et al 1979), pH (Delwiche & Bryan 1976) and nitrogen

supply (Addiscott et al 1991).

Chapter 1 23

Figure 1.1 Conceptual diagram of the nitrogen cycle indicating dominant nitrogen inputs and processes in atmospheric, terrestrial and aquatic environments.

Anthropogenic activities are also responsible for converting gaseous nitrogen into

biologically available forms, primarily due to fossil fuel combustion and industrial

fixation of N2 (Vitousek 1997). Subsequently, the loading of N to the world’s rivers,

estuaries and oceans has strongly been influenced by human population densities and

land use (Howarth et al 1996, Smith & Hitchcock 1994, Vitousek 1994, Vitousek et al

1997). The major sources of fixed N2 to coastal waters include wastewater disposal,

atmospheric contamination and fertilizer use (Figure 1.1) (Cole et al 1993, Lee &

Olsen 1985, Nixon 1986, Nixon & Pilson 1983). Another nitrogen source to marine

environments, likely to increase dramatically in the future, is aquaculture effluent.

24 Introduction

Dependence on aquaculture for marine products will strongly be associated with

population increases and resulting pressures on wild fisheries stocks (FAO 2000).

Three dominant sources of anthropogenic nitrogen inputs to the marine environment

which are assessed in this thesis include urban sewage, sugar cane agriculture and

shrimp aquaculture. Each activity has unique loadings to surface waters, with nutrient

inputs varying spatially and temporally due to the nature of the activities and their

environmental controls.

1.2.1 Urban Sewage

Sewage effluent can be a significant component of coastal nutrient enrichment in

populated coastal communities (Smith Evans & Dawes 1997, Sweeney 1980b). The

degree of influence of sewage discharged to coastal waters depends primarily on

treatment technology preceding effluent disposal and volume discharged. Sewage

treatment is classified as primary, secondary, or tertiary, depending on the degree to

which the effluent is purified. Primary treatment is removal of floating and suspended

solids. Secondary treatment uses biological methods such as digestion to remove

organic matter. Complete, or tertiary, treatment removes all but a negligible portion of

bacterial and organic matter and targets undesirable nutrient species such as NO3- or

PO43- (Henze 1995). The relatively standard approaches of using primary and

secondary sewage treatment, lower nitrogen discharges on average by approximately

20-25%, although there is variation among different treatment plants (NRC 1993a,

Viessman Jr. & Hammer 1998). Additional tertiary treatment for nutrient removal

can lower nitrogen discharges by 80-88% (NRC 1993a), however this process is

expensive and not yet widely used. Subsequently, disposal of treated sewage effluent

can be a significant point-source contributor of nitrogen to marine environments.

Chapter 1 25

Sewage effluent disposal in larger sewage treatment plants is generally continuous,

whereas intermittent discharges do occur in smaller plants, often coinciding with tidal

regimes aimed at decreasing residence times.

1.2.2 Crop Agriculture

Nitrogen is vital to the growth of plants and therefore is applied vigorously to soils, as

fertiliser, to increase crop yields. Over 85% of the nitrogen fixed commercially goes

into the production of fertiliser (Nixon 1995), which equated to ~ 80 million metric

tonnes in 1990 (Vitousek et al 1997). A small, but significant, fraction of the total

agricultural N applied to land is in excess of plant requirements for growth, and

surplus N may: 1) accumulate in soils; 2) move from the land into surface waters; 3)

migrate into groundwaters; or 4) enter the atmosphere via ammonia volatilisation and

nitrous oxide production (Smith et al 1999). Entry of fertiliser N into aquatic systems

is almost always via diffuse sources, making regulations and monitoring difficult.

One aspect that predominantly controls these diffuse inputs is rainfall, which acts as a

conduit of nitrogen between land and surface waters. Subsequently, nitrogen inputs to

coastal waters are not continuous and a large proportion of annual nitrogen loads

coincide with rainfall events, which is strongly seasonal in tropical/subtropical

environments (Eyre 1998).

1.2.3 Shrimp Aquaculture

Discharges of water from shrimp ponds occur during water exchange to improve the

quality in the ponds and during periods of shrimp harvest (Teichert et al 1999).

Shrimp aquaculture is fuelled by high protein feed pellets to produce high rates of

shrimp growth. However, only ~20% of nitrogen in these pellets is retained as shrimp

biomass, thereby producing large concentrated amounts of gaseous, dissolved and

26 Introduction

particulate waste (Burford & Glibert 1999, Lin et al 1993, Moriarty 1997). Untreated

wastes, principally high in NH4+, plankton and suspended solids, are generally

discharged directly into the environment. Pond effluent release is strongly dependent

on the growth cycle of shrimp and the status of pond water quality, discharges

increasing towards shrimp maturity and harvest. With the current projected expansion

of shrimp farming in most coastal areas of the world, large scale increases in nutrients

and suspended solids in receiving waters are likely (Jones et al 2001).

1.3 Understanding the Problem

The challenge of assessing and monitoring coastal systems is the synthesis of classical

water quality measures (physical/chemical parameters) with biotic responses. In the

past, indicators have often been developed haphazardly, often reflecting the interests

of specific user groups and value systems, rather than according to more broad scale

ecological criteria. There are numerous cases where considerable financial and

human resources have been used for sampling aquatic environments, often resulting in

data rich, information poor outcomes due to samples not providing information

required to make informative decisions (Maher et al 1994). While it is relatively easy

to discern differences between a healthy vs. a severely degraded ecosystem, choosing

appropriate indicators of ecosystem change (‘vital signs’) is much more difficult, as it

requires an understanding of the ecosystem and the various pressures on the

ecosystem (Wilson 1994). Yet, choosing appropriate indicators is one of the keys to

developing an effective environmental monitoring program.

Chapter 1 27

1.4 Thesis Outline

Indicators for assessing and monitoring nutrient influences in coastal waters were

developed in a semi-enclosed coastal embayment receiving large sewage inputs

(Chapter 2 & 3). Indicators comprised of δ15N signatures of marine flora and

phytoplankton bioassays, combined with standard physical/chemical parameters. The

potential of these indicators to provide useful information on the influence of shrimp

farm effluent in a tidal mangrove creek (Chapter 4), and agricultural and sewage

inputs into a coastal river estuary (Chapter 5), were also investigated.

Results and interpretation are represented as a series of papers and subsequently

chapters 2 to 5 may be read independently of the thesis as they have been prepared,

submitted or accepted for publication. The details of the publication status appear

below:

Chapter 2. A new approach for detecting and mapping sewage impacts. 2001. Marine Pollution Bulletin, 42: 149-156.

Chapter 3. Nutrient Bioassays – a tool for determining phytoplankton bloom

potential in coastal waters. In preparation for Marine Ecological Progress Series.

Chapter 4. Assessing the influence of shrimp pond effluent on receiving waters

using plant bioindicators. Submitted to Estuarine, Coastal and Shelf Science.

Chapter 5. Assessing the seasonal influence of sewage and agricultural nutrients

in a sub-tropical river estuary. In preparation for Estuaries. Chapter 6. Discussion.

To identify overall trends found and synthesise divergent results into a single

depiction for each study region, several conceptual diagrams were compared and

contrasted. These diagrams depict the major environmental features evident in each

28 Introduction

study region and attempt to encapsulate information attained from the different

systems studied. The application of approaches taken to assess water quality in these

systems is then discussed, and examples of how these techniques have been adopted

by management authorities described.

CChhaapptteerr 22 A new approach for detecting and mapping

sewage nitrogen

Abstract

Increased nitrogen loading has been implicated in eutrophication occurrences world-

wide. Much of this loading is attributable to the growing human population along the

world’s coastlines. A significant component of this nitrogen input is from sewage

effluent, and delineation of the distribution and biological impact of sewage-derived

nitrogen is becoming increasingly important. This study demonstrates a technique

that identifies the source, extent and fate of biologically available sewage nitrogen in

coastal marine ecosystems. This method was based on the uptake of sewage nitrogen

by marine plants and subsequent analysis of the sewage signature (elevated δ15N) in

plant tissues. Spatial analysis was used to create maps of δ15N and establish

coefficient of variation estimates of the mapped values. Results showed elevated δ15N

levels in marine plants near sewage outfalls in Moreton Bay, Australia, a semi-

enclosed bay receiving multiple sewage inputs. These maps of sewage nitrogen

distribution are being used to direct nutrient reduction strategies in the region and will

assist in monitoring the effectiveness of environmental protection measures.

30 Mapping Sewage Impacts

2.1 Introduction

The nitrogen cycle is one of Earth’s most important elemental cycles, and also the

most influenced by human activity (Heaton 1986). A significant component of marine

eutrophication in many near shore environments can be attributed to anthropogenic

inputs of sewage nitrogen (Lee & Olsen 1985, Nixon 1986). Managing and

monitoring the effects of sewage entering marine ecosystems has become a major

environmental challenge (Nixon 1995). The distribution of sewage effluent in marine

ecosystems can be mapped with such parameters as salinity, dissolved nutrients,

bacteria, organic matter composition, radioisotope tracers, dye fluorescence, water

current measurements, and δ15N ratios in water and sediment receiving effluent

(Lindau et al 1989, Smith Evans & Dawes 1997, Sweeney 1980a). These techniques

are useful for determining the physical extent of sewage effluent, but provide little

insight into the biological uptake and impact of sewage nutrients. Furthermore,

temporal and spatial variability confounds interpretation of these parameters. This

study developed a technique, involving nitrogen stable isotopes in marine plants, that

provides temporally integrated information on the biologically available, and

therefore ecologically significant, component of sewage nitrogen.

2.1Nitrogen Stable Isotopes

There are two naturally occurring atomic forms of nitrogen. The common form that

contains 7 protons and 7 neutrons is referred to as nitrogen 14 and is expressed as 14N.

A heavier form that contains an extra neutron is called nitrogen 15 and is expressed as

15N. By measuring the ratio of 15N to 14N in dried plant tissue, and comparing to a

worldwide standard, the relative amount of 15N, or δ15N in the plant can be

determined, as described below:

Chapter 2 31

δ15N (‰) = (Rsample / Rstandard – 1) X 103

where R is defined as the atomic 15N/14N ratio. The global standard used is

atmospheric N2. The isotopic composition of atmospheric N2 is considered to be

uniform with a 15N abundance of 0.3663% (Junk & Svec 1958, Mariotti 1983,

Sweeney et al 1978).

The various sources of nitrogen pollution to coastal ecosystems often have

distinguishable 15N/14N ratios, thereby providing a means to identify the source of

pollution (Heaton 1986). For example, nitrogen fertiliser and sewage-derived nitrogen

have distinct differences in their δ15N signatures. The main method of nitrate and

ammonium fertiliser production is by industrial fixation of atmospheric nitrogen,

resulting in products that have δ15N values close to zero. However in animal or

sewage waste, nitrogen is excreted mainly in the form of urea, which when

hydrolysed, produces a temporary rise in pH. The more basic conditions favour

conversion to ammonia which is easily lost by volatilisation to the atmosphere.

Fractionation during this process results in the ammonia, which is lost from the

system, being depleted in 15N. The remaining ammonium, now correspondingly

enriched in 15N, is subsequently converted to 15N-enriched nitrate which is more

readily leached and dispersed by water (Heaton 1986). The elevated δ15N signature of

treated sewage (~10‰) therefore distinguishes it from other nitrogen sources entering

marine ecosystems (cf. fertilizer nitrogen ~ 0‰) (Heaton 1986).


Figure 2.1 Location and nitrogen loading (tonnes/yr) of sewage effluent outfalls into Moreton

Bay and tidal estuaries.

2.2 Sewage Nitrogen and Marine Plants

Marine plants can absorb and assimilate sewage-derived nitrogen, their tissue δ15N

reflecting exposure to sewage over a given timeframe (Peterson & Fry 1987).

Elevated δ15N signatures have been identified in marine plants exposed to seabird

guano (Wainright et al 1998), septic contaminated groundwater (McClelland et al

1997), and sewage effluent (Cabana & Rasmussen 1996, Grice et al 1996, Hansson et

al 1997, Hobbie & Fry 1990, Udy & Dennison 1997b). Variations of δ15N in

naturally occurring marine plants adjacent to sewage outfalls, therefore, provides a

Chapter 2 33

technique for detecting and mapping the geographical extent of biologically available

sewage nitrogen.

In many cases, however, submersed marine vegetation has disappeared in regions

impacted by sewage. To map the fate of sewage derived N within these regions, a

method was developed which utilised changes in the δ15N values of macroalgae

deployed in situ at any desired location. This technique of mapping sewage nitrogen

with naturally occurring and deployed marine plants was tested in Moreton Bay,

Queensland, Australia.

2.3 Study Region

Moreton Bay, on the east-coast of Australia (27o 15 S, 153o 15 E), is a sub-tropical,

shallow coastal embayment (1.5 x 103 km2). Its drainage catchment (2.1 x 104 km2)

contains an urban center with a population of approximately 1.5 million people.

Development is concentrated on the western side of the Bay and high nutrient

concentrations in surrounding waters reflect this urban influence (Dennison & Abal

1999). The majority of treated sewage effluent is discharged into 4 river estuaries on

the western side of Moreton Bay, with over 50% discharged into the Brisbane River

estuary (~2000 tonnes nitrogen/year) (Figure 2.1). The eastern side of Moreton Bay

receives low-nutrient oceanic water and few anthropogenic nutrients (Udy &

Dennison 1997b).

2.4 δ15N of ambient marine plants

A variety of naturally occurring marine plants were collected throughout Moreton Bay

and the estuarine regions of its catchment. Marine plant species collected included a


seagrass (Zostera capricorni), attached macroalgae (Gracilaria edulis, Catenella

nipae) and a mangrove (Avicennia marina). Following collection, plant material was

oven dried and samples were oxidised in a Roboprep CN Biological Sample

Converter (Dumas Combustion). The resultant N2 was analysed by a continuous flow

isotope ratio mass spectrometer (Europa Tracermass, Crewe, U. K.) for δ15N isotopic

signatures.

Figure 2.2 δ15N values of naturally occurring marine plants (mangroves, seagrass and macroalgae) throughout Moreton Bay and tidal estuaries.

A strong west-east gradient of δ15N signatures were identified in these naturally

occurring marine plants (Figure 2.2). Values were highest (~10‰) in proximity to

sewage outfalls, values decreasing with increasing distance from the central regions of

Chapter 2 35

the western shore. Values around the mouths of the Caboolture R. and Logan R. were

relatively low in comparison to values around the Brisbane and Pine R. mouths.

Sewage nitrogen from upstream outfalls in the Caboolture and Logan R. appears to be

dissipated by flushing of low δ15N seawater resulting in low δ15N signatures

downstream. There are large sewage outfalls located at the mouths of the Brisbane

and Pine Rivers, and marine plants near these sources had elevated δ15N values.

Marine plants collected from the oceanic influenced eastern portions of the bay had

δ15N always less than 3.0‰.

2.5 δ15N of deployed marine plants

Utilising δ15N signatures of naturally occurring marine plants is limited by their

distribution. In regions receiving riverine inputs from urbanised catchments, growth

of benthic marine vegetation is often restricted due to a number of variables,

particularly the lack of light. Restricted marine plant distribution decreases the ability

to map sewage plumes with adequate spatial resolution. In order to overcome

restricted distributional ranges, samples of the red macroalgae, Catenella nipae, were

incubated at approximately 100 sites along the western shore of Moreton Bay. At

each site, macroalgae were housed in transparent, perforated chambers and suspended

in the water column for 4 days at ~50% light (secchi disk/2) using a combination of

buoy, rope and weights (secchi depth varied 0.5 - 4 m) form (Figure 2.3). This

timeframe allowed maximal deployment for the macroalgae to experience the various

driving forces affecting nutrient concentrations in their surrounding environment (e.g.

pulsed nutrient discharges, tidal flushing) before biofouling and siltation of the

polycarbonate cups reduced light availability to the housed macroalgae. C. nipae

(δ15N ~2‰) was collected from a low nutrient environment on the eastern side of


Moreton Bay and deployed in situ in a radiating grid pattern adjacent the 4 major river

mouths entering Moreton Bay (Figure 2.4). Sites were selected using GIS software

(ESRI Arcview 3.1) and located in the field using a differential Global Positioning

System. Following deployment of macroalgae, samples were analysed for δ15N using

the same technique as used for the naturally occurring plants. This technique of

macroalgal deployment was applied in September 1997 and February 1998. There

was little rainfall in the region prior to and during both incubation periods.

Figure 2.3 Macroalgal deployment chambers with embedded image of macroalgae incubating in situ

2.6 Spatial Analysis

Deployment sites were arranged radially around river mouths to provide a spatial

resolution of sewage plumes. δ15N values of the macroalgae following deployment

were spatially interpolated using universal Kriging (Cressie 1993), which also

Chapter 2 37

provided spatially structured estimates of uncertainty. Interpolations were back-

transformed to the original scale. Contours of δ15N were based on the logarithm of

δ15N, which was well described by a normal distribution, and which had stable

variance. An exponential model was fitted to an empirical semi-variogram

constructed from the residuals of a quadratic surface, fitted to the logarithm of δ15N

(to remove large-scale trends). δ15N values were ranked into six categories, ranging

from <3‰ to >9‰. Values above 3‰ indicate sewage-derived nitrogen was

assimilated by the macroalgae. The contours of these six categories define the degree

of biological influence of sewage nitrogen in the bay.

Figure 2.4 Locations of deployed macroalgae in Moreton Bay (red solid circles).


Two sewage plumes were evident emanating from the Brisbane and Pine Rivers

(Figure 2.5a). Macroalgal δ15N increased from less than 3‰ in central-western

Moreton Bay to greater than 9‰ in close proximity to sewage outfalls. δ15N values of

macroalgae were above 3‰ up to 5 km from these river mouths, indicating that the

influence of sewage extended to this distance. Very little or no sewage signal was

detected around the mouths of the Caboolture or Logan Rivers supporting the results

measured with naturally occurring marine plants.

The distribution of δ15N was found to be more widespread in February rela tive to

September (Figure 2.5b). Although two sewage plumes can still be discerned

emanating from the Brisbane and Pine Rivers, the delineation is much less distinct.

Whereas values approached control levels (<3‰) at 5 km from the source in

September, values were around 5‰ at this distance in February. δ15N values were

high in macroalgae deployed throughout Waterloo Bay, a feature not apparent in algae

deployed in September. As with previous surveys, little or no sewage signal was

detected around the Caboolture or Logan Rivers.

The variation in δ15N values of the macroalgae between the two deployment periods

may be related to a variety of biological and environmental factors. In either case, the

biological effect of sewage-derived nitrogen is being measured. Although rainfall

was similar between the two deployment periods, water column temperature increased

by 5oC. This is likely to have increased macroalgal productivity in February,

increasing algal uptake of nitrogen (Peckol et al 1994.).

a b

Figure 2.5 Spatial distribution of deployed macroalgal δ15N values in September 1997 (a) and February 1998 (b). Macroalgae (Catenella nipae, Rhodophyte) was deployed at ~100 sites (yellow solid circles) in Moreton Bay, Australia.


In addition, sediment microbial activity is enhanced with increased temperature

(Jeffrey et al 1995), possibly increasing sediment mineralisation of sewage-derived

organic matter resulting in 15NH4+ fluxes into the water column available for algal

uptake. These explanations may also be pertinent to the large response measured in

Waterloo Bay in February. In addition, a hydrodynamic model of Moreton Bay

predicts a south-westerly water flow in this region at this time of year. This seasonal

circulation pattern may be responsible for driving sewage from Bramble Bay into

Waterloo Bay (Dennison & Abal 1999).

2.7 Statistical verification

An integral component of this technique was the estimation of uncertainty (co-

efficient of variation) of the interpolated data. Interpolation between samples is

subject to error, and maps must, therefore, contain estimates of uncertainty.

Uncertainty was quantified as a coefficient of variation, obtained by expressing the

standard error of prediction as a percentage of the interpolated value. The co-efficient

of variation provides a means of assessing the accuracy of the interpolated data. The

co-efficient of variation map produced for February indicates that the interpolated

data between points is 75-90% accurate for the majority of the western shore (Figure

2.6).

Chapter 2 41

Figure 2.6 Co-efficient of variation map of interpolated δ15N values for February.

2.8 Applications

This technique of mapping biologically available sewage nitrogen has created a

baseline for future water quality management and monitoring in the region. The

results have highlighted the need for improved standards for sewage discharge. As a

direct result of this study, six regional authorities have initiated improvements to

sewage treatment facilities with emphasis on nitrogen removal and begun

investigations and implementation of wastewater re-use schemes (MBCWQMST


1999). This method of mapping sewage plumes has also been incorporated into a

regional water quality monitoring program.

2.9 Summary

Previous techniques for detecting and mapping sewage effluent have provided

information on various sewage components, however, in coastal ecosystems limited

by nitrogen availability, the detection of nitrogen is of paramount importance. By

using δ15N signatures of marine plants, biologically available and therefore

ecologically significant nitrogen can be detected and mapped in coastal ecosystems.

The measurement of sewage impacts using marine plants has proven to be valuable

for environmental management and monitoring of sewage nitrogen in Moreton Bay.

It is believed that this novel approach will lead to a reassessment of methods for

detecting and mapping sewage impacts in many other coastal ecosystems.

CChhaapptteerr 33 Nutrient Bioassays: a tool for determining phytoplankton bloom potential in estuarine

and coastal waters

Abstract This study demonstrated that measuring physical/chemical variables alone was not

sufficient for predicting the responsiveness of a system to phytoplankton blooms. A

Phytoplankton Response Index (PRI) was developed to assess bloom potential of

natural phytoplankton communities. This index was a function of growth rate and

maximum abundance of phytoplankton following increased nutrient and /or light

availability. Bioassays were cultured for 7 days and phytoplankton growth measured

using in vivo fluorescence. Bioassay responses were investigated along four

eutrophication gradients (river-bay) in the Moreton Bay region, Australia. Temporal

patterns in bioassay responses were also investigated at 11 sites throughout the region

over a nine month period. River regions consistently displayed elevated nutrient

concentrations (NO3- - 10.5 µM) and light attenuation (secchi - 0.7m) than the bays

(NO3- - 0.5µM and secchi - 1.1m, respectively), though salinities between these

regions were comparable due to extended dry periods prior to sampling (~32 ‰).

Both nutrient and light PRI’s were greater in the river (26.4 ± 2.3 and 8.3 ± 1.4 fsu.d-1,

respectively) than the bay regions (15.5 ± 2.1 and 1.4 ± 0.6 fsu.d-1, respectively)

(p<0.001). Nutrient PRI’s varied seasonally (range of PRI) with greater responses in

summer (Feb; p<0.001), however no significant seasonal difference was measured in

light PRI’s. Relationships between physical/chemical variables and phytoplankton

bioassay responses were investigated using multiple regression analysis. Temperature

and NO3- explained the most variability in nutrient and light PRI’s, respectively.

Regression equations, obtained from the spatial data set, were applied to

physical/chemical data obtained from seasonal sampling to assess predictability of

PRI. Predictive ability was poor (R2 nutrient PRI = 0.15; R2 light PRI = 0.35)

indicating that bioassay response was not simply a function of physical/chemical

parameters. This justifies the use of bioassays in assessing a systems susceptibility to

phytoplankton blooms.

44 Phytoplankton Bioassays

3.1Introduction

Cultural eutrophication can lead to deteriorating ecosystem health (Dennison & Abal

1999). Eutrophication can be driven by either non-point sources of nutrients such as

agricultural, urban runoff and septic tank effluents, or point sources such as industrial

effluents, aquaculture, and municipal sewage. Symptoms of declined ecosystem

health are repeatedly demonstrated by nuisance algal blooms which have no historical

presence (Smayda 1990), declines in recreational and commercial fisheries value

(Wolter et al 2000), the presence of pathogenic organisms (Wu 1999), anoxia of

bottom waters and sediments (Malakoff 1998), and foul odours and tastes (Paerl &

Bowles 1987). To arrest and reverse cultural eutrophication of coastal waters,

regulatory agencies must be able to predict the response of coastal ecosystems to

changes in nutrient loading.

Typically, the first sign of eutrophication is an increase in phytoplankton biomass

above that of seasonal variation (D'Elia 1987). The intimate relationship between

phytoplankton and dissolved inorganic nutrients (Valiela 1995), can be used to

estimate the response of aquatic systems to nutrient enrichment. However, ambient

nutrient concentrations are not always the best predictors of nutrient status of the

phytoplankton themselves, as deficient cells can take up nutrients in excess of their

immediate growth requirements (Borum & Sand Jensen 1996, Healey 1979, Lean

1981).

Several approaches have been used to assess nutrient limitation in estuaries and

coastal waters (D'Elia 1985). These methods include: 1) assays of a metabolic

indicator, such as enzyme activity, 2) mathematical models which utilize input-output

Chapter 3 45

response to nutrients, 3) inferences based upon stoichiometric relationships of

dissolved nutrient concentrations, and 4) bioassays involving time-dependent growth

of nutrient enriched phytoplankton – phytoplankton bioassays (Smith & Hitchcock

1994 and references therein). This last method has been widely used as a tool for

identifying growth- limiting nutrients and formulating nutrient input constraints

(Bonetto et al 1991, Carignan & Planas 1994, Carrick et al 1993, Cowell & Dawes

1991, Gerhart & Likens 1975, Paerl & Bowles 1987, Smith & Hitchcock 1994,

Thompson & Hosja 1996). Determining growth limiting nutrients is generally

achieved by monitoring phytoplankton growth responses to either individual or

combined additions of inorganic or organic nutrients (Paerl & Bowles 1987).

Phytoplankton are particularly applicable as indicators of nutrient limitation, due to

their ubiquitous distribution and their sensitivity to perturbations in their environment

(Valiela 1995).

Hence, it is evident that phytoplankton bioassays have been successful in identifying

growth limiting nutrients of aquatic environments. However, this study has

endeavoured to use phytoplankton bioassays to infer their bloom potential to elevated

nutrients in Moreton Bay, Australia. Moreton Bay has previously been described as

nitrogen limited (Dennison & Abal 1999, Horrocks et al 1995, O'Donohue et al 1997,

Udy & Dennison 1997a). To assess phytoplankton bloom potential, a Phytoplankton

Response Index (PRI) was developed which provided a measure of phytoplankton

bloom potential as a function of maximum growth and response time. The

relationships between PRI’s and ambient physical/chemical conditions were then

investigated to determine which physical/chemical parameter/s best predict

phytoplankton responses in Moreton Bay.


3.2 Materials & Methods

3.2.1 Site Location

Moreton Bay (27o 15 S, 153o 15 E) is a sub-tropical, shallow coastal embayment (1.5

x 103 km2) on the east coast of Australia (Plate 3.1). Its drainage catchment area is

dominated by the large catchment of the Brisbane River (13,100 km2), followed by

the Logan/Albert (3,157 km2), Pine Rivers (808 km2), and the Caboolture River (354

km2) catchments. Development is concentrated on the western side of the Bay and

high nutrient and suspended sediment concentrations in surrounding waters reflect

this influence (Dennison & Abal 1999). Strong gradients in water quality exist from

western to eastern sides of the bay, the eastern side receiving relatively low-nutrient

oceanic water and few anthropogenic inputs (Udy & Dennison 1997b).

Plate 3.1 Satellite image of Moreton Bay and catchment. Four major rivers entering Moreton

bay are highlighted in blue.

Chapter 3 47

3.2.2 Sampling Strategy and Analysis

Relationships between phytoplankton responses and various grades of water quality

experienced within the Moreton Bay region were examined at various spatial and

temporal scales:

3.2.2.1Spatial Comparison

In order to assess phytoplankton responses to strong water quality gradients, 4x

~20km long transect were surveyed. The transects began ~10km upstream from each

major river mouth and ran linearly ~10km out from each river mouth (Figure 3.1)

based on plume data collected by Costanzo et al. (2001) and nutrient surveys

presented in Dennison and Abal (1999). Survey sites were positioned at

approximately 2km intervals along transects. At each site, physical/chemical

variables were measured and water was collected for phytoplankton bioassays. This

assessment was conducted twice, once in early September 1997 (austral winter /

spring) and February 1998 (austral summer).

3.2.2.2 Temporal Comparison

Secondly, 11 sites were positioned around the Moreton Bay region and bioassay

responses and physical/chemical parameters measured at nine monthly intervals from

September 1997 - June 1998 in order to assess seasonal responses (Figure 3.1).


Monthly sample site

Transect sample site

Figure 3.1 Study location showing the four major rivers entering Moreton Bay and the study transects with sampling sites along each (solid circles) and monthly sampling sites (open squares).

3.2.3 Physical/Chemical parameters

Physical parameters measured in this study included secchi depth, salinity, chl a, pH

and temperature. Salinity and light attenuation were measured in the field using a

Horiba Water Quality Checker Model U-10 and a secchi disk, respectively. Duplicate

water samples for chlorophyll a analysis were collected from approximately 0.1 m

below the surface at each site, filtered (Whatman GF/F filters; nominal pore size 0.7

Chapter 3 49

µm) and then stored over dry ice. Samples for chlorophyll a were analysed on return

to the laboratory using a specific formula for chl a (Parsons et al 1984).

Inorganic nutrient samples (NH4+, NO3

-, PO43-) were collected from approximately

0.1 m below the water surface. For dissolved nutrients, samples were filtered in the

field to remove particulate matter using a 60 ml syringe and Sartorius Minisart 0.45

µm disposable glass fibre polycarb filters. Filtered and unfiltered water samples were

frozen immediately on dry ice and transported to the laboratory where samples were

analysed using auto-analyser chemical techniques (Clesceri et al 1998).

3.2.4 Phytoplankton Bioassays

Bioassays were run over 7 days by incubating 4 L samples of water from each study

site with added nutrients. Water samples were collected in 30 L black barrels (acid

washed) and transported to an outdoor incubation facility. Water from each site was

filtered through a 200 µm mesh (to screen out large zooplankton grazers) into sealed

transparent 6 L plastic polycarbonate containers (Plate 3.2).

Plate 3.2 Bioassay containers used for phytoplankton bioassays. Each container holds 4 L of site water. The container on the right at the front was a nutrient treatment and the subsequent phytoplankton response is evident by the green coloured water.


An artefact of the technique is that particulates, generally kept in resuspension in situ,

settled out in the bioassay containers permitting light penetration, thereby allowing

the control bag to be used as a light stimulation experiment. Further light stimulation

would be contrary to scenarios experienced in the natural environment as light cannot

increase above ambient levels, whereas nutrient concentrations can increase well

above ambient concentrations dependent on inputs.

The response of the phytoplankton to nutrient addition and enhanced light availability

was determined by measuring biomass increases (via in vivo fluorescence

measurements) relative to a control sample with no added nutrients. Units used for

these measurements were fluorescence units (fsu). Nutrient addition bioassays

comprised of four nutrient applications (added at the beginning of the experiment) to

one bioassay container and were applied at sufficient concentrations to ensure

saturation of phytoplankton (nitrate- NaNO3 (200 µM), ammonium- NH4Cl (30 µM),

phosphorus- Na2HPO4.H2O (20 µM), silica NaSiO3 (66 µM)) (O'Donohue et al 1997).

Bioassay containers were then incubated outdoors in large tanks (2m diameter, 0.5m

deep) under 50% ambient surface irradiance (covered with 50% shade cloth) with a

flow through system to maintain a steady ambient water temperature (± 2oC of the

ambient found at each site) (Plate 3.3). Bioassay containers were removed at the

same time each day, gently shaken and a 20 ml sub-sample was poured into 30 ml

clear glass cuvette that was then dark-adapted for a period of 30 minutes. Cuvettes

were mixed and wiped clean before being placed in a Turner TD1000 design

fluorometer to measure fluorescence. As the total volume removed from bioassay

containers each day was small (<1%), it was assumed that decreasing volume had

little influence on the outcome of the experiment s.

Chapter 3 51

Plate 3.3. Outdoor incubation tank with bioassay containers inside.

3.2.4.1 Phytoplankton Response Index

In order to analyse the large amounts of data that result from this procedure, and to

quantify the response-potential of phytoplankton from each site to simulated nutrient

additions, a Phytoplankton Response Index (PRI) was developed. Three aspects of

phytoplankton bioassays that were deemed important in assessing phytoplankton

bloom potential were: 1) initial fluorescence, 2) maximum response, and 3) time to

maximum response, all of which are incorporated into the PRI (Figure 3.2).

A light stimulated response was inferred from growth of phytoplankton in the control

bag where no nutrients were added. Light Phytoplankton Response Index (LPRI) was

derived from the following equation:

Light PRI = (FC – FO) / TC

where FC is maximum control fluorescence, FO is initial fluorescence and TC is time to

maximum control fluorescence.


0

10

20

30

40

0 1 2 3 4 5 6 7

Day

Fluo

resc

ence

(FS

U)

Nutrient

Control

FN

TN

FC

TC

FCN

FO

Figure 3.2 Graphic description of phytoplankton bioassay responses over 7days to added nutrients and no added nutrients. FO – initial fluorescence; FN – maximum nutrient fluorescence; FC – maximum light fluorescence; FCN – control fluorescence at maximum nutrient fluorescence; TN – time to maximum nutrient fluorescence; TC – time to maximum light fluorescence.

Nutrient stimulated responses were inferred growth of phytoplankton above that of the

control phytoplankton growth. Nutrient Phytoplankton Response Index (NPRI) was


Nutrient PRI = (FN – FCN) / TN

where FN is maximum nutrient fluorescence, FCN is control fluorescence at the time of

FN and TN is time to maximum nutrient fluorescence. Units for PRI were fsu.d-1.

3.2.5 Data Analyses

Bioassay data were analysed in a hierarchical structure, with investigations of

variability between transects (Caboolture, Pine, Brisbane and Logan), surveys

(September ’97 and February ’98) and regions (rivers and bays). This approach was

undertaken in order to assess both temporal and spatial variability in phytoplankton

bioassay responses to nutrients and light in Moreton Bay.

Chapter 3 53

Differences in phytoplankton responses between transects, surveys and regions were

tested by three-way ANOVA’s (data was v (+ 0.5) transformed). Multiple regressions

were conducted on the transect data to determine which explanatory variables

(chlorophyll a, secchi depth, temperature, salinity, pH, NH4+, NO3

- and PO43-) best

explained the response variables (nutrient PRI and light PRI). Regression equations

were then applied to the monthly data set to assess the predictability of PRI’s from

physical/chemical parameters alone. Correlations were then conducted between

observed and predicted PRI’s to assess predictive power. Data were assessed for

homogeneity of variances using Cochran’s test (Underwood 1981). Data used in the

correlation required a v (+ 0.5) transformation (Zar 1984).

3.3 Results

3.3.1 Physical / Chemical Properties

Variability was observed in physical/chemical parameters between transects, surveys

and regions. The highest overall mean chlorophyll a concentrations were recorded in

the Pine transect (4.3 ± 0.9 µg/L), lower concentrations measured respectively in the

Logan (3.5 ± 0.9 µg/L), Caboolture (2.0 ± 0.3 µg/L) and Brisbane (0.9 ± 0.1 µg/L)

transects (Table 3.1). The highest overall nutrient concentrations were recorded in the

Brisbane transect, followed by the Pine, Logan and Caboolture transects, respectively.

The relative concentrations of different nutrient species differed between these rivers

with highest mean NO3-N concentrations in the Brisbane transect (15.3 ± 3.4 µM) and

highest mean NH4-N concentrations in the Pine transect (4.3 ± 1.3 µM).

Temperatures were approximately 3oC higher in the February survey than the

September survey (23.1 and 26.3 oC, respectively). The ‘river’ regions were typically


higher in chlorophyll a and nutrient concentrations than ‘bay’ regions (3.3 ± 0.6 and

1.6 ± 0.3 µg/L, respectively). Secchi depths and salinities were lower in the ‘river’

(0.7 ± 0.0 m and 30.4 ± 1.7 ‰, respectively) than the ‘bay’ regions (1.1 ± 0.1m and

36.8 ± 0.2 ‰, respectively).

Table 3.1 Mean (± SE) water column physical/chemical measurements between transects, surveys and regions.

Transect CaboolturePineBrisbaneLogan

Survey SeptemberFebruary

Region RiverBay

8.28.18.28.2

8.08.4

8.18.3

(0.1)(0.1)(0.1)(0.1)

(0.0)(0.1)

(0.0)(0.0)

pH

2.04.30.93.5

2.92.7

3.31.6

(0.3)(0.9)(0.1)(0.9)

(0.6)(0.6)

(0.6)(0.3)

Chlorophyll a(µg/L)

1.10.80.90.8

0.80.8

0.71.1

(0.1)(0.1)(0.1)(0.0)

(0.1)(0.1)

(0.0)(0.1)

Secchi(m)

24.924.924.524.4

23.126.3

24.624.4

(0.5)(0.4)(0.5)(0.4)

(0.1)(0.1)

(0.3)(0.4)

Temperature(oC)

35.226.235.733.6

32.832.1

30.436.8

(0.8)(3.8)(0.5)(1.1)

(1.5)(1.8)

(1.7)(0.2)

Salinity(‰)

0.74.33.01.0

3.41.5

3.21.0

(0.2)(1.3)(0.6)(0.1)

(0.7)(0.4)

(0.6)(0.2)

NH4+

(µM)

1.34.2

15.93.9

9.06.0

10.50.5

(0.7)(1.7)(3.4)(2.0)

(2.0)(2.2)

(1.9)(0.2)

NO3-

(µM)

0.41.62.21.7

2.11.3

2.10.6

(0.1 )(0.4 )(0.3 )(0.6 )

(0.4 )(0.2 )

(0.3 )(0.1 )

PO43-

(µM)Factor


3.3.2.1 Transect Data

3.3.2.1.1 Nutrient Phytoplankton Response Indices (PRI)

Mean nutrient PRI’s did not differ between whole transects (p>0.05) (Table 3.2;

Figure 3.3 a). Nutrient PRI’s differed significantly between surveys and regions

(p<0.001), but not regions (Table 3.1; Figure 3.3 b & 3.3 c, respectively). Mean

responses were approximately two times higher in February than in September, and

higher in ‘rivers’ than in ‘bays’. Three-way ANOVA’s found significant interactions

between ‘transect’ and ‘region’ indicating that trends within transects were not

uniform (Table 3.2). Nutrient PRI’s in the Logan transect was responsible for this

interaction between transects and regions, as river responses were similar to bay

responses in this southern poorly flushed section of the bay.

Chapter 3 55

0

10

20

30

Caboolture Pine Brisbane Logan

Transect

Phyt

opla

nkto

n R

espo

nse

Inde

x (f

su.d

-1)

a

0

10

20

30

40

September ‘97 February ‘98

Survey

b

Phyt

opla

nkto

n R

espo

nse

Inde

x (f

su.d

-1)

0

10

20

30

40

Rivers BaysRegion

c

Phyt

opla

nkto

n R

espo

nse

Inde

x (f

su.d

-1)

Nutrient Light

Figure 3.3. Comparison of mean nutrient and light phytoplankton response indices (PRI) (± SE) between a) transects, b) surveys, and c) regions (river or bay)


3.3.2.1.2 Light Phytoplankton Response Indices (LPRI)

Mean light PRI’s did show trends between transects, highest light responses recorded

in the Brisbane transect, followed by the Pine, Logan and Caboolture transects,

respectively (p = 0.05) (Figure 3.3 a). There was, however, no significant difference

between surveys (September and February) (p>0.05) (Figure 3.3 b). Phytoplankton

response to light was greater in the ‘rivers’ than in the ‘bays’ (p<0.001) (Figure 3.3 c).

Table 3.2 Significance values for nutrient and light PRI comparisons between transects (Caboolture, Pine, Brisbane and Logan), surveys (September 1997 and February 1998), regions (river/bay) and interactions between these factors.

Factor

Nutrient PRI Light PRI

df MS p-value

SurveyTransect

RegionTransect x Survey

Survey x RegionTransect x Region

Transect x Survey x RegionError

1 16.72 <0.0013 0.018 0.92

1 19.82 <0.0013 1.16 0.36

1 2.21 0.153 9.88 <0.001

3 6.93 <0.00161 1.06

1 5.9 x 10-4 0.933 0.23 0.05

0.333 0.1

0.911 1.0 x 10-3

0.63 0.05

0.223 0.1361 0.08

1 1.67 <0.001

df MS p-value

3.3.2.2 Monthly Data

Seasonal trends were observed in both nutrient and light PRI’s from the monthly data

set (p<0.05) (Figure 3.4). PRI values increased in late spring (November), reaching a

mean maximum response index of 43 fsu.d-1 at the end of summer (February). Values

then rapidly decreased for the remaining 4 months to < 10 fsu.d-1. Elevated light

responses persisted from the beginning of summer (December) through to the end of

autumn (May).

Chapter 3 57

0

10

20

30

40

50

Oct

ober

Nov

embe

r

Dec

embe

r

Janu

ary

Febr

aury

Mar

ch

Apr

il

May

June

Month

10

15

20

25

30

Tem

pera

ture

(o C)

NPRI

LPRI

TempPh

ytop

lank

ton

Res

pons

e In

dex

(fsu

.d-1

)

Figure 3.4 Mean monthly nutrient and light bioassay responses expressed as phytoplankton

response indices (PRI) and mean monthly water column nutrients.

3.3.3 Bioassay Responses and Physical/Chemical variables

Stepwise forward multiple regression analysis of secchi depth, salinity, chl a, pH,

temperature, NH4+, NO3

-, PO43- and PRI’s, identified variables that explained most of

the variation observed in bioassay responses measured in this study, as described

below.

3.3.3.1 Transects

Temperature, salinity, secchi and NH4+ together explained 48% of the variation in

nutrient bioassays performed along the transects (Table 3.3). NO3-, salinity and

temperature, respectively, explained 75% of the variation observed in light bioassays.

3.3.3.2 Monthly

Temperature, PO43- and NH4

+ together explained 57% of the variation observed in

nutrient bioassays in the monthly data set. Physical/chemical variables NO3-, NH4

+

and PO43-, respectively, explained 51% of the variation observed in light bioassays.


Table 3.3 Significant variables that exp lain variability experienced in nutrient and light PRI’s in the transect and monthly data sets.

Data setNutrient PRI Light PRI

Variable R2 change (%)

p - value Variable R2 change (%)

p - value

Transect

Monthly

Temperature

Salinity

Secchi depth

NH4+

Temperature

PO43-

NH4+

NO3-

Salinity

Temperature

NO3-

PO43-

NH4+

21

16

6

5

<0.001

<0.001

<0.05

<0.05

35

15

7

35

15

7

<0.001

<0.001

<0.05

51

22

2

<0.001

<0.001

<0.05

28

13

10

<0.001

<0.01

<0.01

Total

Total

48

57

75

51

3.3.4 Bioassay Predictions

Regression coefficients for significant physical/chemical variables were applied to the

monthly data set, in order to assess the predictive ability of bioassay responses from

physical/chemical indices alone. The regression between predicted and observed

nutrient PRI’s did not provide a strong basis to infer bioassay responses from

physical/chemical variables alone (R2 = 0.15, n = 27) (Figure 3.5a). The regression

for light bioassays provided a slightly stronger predictive ability from

physical/chemical variables, though the poor strength of this regression does not allow

accurate predictions to be made (R2 = 0.35, n = 48) (Figure 3.5b).

Chapter 3 59

R2 = 0.15

0

2

4

6

8

10

0 2 4 6 8 10

Observed NPRI

Pred

icte

d N

PRI

R2 = 0.15

0

2

4

6

8

10

0 2 4 6 8 100

2

4

6

8

10

0 2 4 6 8 10

Observed NPRI

Pred

icte

d N

PRI

a

R2 = 0.350

1

2

3

0 1 2 3

Observed LPRI

Pred

icte

d L

PRI

R2 = 0.350

1

2

3

0 1 2 3

Observed LPRI

Pred

icte

d L

PRI

b

Figure 3.5 Correlations between monthly predicted and monthly observed a) nutrient bioassays and b) light bioassays.

3.4 Discussion

Phytoplankton bioassays allowed spatial and temporal characterisation of

phytoplankton responses to light and nutrients in Moreton Bay. The poor power of

physical/chemical variables in predicting bioassay responses in this study highlights

the complexity of phytoplankton dynamics and signals the lack of information on

biotic influences gained from physical/chemical analyses alone. Despite this poor

predictive power, certain physical/chemical variables were found to primarily govern


light and nutrient bioassay responses, providing insight to phytoplankton processes

within the Moreton Bay region.

3.4.1 Physical/Chemical Water Quality

Physical/chemical water quality data collected in this study complements previous

work done in the region, which identifies the rivers to be nutrient and sediment laden

(Dennison & Abal 1999, Jones et al 1996). This has been attributed to long water

residence times, large urban inputs of suspended matter and nutrients, and strong tidal

resuspension of particulates (Dennison & Abal 1999). The elevated nutrient

concentrations in the Pine and Brisbane Rivers were largely a result of high volumes

of sewage effluent that is discharged into these rivers (Costanzo et al 2001). The

lower nutrient concentrations in the bay regions were primarily attributable to shorter

residence times as a result of oceanic flushing. Of the four major rivers entering

Moreton Bay, the Brisbane River is the only river that has routine navigational

dredging at the river mouth. This allows large tidal excursions in and out of the river

which leads to further resuspension of particulates in the river, as indicated by the low

secchi depths observed in this study. However, the low secchi depths in the Pine and

Logan River’s are likely to be a result of the elevated phytoplankton as represented by

elevated chlorophyll a values. The best water quality of the four rivers was in the

Caboolture River, the river with the smallest catchment and minimal sewage inputs

(Costanzo et al 2001, Dennison & Abal 1999).


There were three aspects of phytoplankton bioassays that were considered important

in this study in describing the susceptibility of the system to algal blooms.

Chapter 3 61

Historically, the magnitude of response following nutrient additions has commonly

been used as one indicator of the vulnerability of a system to nutrient additions (Aidar

et al 1996, Carrick et al 1993, Jones et al 1998, Smith & Hitchcock 1994, Van Donk

et al 1988). This value alone was not considered representative in describing

phytoplankton bloom potential, as other factors such as initial phytoplankton

abundance and time taken to reach maximum response, varied considerably between

sites and was believed to infer further information on bloom dynamics and system

susceptibility to nutrient loads. The time taken to reach a maximum response is

important, as there is a lag period in phytoplankton biomass accumulation following

nutrient additions. This lag time has even been used as an index of the capacity of

phytoplankton to bloom (O'Donohue et al 1997), where shorter lag times were

indicative of regions with a greater propensity for phytoplankton blooms, and vice-

versa.

3.4.2.1 Nutrient Responses

It was evident that nutrients were stimulating phytoplankton growth largely in the

river regions and these responses were similar between rivers. Greater nutrient

limitation found in the rivers than the bays was somewhat unexpected as the river

regions were more replete in nutrients than the bays, hence a lower nutrient

requirement by phytoplankton for growth in the bioassays. There are a number of

possible explanations why nutrients were stimulating growth in the rivers. Firstly,

there were typically higher chlorophyll a concentrations in the rivers than the bays.

Subsequently, growth of phytoplankton in the rivers had a ‘head start’ due to a higher

initial abundance of phytoplankton, likely already dominated by ‘bloom’ species that

are capable of rapid uptake and growth. Consequently, demand for nutrients was


likely to be higher in the rivers than the bays. Secondly, it is possible that the rivers

had a greater supply of micronutrients required for optimal phytoplankton growth.

This is likely as the rivers were elevated in all nutrient species measured and it’s

assumed the rivers would have higher concentrations of other non-measured nutrients

than in the bays. Thirdly, high nutrient concentrations and high light attenuation (low

light) would suggest that heterotrophic and not autotrophic processes dominate in the

rivers (O'Donohue & Dennison 1997). Removal of macro-grazers in the bioassay

containers may have removed the biological agent controlling phytoplankton growth

in the rivers. Lastly, settled suspended material inside the bioassay containers was

more typical in river bioassays. This material may act as a slow release fertiliser

aiding phytoplankton growth over the duration of the bioassay period (Uusitalo et al

2000).

Nitrogen uptake kinetics for phytoplankton in Moreton Bay, calculated from uptake

rates described by O'Donohue et al (2000), indicate a potential two-fold increase in

uptake rates from 0.34 to 0.76 µM N hr-1 with the variation of temperatures observed

between transect surveys in this study (winter/spring - 23 oC and summer - 26oC,

respectively). For the monthly data set there was an estimated 25 times increase in

potential nitrogen uptake rates (0.05 - 1.26 µM N hr-1) between winter (16oC) and

summer (28oC). This demonstrates the large-scale variations in phytoplankton

productivity that can occur seasonally in Moreton Bay and provides managers with

quantitative evidence of periods when added nutrients to the system are likely to have

the largest impact.

Chapter 3 63

3.4.2.2 Light Responses

Light limitation of phytoplankton growth was largely restricted to the river regions,

which was not unexpected, as these regions had the lowest secchi depths, highest

nutrient concentrations and typically higher chlorophyll a concentrations than the

bays. The high light responses in the Pine, Brisbane and Logan Rivers coincided with

highest nutrient concentrations and shallowest secchi’s in these rivers, as opposed to

the Caboolture River, which had lower nutrients and deeper secchi depths. The lack

of variation in light PRI’s between surveys can be explained by the minimal variation

in NO3- and salinity between the two surveys, both of which were found to be highly

significant in explaining variability in light PRI’s. Temperature explained only a

small proportion of the variability in light PRI’s and hence the lack of difference in

light bioassays between transects surveys and the monthly data set.

3.4.3 Bioassay Response Predictability

The lack of predictive power of bioassay responses, from physical/chemical variables

alone, indicates that these variables are not sufficient in explaining the variability of

bioassay responses measured in this study. In addition, the temporal variability of

physical/chemical parameters, often experienced in estuarine and coastal systems (e.g.

Eyre & Pepperell 1999), is likely to confound any predictive ability. Obviously there

were many parameters that were not taken into account purely due to logistical

reasons. Primarily, phytoplankton species composition may govern the type and

magnitude of response. However in the context of this study, the aim was purely to

ascertain bloom potential of a region following nutrient addition, irrespective of

phytoplankton composition.


It was evident in this study, however, that temperature and NO3- explained most of the

variability observed in nutrient and light PRI’s, respectively. Temperature has been

shown to be positively correlated with phytoplankton productivity (e.g. Berounsky &

Nixon 1990, Eppley 1972, Fisher et al 1999, Fong & Zedler 1993, O'Donohue &

Dennison 1997, Van Donk et al 1988), which supports results found in this study that

temperature was the variable (of the variables measured) that mostly influenced

phytoplankton growth in the nutrient bioassays. Temperature was not, however,

responsible for the variability observed in light bioassay responses. Theoretically,

responses in light bioassays should depend on the ambient concentrations of nutrients

and the abundance of phytoplankton in the water column at the time of bioassay water

collection. Without sufficient concentrations of these variables for phytoplankton

growth, bioassay responses to light are likely to be minimal. Hence the significance

of a nutrient (NO3-) being responsible for much of the variability observed in the light

bioassays. Why did NO3- explain most of the variability observed in light PRI’s?

NO3- was the predominant nutrient species in the rivers, primarily due to sewage

discharges (Costanzo et al 2001) and high rates of nitrification (Dennison & Abal

1999), resulting in nitrate concentrations ~20 times greater in the rivers than the bays.

The difference between rivers and bays was much lower for NH4+ (~3 times greater in

the river) and PO43- (~4 times greater in the river). Considering light responses were

restricted primarily to the NO3- laden river regions and negligible in the bays,

highlights why a relationship was found between light responses and NO3-.

Whereas temperature and NO3- differentiated between nutrient and light PRI’s,

secondary variables delineated between transect and monthly data sets. Salinity was

identified as the next significant variable for explaining variability in nutrient and

Chapter 3 65

light PRI’s in the river transect data, though not in the monthly data. Salinity is a

conservative component representative of seawater, of which has generally lower

nutrient concentrations than riverine and estuarine waters (Dennison & Abal 1999).

The gradient in salinity from the rivers to the bays reflects changes observed in many

of the physical/chemical variables, particularly nutrients, secchi and chlorophyll a, all

of which are likely to affect phytoplankton growth.

Nutrients (NH4+ & PO4

3-), not salinity, were secondary significant explanatory

variables in the monthly data set. This can be explained by the location of sampling

sites for the monthly data set, which were predominantly in the bay. The bay does not

experience strong trends in salinity, though does vary considerably from west to east

in nutrient concentrations (Dennison & Abal 1999), thereby nutrients having

significance in describing bioassay responses.

3.5 Conclusion

Discriminating between nutrient and light augmented phytoplankton responses is

particularly important for turbid, nutrient laden rivers, as the removal of e.g.

suspended particulates, generally promotes algal growth. Understanding and

predicting this process is important for management purposes and highlights

coordinated remedial actions of both nutrient and suspended solid loads. The poor

ability to predict bioassay responses from physical/chemical variables alone justifies

the value of conducting phytoplankton bioassays as a tool for monitoring water

quality. Subsequently, this technique has been incorporated into the regions ongoing

water quality-monitoring program aimed at assessing the success of remedial actions

taken in the bay and catchment (Dennison & Abal 1999).


This study also highlighted certain physical/chemical parameters that can guide

managers of appropriate periods and locations for nutrient release. In this study

region, there is discussion of diverting sewage effluent from the rivers and bays, back

up into the catchment for crop irrigation purposes. The results found in this study

would indicate that effluent diversion and reuse would have most effect in the summer

months when the influence of effluent on the receiving environment is greatest. This

timing also coincides with highest crop water requirements. Alternatively in the

winter months, as crop water requirements decrease, effluent can be discharged into

the bay with less negative impacts on phytoplankton biomass than the summer

months. Unfortunately, the principal problem in this strategy is changing public

perception towards sewage effluent reuse, though with ensuing environmental

problems opinions are likely to change.

CChhaapptteerr 44 Assessing the influence and distribution of

shrimp pond effluent in a tidal mangrove creek

Abstract Effluent from a land based shrimp farm was detected in a receiving creek as changes

in physical, chemical and biological parameters. The extent and severity of these

changes depended on farm operations. This assessment was conducted at three

different stages of shrimp-pond maturity, including when the ponds were empty, full

and being harvested. Methods for assessing farm effluent in receiving waters

included physical/chemical analyses of the water column, phytoplankton bioassays

and δ15N signatures of marine flora. Comparisons were made with an adjacent creek

that served as the farms intake creek and did not receive effluent. Physical/chemical

parameters identified distinct changes in the receiving creek with respect to farm

operations. Elevated water column NH4+ (18.5 ± 8.0 µM) and chlorophyll a

concentrations (5.5 ± 1.9 µg/L) were measured when the farm was in operation, in

contrast to when the farm was inactive (1.3 ± 0.3 µM & 1.2 ± 0.6 µg/L, respectively).

At all times, physical/chemical parameters at the mouth of the effluent creek, were

equivalent to control values, indicating effluent was contained within the effluent-

receiving creek. However, elevated δ15N signatures of mangroves (up to ~8 ‰) and

macroalgae (up to ~5 ‰) indicated a broader influence of shrimp farm effluent,

extending to the lower regions of the farms intake creek. Elevated concentrations of

chl a (9.6 µg/L) when the farm was in operation were measured at upstream sites

close to the location of farm effluent discharge. Bioassays indicated that

phytoplankton at these sites did not respond to further nutrient additions, however

downstream sites showed large growth responses. This suggested that further nutrient

loading from the shrimp farm, resulting in greater nutrient dispersal, will increase the

extent of phytoplankton blooms downstream from the site of effluent discharge. When

shrimp ponds were empty water quality in the effluent and intake creeks was

comparable. This indicated that observed elevated nutrient and phytoplankton

concentrations were directly attributable to farm operations.

68 Assessing Influences of Shrimp Farm Effluent

4.1 Introduction

Global demand for seafood continues to rise despite most wild fisheries being at their

maximum level of sustainable exploitation (Moriarty 1997). Predictions are that by

2030 more than 50% of fisheries production will need to come from aquaculture due

to human population growth, continuing demand for seafood, and static or declining

fish harvests (FAO 2000). Shrimp farming complements existing wild fisheries

operations, reduces demand on wild fisheries stocks and avoids environmental

damage resulting from fishing practices. However, shrimp farming has the potential

to adversely affect receiving water quality if not planned and managed appropriately

(Lin 1989, Naylor et al 1998, Naylor et al 2000).

A major problem experienced with shrimp farms, subsequent to issues surrounding

their initial construction, is the discharge of pond waters. Approximately 80% of

nitrogen added to ponds as shrimp feed is not retained as shrimp biomass (Briggs &

Funge 1994, Jackson et al In prep). Remaining nitrogen acts to fuel plankton and

microbial production within ponds, often resulting in negative effects on pond water

and sediment quality such as anoxia, nutrient toxicity, and blooms of undesirable algal

species (Burford & Glibert 1999, Moriarty 1997). Subsequently, pond operators

periodically flush ponds to minimise these effects, discharging untreated pond waters

(containing 27-57% of initial added nitrogen), into nearby rivers, creeks and estuaries

(Funge-Smith & Briggs 1998, Jackson et al In prep, Preston et al 2000), potentially

leading to deleterious effects in receiving waters such as eutrophication (Naylor et al

1998, Sansanyuth et al 1996). Often these same waters serve as intake (influent)

water for the same or neighbouring shrimp farms. The process of water reuse acts to

Chapter 4 69

further degrade water quality and provides a mechanism to spread water borne disease

agents (e.g. viral pathogens) from farm to farm or pond to pond (Pruder 1992).

Research on the influence of shrimp farm effluent on the receiving environment is

limited and data often confounded by other influences such as sewage, agricultural,

urban and/or industrial inputs (Grant et al 1995, Jones et al 2001). Studies using

physical/chemical parameters have identified that symptoms of aquaculture effluent

(e.g. elevated nutrient and chlorophyll concentrations) are only measurable in close

proximity to the discharge point (Hensey 1991, Samocha & Lawrence 1997), when in

fact the biological influence of shrimp effluent can extend further (Jones et al 2001).

In addition, the spatial and temporal variability associated with physical/chemical

techniques can make assessments of impacts unreliable (Wolanski et al 2000).

Subsequently, there is a need for a more comprehensive approach to assessing the

influence of shrimp farm effluent on the surrounding environment, with ability to

predict changes in the environment following further changes in farm management

practices. In order to achieve this goal, a broader range of indicators are required.

These indicators should be relevant to shrimp farm effluent and integrate temporal

variations experienced in shrimp farm receiving waters.

Therefore it was the aim of this study to augment typical physical/chemical water

quality monitoring techniques with biological ind icators. This was assessed in

receiving waters from a shrimp farm located on the north-east coast of Australia. The

location of this shrimp farm was advantageous as the farms receiving waters were not

influenced by other anthropogenic activities. This allowed responses in the biological

indicators to be specifically related to the shrimp farm effluent.


4.2 Materials and Methods

4.2.1 Site Location

This study was conducted in the surrounding waters of a shrimp farm located on the

north-east coast of Aus tralia (18o20’S; 146o4’E). This farm is located on the edge of a

world heritage area “Hinchinbrook Island”, and consequently there were little other

anthropogenic influences to interfere with indicator responses. The farm is positioned

between two natural mangrove lined tidal-creeks; one that is used for influent pond

water, the other for discharging effluent into (Plate 4.1). This enabled the influent

creek to be used as the control creek in relation to the effluent creek. The two creeks

differed in size, the influent creek being deeper (2-3m MHW), shorter (~1.5 km from

the mouth to the farm) and wider (10m) than the effluent creek (1-2m deep MHW,

~3km long and 4-6m wide). These two creeks discharge into Hinchinbrook Channel

that is approximately 10km in width, the eastern side bounded by Hinchinbrook

Island. Tidal range in this region was approximately 2m.

4.2.2 Sampling regime

Three sites were positioned in the intake creek and 4 in the effluent creek (Figure 4.1).

Sites were also positioned outside these creeks in the channel in order to detect

effluent effects outside the creeks. There were three sampling periods at different

stages of shrimp-pond maturity; empty ponds with no shrimp, full ponds with juvenile

shrimp, and during pond harvest where whole ponds were being emptied daily. All

sampling occurred near the top of the flood tide.

Chapter 4 71

Plate 4.1 Aerial photo of the studied shrimp farm with the influent creek highlighted in blue and the effluent creek in red. The dotted lines show the continuation of the natural creeks.

0 5 10 15 20 25 30

Kilometers

0 5 10 15 20 25 30

Kilometers

Figure 4.1 Site map showing locations of sites within the influent and effluent creeks and sites

located in Hinchinbrook Channel.


4.2.3 Physical/chemical parameters

Physical parameters measured in this study were total suspended solid concentrations,

secchi depth, salinity, chlorophyll a (chl a), dissolved oxygen, pH and temperature.

Salinity, pH and temperature were measured in the field using a Horiba Water

Checker Model U-10. Light attenuation was determined using a 30 cm diameter

secchi disk. Duplicate water samples for chlorophyll a analysis were collected from

approximately 0.1m below the surface at each site, and filtered immediately

(Whatman GF/F filters, nominal pore size 0.7µm) then stored over dry ice for analysis

on return to the laboratory (Parsons et al 1984). Duplicate 2 L samples of water from

each site were stored in rinsed plastic containers until return to the field laboratory

where samples were filtered onto pre-weighed Whatman GF/F filters. Filters were

then rinsed twice with deionised water to remove salts and then re-weighed. The

difference in filter weight before and after filtration was equal to the concentration of

suspended solids greater than 0.7µm.

Nutrient samples (dissolved and total) were collected from approximately 0.1m below

the water surface. Dissolved nutrient samples were filtered in the field to remove

particulate matter using a 60ml syringe and disposable filters (Sartorius Minisart 0.45

µm). Total nutrients were collected using a 60ml syringe without a filter in order to

obtain a whole water sample. Collected samples were stored in polyethylene 100ml

bottles. Filtered and unfiltered water samples were frozen immediately on dry ice and

transported to the laboratory where they were analysed colourimetrically using auto-

analyser chemical techniques (Clesceri et al 1998).

Chapter 4 73

4.2.4 Biological Indicators

4.2.4.1Nitrogen Stable Isotope Mapping

Mangrove leaves (Rhizophora stylosa) were collected at each site where available. In

order to standardise collection, the second youngest leaf of a rosette was selected from

5 different trees at each site. This age leaf was found to be physically mature (in

comparison to the youngest leaf) and continuously free from insectivorous galls or

fungal infection. In order to overcome restricted distributional ranges and assess

water column δ15N, red macroalgae (Catenella nipae) was incubated at each site using

the methods of Costanzo et al (2001). Three samples of the macroalgae, collected

from a low nutrient environment (δ15N ~2‰), were deployed at each site for a period

of four days. At each site, macroalgae were housed in transparent, perforated

chambers and suspended in the water column for 4 days at ~50% light (half secchi

depth) (Costanzo et al 2001).

Following collection, plant material was composited, oven dried, ground to a fine

powder using a ball-mill grinder and then oxidised in a Roboprep CN Biological

Sample Converter (Dumas Combustion). The resultant N2 was analysed by a

continuous flow isotope ratio mass spectrometer (Europa Tracermass, Crewe, U. K.)

4.2.4.2 Phytoplankton Bioassays

Seven day bioassays were employed in this study, incubating samples of water from

each study site with added nutrients. Water samples were collected in 30L black barrels

(acid washed) and transported back to an outdoor incubation facility. Water (4L) from

each site was filtered through a 200µm mesh (to screen out large zooplankton) into

sealed transparent 6L containers.


The response of phytoplankton to nutrient additions was determined by measuring

biomass increases (via fluorescence measurements) relative to a control sample to which

no nutrients were added. Nutrients added were nitrate - NaNO3 (200µM), ammonium-

NH4Cl (30µM), phosphate - Na2HPO4.H2O (20µM), silica - NaSiO3 (66 µM) and a

treatment receiving all the above nutrients. These concentrations were chosen to ensure

saturation of phytoplankton response by each particular nutrient (O'Donohue et al 1997).

Control and nutrient treatments were designed such that the potential of both light and

nutrients to stimulate phytoplankton growth could be determined. The control bioassay

treatment functions as a light response treatment as suspended solids in the collected

water settle during the 7 day incubation allowing light penetration.

Containers were then incubated at ambient water column temperature (± 2oC) and 50%

light intensity in outdoor flow-through aquaria. The response of phytoplankton was

monitored daily by withdrawing 30mL aliquots from each container, dark adapting

aliquots for 15 min then measuring fluorescence with a fluorometer (Turner Designs

TD700) under UV-excitation. In vivo fluorescence data from each treatment and site

was plotted daily over the course of the incubation. To quantify the response-potential

of phytoplankton to simulated light and nutrient additions, a Phytoplankton Response

Index (PRI) was used (as described in Chapter 3) that incorporated initial fluorescence,

maximum response, and time to maximum response. A light stimulated response was

inferred from growth of phytoplankton in the control bag where no nutrients were

added. Light Phytoplankton Response Index (LPRI) was derived from the following

equation:


Chapter 4 75



Nutrient stimulated responses were inferred growth of phytoplankton above that of the

control phytoplankton growth. Nutrient Phytoplankton Response Index (NPRI) was




FN and TN is time to maximum nutrient fluorescence. Units for PRI were fsu.d-1

4.2.5 Statistical Analyses

Differences between sites and surveys were tested for significance using a two-factor,

with replication ANOVA. Data closely followed a normal distribution according to

the ‘Studentised Range Test’ (Madansky 1988).

4.3 Results

In the sections below, the physical/chemical data is first examined from the receiving

waters of a tropical shrimp farm located in north-east Australia. The nitrogen isotope

values of mangroves and deployed macroalgae are then examined in order to assess

the ability of each technique to measure the influence of nutrients from shrimp farm

effluent. Finally examined are the phytoplankton bioassay data and phytoplankton

response index as a tool for assessing bloom potential of phytoplankton in effluent

receiving waters. The three surveys of the region are referred to as ‘empty’, ‘full’ and


harvest’ relating to the status of the shrimp ponds and correspond to the maturity of

the prawns (none, juveniles and adults, respectively).

4.3.1 Physical / Chemical Properties of Receiving Waters

Strong contrasts in physical/chemical variables were evident between the influent and

effluent creeks and between empty, full and harvest pond stages (Table 4.1). The

effluent creek displayed poorest water quality when the ponds were full; best water

quality when the ponds were empty. All physical/chemical variables continually

approached control values at the mouth of the effluent creek in all surveys, indicating

that the influence of shrimp effluent was contained within the effluent creek. This

generic trend is demonstrated here by transect data of dissolved inorganic nitrogen

(DIN) and total nitrogen (TN) concentrations (Figure 4.2).

Table 4.1 Mean water column physical/chemical parameters of the influent and effluent creeks associated with the shrimp farm. Data represent surveys conducted over three stages of shrimp -pond maturity – empty, full and harvest. (TSS – Total Suspended Solids; FRP – Filterable Reactive Phosphorus; TN/TP – Total Nitrogen / Total Phosphorus).

Nutrient concentrations were considerably variable in the effluent creek over the three

surveys, whereas concentrations in the influent creek were relatively uniform (Table

Chapter 4 77

4.1). Highest nutrient concentrations were recorded in the effluent creek when the

ponds were full, standard errors indicating the large range of values recorded between

sites within the creek. NH4+ was the dominant nutrient species measured in the

effluent creek when the ponds were in operation. Mean NH4+ concentrations in the

effluent creek when the ponds were full (18.5µM) were ~20 times concentrations

measured in the influent creek (0.9µM). During pond harvest, NH4 concentrations

were still elevated in the effluent creek (5.3µM) but not to the extent of when the

ponds were full. Filterable reactive phosphorus (FRP) was at or below detection

limits in both creeks during each of the three surveys (≤ 0.1µM). DIN:DIP ratios

were similar in both influent and effluent creeks (44 and 41, respectively) when the

ponds were empty. However, large shifts in the ratio occurred in the effluent creek

when the ponds were full and being harvested (214 and 152, respectively) in

comparison to smaller shifts in the influent creek during these periods (14 and 31,

respectively).

Salinity measurements highlighted the freshwater influence of heavy rainfall when the

ponds were empty, with mean salinities in the effluent creek of 10.3‰ as opposed to

mean salinities of 25.4‰ and 20.1‰ when the ponds were full and being harvested,

respectively. Salinities were always higher in the influent creek than in the effluent

creek, most likely due to a greater tidal excursion and the shorter distance surveyed in

the influent creek. Chlorophyll a values were similar in the influent and effluent

creeks when the ponds were empty (1.4 & 1.2 µg/L, respectively), though higher in

the effluent creek when the ponds were full and harvesting (5.5 & 2.5µg/L,

respectively) than in the influent creek (1.0 & 0.5 µg/, respectively).


0

40

80

120

TN

(µM

)

1 2 3 4 5 6 7Site

0

10

20

30

40

DIN

(µM

)

FullEmpty

Harvest

0

40

80

120

TN

(µM

)

1 2 3 4 5 6 7Site

0

10

20

30

40

DIN

(µM

)

FullEmpty

Harvest

0

10

20

30

40

DIN

(µM

)

FullEmpty

Harvest

Figure 4.2 Transect data of a) dissolved and b) total nitrogen concentrations along the influent creek (sites 1-3) and the effluent creek (sites 4-7) for each stage of shrimp farm maturity – empty, full and harvest.


4.3.2.1 Vegetation δ15N Mapping

δ15N signatures of mangrove leaf tissue and deployed macroalgae varied in response

to the status of the shrimp ponds and with distance away from the site of pond effluent

discharge. Mean δ15N values for mangroves were typically higher in the effluent

creek than values measured in the influent creek (P = 0.054) (Figure 4.3 a; Table 4.2).

Mangrove δ15N signatures decreased with distance downstream in the effluent creek

from a mean value of 7.3 ‰ upstream to 3.6 ‰ at the mouth (p<0.05) (Figure 4.3 b).

A reverse trend was observed in the influent creek with highest signatures measured

Chapter 4 79

in mangroves at the mouth (3.7‰) and lower values measured upstream towards the

farm water intake (2.2‰).

0

2

4

6

8

Man

grov

e δ15

N (‰

)

a

Empty Full Harvest

Influent Creek

Effluent Creek

Influent Creek

Effluent Creek

Man

grov

e δ15

N (‰

)

b

2

4

6

8

10

1 2 3 4 5 6 7

Site Cha

nnel

Figure 4.3 Mean δ15N values for mangroves sampled in the influent and effluent creeks and Hinchinbrook Channel for each: a) survey, and b) site. Bars indicate standard error.

δ15N signatures of deployed macroalgae, following the 4-day incubation, displayed

similar trends as mangroves of elevated δ15N signatures in the effluent creek

compared to the influent creek and Hinchinbrook Channel (p<0.001), however only

when the ponds were full and being harvested (Figure 4.4 a; Table 4.2). When the

ponds were empty there was no difference in macroalgae δ15N between the influent

and effluent creeks and Hinchinbrook Channel (p>0.05). Initial δ15N values for

macroalgae were below 3‰ (not shown), indicating an increase at all sites in the


effluent creek during periods when the ponds were full and harvesting. Sites in the

channel generally displayed values equivalent to values recorded in the influent creek.

Elevated macroalgae δ15N signatures (4.2 - 4.3‰) were observed throughout the

effluent creek relative to signatures recorded in the channel (~3‰) (Figure 4.4 b). As

recognised in the mangrove δ15N data, an elevated macroalgal signature was

measured at the mouth of the influent creek (3.4‰) with values decreasing upstream

towards the farm intake (2.9‰) (Figure 4.4 b)

0

2

4

6

8

Mac

roal

gal δ

15N

(‰)

Empty Full Harvest

a

Mac

roal

gal

δ15N

(‰)

2

4

6

8

10 b

Cha

nnel1 2 3 4 5 6 7

Site

Influent Creek

Effluent Creek

Channel

Figure 4.4 Mean δ15N values for macroalgae sampled in the influent and effluent creeks and Hinchinbrook Channel for each: a) survey, and b) site. Bars indicate standard error.

Chapter 4 81

Table 4.2 Significance values for comparisons between surveys (empty, full and harvest), locations (influent and effluent creeks) and interactions between surveys and locations for δ15N of mangroves and macroalgae.

df MS P-value

Mangrove δ15N Survey 2 1.56 0.670

Location 1 16.80 0.054Survey x Location 2 1.50 0.068

Error 14 3.80

Macroalgal δ15N Survey 2 1.25 <0.05Location 1 7.41 <0.001

Survey x Location 2 1.76 <0.01Error 13 0.24

FactorParameter


Response indices of phytoplankton bioassays to N, P, Si and +All additions indicated

a predominant N+P limitation at most sites over the three surveys (Table 4.3). Light

responses (established from responses in the control bag) were recorded in the

effluent creek when the ponds were full and being harvested (12.0 and 10 fsu.d-1,

respectively). Additions of nitrogen did not result in strong responses in either creek

where indices ranged from 1.0 – 5.0 fsu.d-1. An elevated phosphorus response was

recorded in the effluent creek during the ‘harvest’ survey (23 fsu.d-1) indicative of P

limitation at this time. Silica additions did not stimulate phytoplankton in either

survey, phytoplankton response indices ranging from 0.0 – 1.0 fsu.d-1. N+P additions

resulted in elevated phytoplankton responses in each creek over all surveys except for

the influent creek when the ponds were full (Table 4.3).


Table 4.3 Mean Phytoplankton Response Indices (PRI) following nutrient additions for the influent and effluent creeks associated with a shrimp farm. Sampling was conducted during 3 stages of shrimp -pond maturity – empty, full and harvest.

Highest initial fluorescence (Fo) values (proxy for chlorophyll a) were recorded in the

effluent creek when the ponds were full, followed by when the ponds were being

harvested and empty (Figure 4.5 a). Values decreased downstream in the effluent

creek reaching values equivalent to those measured in the influent creek. Low Fo

readings were recorded at all sites in the influent creek during all three surveys.

As ‘+ALL’ nutrients generally caused the greatest phytoplankton response in each

creek over the three surveys, transect data relating PRI trends for the ‘+ALL’

treatments are represented in Figure 4.5 b & 4.5 c. Phytoplankton responses to

nutrients were evident in the influent creek when the ponds were empty and

harvesting, with minimal responses when the ponds were full. The effluent creek

displayed a strong transition between high and low phytoplankton responses (Figure

4.5 b), indicating regions where added nutrients did not stimulate phytoplankton

growth despite high initial phytoplankton biomass as indicated by Fo (Figure 4.5 a).

Correspondingly, in regions downstream where chlorophyll a decreased, nutrient

PRI’s increased. PRI’s at the creek mouths were comparable for each season.

Chapter 4 83

Elevated responses of phytoplankton in the control treatment (response to light) were

restricted to the upper and middle reaches of the effluent creek during each survey

(Figure 4.5 c). Minimal light responses were recorded in the influent creek when the

ponds were empty and during harvest, with no light response when the ponds were

full.

0

Nut

rient

PR

(fs

u.d-1

)

0

20

40

60

20

40

60

Lig

ht P

RI (

fsu.

d-1 )

FullHarvest

Empty

0

Initi

al F

luor

esce

nce

(fsu

)

1 2 5 6 73 4

Site

a

c

b

20

40

60

20

40

60

20

40

60

20

40

60

Figure 4.5 Phytoplankton bioassay data along the influent creek (sites 1-3) and the effluent

creek (sites 4-7) for each stage of shrimp -pond maturity – empty, full and harvest. a) ambient fluorescence values of creek water at the time of bioassay collection (proxy for chlorophyll a concentrations); b) nutrient (N+P) PRI’s along each creek for each pond stage; and c) light PRI’s along each creek for each pond stage.


4.4 Discussion

This study showed that there were measurable changes in physical, chemical and

biological parameters of receiving waters as a result of effluent discharge from a

shrimp farm. Each parameter highlighted the variability experienced in the receiving

environment at different pond stages. The bioindicators employed in this study each

served a purpose in establishing an understanding of various influences on the

environment. It is evident that farm waste is moving outside the effluent creek and

influencing sites within the influent creek, thereby acting to potentially recycle

effluent back into the shrimp farms.

4.4.1 Water Column Physical/Chemical Parameters

Physical/chemical parameters indicated poor water quality (i.e. high NH4+ and

chlorophyll concentrations) in the effluent creek when the ponds were in full

production, temperatures were highest and flow conditions minimal (as indicated by

elevated salinities). Effluent discharge, when ponds are full, can still be quite

significant as farmers regularly flush ponds to control excessive nutrient and/or

phytoplankton concentrations (Jackson et al In prep), particularly towards the latter

half of shrimp production. Up to 5-10% of pond water is exchanged per day in

intensive shrimp ponds (Funge-Smith & Briggs 1998). Highest salinities within the

effluent creek during the time when ponds were full, suggest longer residence times

which would facilitate settlement of particulate material, including phytoplankton,

leading to anoxia/hypoxia of bottom waters (Boyd 1995). In turn, this can lead to

mineralisation of settled organic matter resulting in ammonium fluxes to the overlying

water column, which further serves to stimulate phytoplankton (Jeffrey et al 1995,

Jensen et al 1990, Lavery & McComb 1991). This would explain, in part, the

Chapter 4 85

elevated ammonium concentrations measured in the effluent creek when the ponds

were full.

All physical/chemical variables indicated that water quality increased downstream

reaching values equivalent to those measured in the influent/control creek. Based on

this data, it would suggest that the influence of effluent is contained within the creek.

The physical/chemical data discussed in this study was only a snapshot in time of the

highly variable water quality that likely occurs in these tidal creeks (Wolanski et al

2000). It was not logistically possible to assess temporal variability of

physical/chemical variables as boat access to these shallow creeks was limited to the

flood tide. This further emphasised the need for time- integrated parameters and hence

the use of biological indicators in this study.


4.4.2.1 Vegetation δ15N Mapping

Elevated δ15N signatures of mangroves and deployed macroalgae indicated that pond

effluent was influencing biotic process in the effluent creek and the lower portions of

the influent creek. This indicated that hydrodynamic forces are distributing effluent

into the influent creek, a scenario that was not identified with physical/chemical

parameters.

A close association between incubated macroalgae and the water column was evident

in this study. Unlike the mangroves, macroalgal δ15N signatures in the effluent creek

changed significantly between when the ponds were empty and when they were full

and being harvested. Therefore, incubated macroalgae in the water column provided


information on the current status of water quality in the effluent creek, in addition to

providing information on areas outside the creeks in Hinchinbrook Channel where

mangroves were not present.

The slow growth of mangroves, in comparison to most other marine plants, provides a

useful bioassay for assessing long term environmental variations in water and

sediment quality. Rhizophora stylosa growing in this region has leaf longevities of up

to 1-2 years (Duke et al 1984). Also, mangroves utilise nutrients from interstitial

pore-water within the sediment, not directly from the water column. Therefore,

interpretation of mangroves as biological indicators is different in comparison to algae

and phytoplankton which have quicker biomass turnover times and use nutrients

directly from the water column (Campbell 1996).

Mangroves typically had higher mean δ15N signatures than the deployed macroalgae.

Slow plant growth of mangroves and high ammonium concentrations should favour

large fractionations against 15N, resulting in lower mangrove δ15N signatures than the

source (Fry et al 2000). Shrimp farm effluent has been found to have δ15N signatures

of ~6‰ (Preston, unpubl. data), as opposed to sewage effluent ~10‰, likely due to

depressed nitrification, denitrification and NH4+ volatilization in the ponds (Alongi

1999; Alongi 2000), which is reflected in this study by the δ15N signatures of

macroalgae in the effluent creek which approach this value, and which have been

found to typically reflect the source δ15N signature (Costanzo et al 2001). This would

suggest that the source of nitrogen that the mangroves are accessing have even higher

δ15N values than the mangroves reflect. This is possible as microbial processing of

nitrogen in mangrove sediments can be quite high, microbes preferring the lighter 14N

Chapter 4 87

for metabolism, leading to enrichment of 15N in the free nitrogen pool (Mariotti et al

1988). Alternatively, 15N enrichment may result from volatilisation (Heaton 1986). It

is evident in these tidal creeks that water can stand for long periods of time,

particularly within the mangrove forest, where water can pool during the low tide

resulting in volatilisation of 15N-depleted nitrogen into the atmosphere, leaving 15N-

enriched nitrogen in the sediment pool.


Phytoplankton response indices proved useful in this study for characterising limiting

factors for phytoplankton growth and regions that were susceptible to blooms

following nutrient additions. Largest responses were generally to ‘+ALL’ added

nutrients indicating nutrient co- limitation. An artefact of the technique is that

particulates, generally kept in resuspension in situ, settle out in the bioassay

containers allowing light penetration. If ambient nutrients are sufficient,

phytoplankton will subsequently increase their biomass (Chapter 3). Therefore, light

limitation was inferred by a response in the control treatment. Light responses were

detected in the effluent creek when the ponds were full and harvesting as elevated

nutrients were present to allow growth.

Phytoplankton responses to nitrogen in the effluent creek when the ponds were full

and being harvested were minimal as ambient nitrogen concentrations were present in

high concentrations. The large phosphorus response in the effluent creek when the

ponds were being harvested indicated a strong P limitation during this time as also

indicated by the large DIN:DIP ratio of 152:1. However, the DIN:DIP ratios

predicted P limitation in the effluent creek when the ponds were empty and full also,


though this was not evident. This would suggest that other factors were inhibiting

growth such as a lack of micro-nutrients, temperature, salinity or competition for

nutrients with bacteria (Fukami et al 1997)

Variations in phytoplankton response indices were apparent within each creek though

differed between surveys. There appeared to be a ‘trade-off’ between initial standing

stock of phytoplankton and the ability to respond to nutrients. This was particularly

apparent at the upper site in the effluent creek, where there was no response to

nutrients, despite high initial phytoplankton biomass. This region was generally

where highest nutrient and suspended solid concentrations were found. Subsequently,

added nutrients did not influence growth any further. Light did however influence

growth in this upper region as ambient nutrients were sufficient as indicated by

elevated light PRI’s. However with distance downstream, ambient phytoplankton

biomass and nutrient concentrations decreased, primarily due to flushing and biotic

uptake, resulting in increased bioassay sensitivity to added nutrients.

4.5 Conclusions

It was evident that the receiving environment in this study had been influenced as a

result of waste from a shrimp farm. Using a combination of indicators acted to

increase the resolution of understanding of how pollutants were affecting the study

region. It was apparent that the influence of shrimp effluent on receiving waters was

variable and related to shrimp-pond maturity, suggesting that impact monitoring

should be timed accordingly in order to fully assess pollutant distribution and effects.

Physical/chemical parameters indicated poor water quality was contained within the

effluent creek, whereas bioindicators indicated a broader influence of effluent

Chapter 4 89

extending to the influent creek, suggesting potential recycling of effluent back into the

farm. This has serious implications towards management decisions in the region as

effluent recycling in shrimp ponds is potentially hazardous to pond operations

(Gowen et al 1990). Phytoplankton bioassays identified regions where additional

nutrients are likely to be detrimental - information that is essential when planning

future farm expansions and elevated discharge loads. Alternatively, there are options

for reducing the current impact of shrimp farm effluent on the environment. These

include a range of effluent treatment options aimed at reducing nutrient, chlorophyll

and sediment concentrations and improved farm management practices that reduce

waste concentrations and thereby reduce costs associated with lost feeds (Samocha &

Lawrence 1997, Sansanyuth et al 1996, Teichert et al 1999, Troell et al 1999).

CChhaapptteerr 55 Assessing the Seasonal Influence of Sewage

and Agricultural Nutrient Inputs in a Sub-Tropical River-Estuary

Abstract A combination of physical/chemical measurements and biological indicators

identified nutrient impacts throughout an Aus tralian sub-tropical river estuary. This

was a balance of sewage inputs in the lower river and agricultural inputs in the mid-

upper river, the combined influence being greater in the wet season. Field sampling

in the region was conducted at 6 sites within the river, over 5 surveys to encapsulate

both wet and dry seasonal effects. Parameters assessed were tissue nitrogen contents

and δ15N signatures of mangroves and macroalgae, phytoplankton nutrient addition

bioassays and standard physical/chemical variables. Strong spatial (within river) and

temporal (seasonal) variability was observed in all parameters. Poorest water quality

was detected in the mid-upper (agricultural) region of the river in the wet season,

attributable to large diffuse inputs in this region. Water quality towards the river

mouth remained constant irrespective of season due to strong oceanic flushing.

Mangrove and macroalgal tissue δ15N and %N proved a successful combination for

discerning sewage and agricultural inputs. Elevated δ15N and %N represented sewage

inputs, whereas low δ15N and elevated %N was indicative of agricultural inputs.

Phytoplankton bioassays found the system to be primarily responsive to nutrient

additions in the warmer wet season, with negligible responses observed in the cooler

dry season. These results indicate that the Tweed River is sensitive to the different

anthropogenic activities in its catchment and that each activity has a unique influence

on receiving water quality. Subsequently, nutrient influences could not be assessed

with a single indicator and required a combined approach in order to better

characterise nutrient influences in the system.

92 Nutrient Influences in a Sub-Tropical River Estuary

5.1 Introduction

Increased acceptance of relationships between land use, nutrient dynamics and biotic

responses, has resulted in a move towards a greater understanding of the relative

influences of land-based activities on adjacent waterways. Agricultural development,

together with increased urbanisation of coastal zone land, has lead to increased

nutrient loads entering estuarine and coastal waters in recent decades (Cosser 1997,

Nixon 1995). This corresponds with an increase in eutrophication events at many

sites along the Australian coastline (Dennison & Abal 1999, Heggie et al 1999,

Lavery et al 1991, Longmore et al 1999). Land uses are often diverse along

waterways making connections between water quality and land use activities

particularly difficult. Confounding this difficulty, in tropical/subtropical systems, is

seasonally driven spatial and temporal variability in water quality (Eyre 1998).

Subsequently, there is a requirement for indicators that are both specific to individual

nutrient sources and integrate external environmental parameters over time.

Biological indicators show merit in achieving these goals and provide information on

the biotic response to environmental influences. The use of nitrogen stable isotopes

of marine flora and phytoplankton bioassays, have respectively proved successful

techniques for tracing nitrogen in marine environments (Cabana & Rasmussen 1994,

Costanzo et al 2001, Hobbie & Fry 1990, McClelland et al 1997, Wainright et al

1998) and testing the responsiveness of these environments to nutrient additions

(Bonetto et al 1991, Cowell & Dawes 1991, Fisher et al 1999, Smith & Hitchcock

1994).

Chapter 5 93

Tracing nitrogen sources in the marine environment using stable isotope analysis of

marine plants, is based upon isotopic fractionations following biotic and industrial

nitrogen processing and resulting transformations of the nitrogen species. Although

nitrogen stable isotopes (14N and 15N) occur on earth in a fixed proportion of

approximately 273 14N atoms for each 15N atom, the ratio of 15N to 14N however,

differs among specific N pools in the environment (Peterson & Fry 1987). This ratio,

when normalised relative the 15N/14N of atmospheric N2 (denoted as δ15N), is what

defines many natural and anthropogenic processes on earth, making N sources

identifiable and traceable within a system (Heaton 1986). For example, the δ15N

signature of sewage effluent (~10 ‰) distinguishes it from most agricultural fertilisers

(~0 ‰) (McClelland et al 1997).

Phytoplankton are useful as indicators of the responsiveness of a system to nutrient

additions, due to their ubiquitous distribution and their sensitivity to perturbations in

their environment (Valiela 1995). Interest in phytoplankton responses to added

nutrients is two-fold. Firstly, elevated phytoplankton populations are a common cause

of ecosystem deterioration resulting in anoxia of bottom waters and sediments, fish

kills, toxic microalgae and foul odours (Paerl & Bowles 1987). Secondly, assaying

the response of phytoplankton to elevated nutrient concentrations can provide signs of

impending phytoplankton blooms and thereby direct nutrient input diversions and

reductions (Dennison & Abal 1999).

The Tweed River Estuary, on the central east coast of Australia, is an example of a

sub-tropical estuarine system undergoing change due to a variety of intensive land

uses in the region. These land uses are dominated by agriculture in the central and


upper reaches of the catchment and urban development in the lower catchment.

Subsequently, there is a variety of point and diffuse nutrient sources including sewage

effluent discharges and large seasonal inputs from agricultural nutrient runoff

(NSWSPCC 1985). This provided a unique opportunity to assess the influence of

various nutrient inputs on water quality in the system, using physical/chemical

parameters, stable isotope analyses of marine flora and phytoplankton nutrient

addition bioassays. This was the first thorough study of spatial and temporal

variations of water quality and biotic responses conducted in the Tweed River.

Therefore, this study provided an opportunity to investigate the applicability of

bioindicators in this region and produce resource managers with criteria that will aid

in maintaining or improving the integrity of their managed environment.

5.2 Materials & Methods

5.2.1 Site Location

The Tweed River catchment (28o 10 S, 153o 33 E) is approximately 1100km2 and is

dominated by the Tweed River, with major tributaries including the Rous River and

the Terranora system of broad-waters (Plate 5.1). The Tweed River weaves a path

across a flood plain dominated by sugar cane farming in the middle-upper regions

(Plate 5.2 a) and urban development in the lower region. Tidal influence penetrates to

most of the system below the weir which is located approximately 40km upstream

(Plate 5.2 b). The mouth of the river has been stabilised and is routinely dredged for

navigational purposes (Plate 5.2 c).

Chapter 5 95

Plate 5.1 NOAA true colour satellite image of the Tweed River Estuary and Catchment,

surrounded by the Border ranges to the north and Mt Warning to the south. Approximately 49% of catchment area is rural as evident in this image. Residential development (pale white) can be seen towards the mouth and in the central region (Murwillimbah).

a b c

Plate 5.2 Aerial images of a) agricultural dominated mid-upper region of the Tweed River

showing cane drainage channel joining the river; b) weir dividing river at ~40km upstream from the mouth; and c) urban dominated lower region of the Tweed River with stabilised river mouth.


5.2.2 Sampling Strategy

This study comprised 5 surveys over 4 years aimed to encompass the seasonal

variability experienced in the region. These surveys included 3 wet and 2 dry season

surveys (Figure 5.1). Wet surveys corresponded with summer months (Dec-Feb) and

dry surveys with winter months (August). Average rainfall for the month preceding

wet samplings was ~10 mm and for dry samplings <2 mm. Included in these surveys

were measurements of standard physical/chemical parameters of the water column

and biological indicators (δ15N and tissue nitrogen content of mangroves and

incubated macroalgae; and phytoplankton nutrient bioassays). Six sites were chosen

within the Tweed River to encompass both urban and agricultural regions (Figure

5.2).

0

40

80

120

160

200

Jun

-97

Aug

-97

Oct

-97

Dec

-97

Feb

-98

Apr

-98

Jun

-98

Aug

-98

Oct

-98

Dec

-98

Feb

-00

Apr

-00

Jun

-98

Aug

-00

Oct

-00

Dec

-00

Feb

-01

Rai

nfal

l (m

m)

Month

Wet

‘97

Dry

‘98

Wet

‘00

Dry

‘00

Wet

‘01

0

40

80

120

160

200

Jun

-97

Aug

-97

Oct

-97

Dec

-97

Feb

-98

Apr

-98

Jun

-98

Aug

-98

Oct

-98

Dec

-98

Feb

-00

Apr

-00

Jun

-98

Aug

-00

Oct

-00

Dec

-00

Feb

-01

Rai

nfal

l (m

m)

Month

0

40

80

120

160

200

Jun

-97

Aug

-97

Oct

-97

Dec

-97

Feb

-98

Apr

-98

Jun

-98

Aug

-98

Oct

-98

Dec

-98

Feb

-00

Apr

-00

Jun

-98

Aug

-00

Oct

-00

Dec

-00

Feb

-01

0

40

80

120

160

200

Jun

-97

Aug

-97

Oct

-97

Dec

-97

Feb

-98

Apr

-98

Jun

-98

Aug

-98

Oct

-98

Dec

-98

Feb

-00

Apr

-00

Jun

-98

Aug

-00

Oct

-00

Dec

-00

Feb

-01

Rai

nfal

l (m

m)

Month

Wet

‘97

Dry

‘98

Wet

‘00

Dry

‘00

Wet

‘01

Figure 5.1 Average Tweed Heads and Murwillumbah daily rainfall data from June 1997 – January 1999 and from January 2000 – March 2001.

Chapter 5 97

0 3 6

KilometersKilometers

N

EW

S

N

EW

S

Queensland

New South Wales

Agriculture (sugar)

Urban

Undisturbed

Tweed River

Rous River

Terranora System

Sewage Outfall

Pacific Ocean

New South Wales

0 km

4.5 km

9 km

13.5 km18 km

25 km

Figure 5.2 Map of Tweed River system with sampling sites (solid circles) and distance up-river, sewage effluent outfalls (open squares) and major land uses (urban/agricultural).

5.2.3 Physical/chemical parameters

Physical aspects of water quality measured in this study included total suspended

solids (TSS), secchi depth, salinity, chlorophyll a, pH and temperature. Salinity, pH

and temperature were measured in the field using a Horiba Water Quality Checker

Model U-10. Samples for chl a and TSS were analysed on return in the laboratory.

Duplicate water samples for chlorophyll a analysis were collected from approximately

0.1m below the surface at each site, and filtered (Whatman GF/F filters) then stored

over dry ice. Duplicate 2 L samples of water from each site were stored in rinsed


plastic containers until return to the field laboratory where samples were filtered onto

pre-weighed Whatman GF/F filters for determination of TSS.

Nutrient samples (NH4+, NO3

- , PO4

3-, Total N and Total P) were collected from

approximately 0.1m below the water surface. For dissolved nutrients, samples were

filtered in the field to remove particulate matter using a 60ml syringe and Sartorius

Minisart 0.45 µm disposable glass fiber filters. Total nutrients were collected using a

60ml syringe without a filter in order to obtain a whole water sample. Collected

samples were stored in plastic 100ml bottles. Filtered and unfiltered water samples

were frozen immediately using dry ice and transported to the laboratory where they

were analysed using auto-analyser chemical techniques (Clesceri et al 1998).


5.2.4.1 Nitrogen Stable Isotope and %N Mapping

Mangrove leaves (Avicennia marina) were collected at each site where available.

Mangroves were not present at the mouth of the river as that region was heavily

modified for stabilisation of the river mouth. In order to standardise mangrove

collection, the second youngest leaf of a rosette was selected from 5 different trees at

each site. This age leaf was found to be physically mature (in comparison to the

youngest leaf) and continuously free from insectivorous galls or fungal infection. In

order to overcome restricted distributional ranges and assess water column nutrients, 3

samples of the red macroalgae, Catenella nipae, were deployed at each site for a

period of four days as prescribed by (Costanzo et al 2001). Catenella nipae was

collected from a low nutrient environment and had initial δ15N values <3‰.

Chapter 5 99

Following collection, plant material was composited, oven dried, ground to a fine

powder using a ball-mill grinder and samples oxidised in a Roboprep CN Biological

Sample Converter (Dumas Combustion). The resultant N2 was analysed by a

continuous flow isotope ratio mass spectrometer (Europa Tracermass, Crewe, U. K.)

for tissue nitrogen content (%N) and δ15N isotopic signatures.


Seven day bioassays were employed in this study, incubating 4 L samples of water from

each study site with added nutrients. Water samples were collected in 30L black barrels

(acid washed) and transported back to an outdoor incubation facility. Water from each

site was filtered through a 200µm mesh (to screen out larger zooplankton grazers) into

sealed transparent 6L plastic containers tied together.

The response of the phytoplankton was determined by measuring their biomass increases

(via fluorescence measurements) relative to a control sample to which no nutrients were

added. Nutrients were added to ensure saturation and included nitrate - NO3 (200µM),

ammonium - NH4 (30µM), phosphorus - PO4 (20µM), silica - NaSiO3 (66 µM) and a

treatment receiving all the above nutrients (O'Donohue et al 1997). Bioassay containers

were then incubated in outside incubation tanks (2m diameter, 0.5m deep) under 50%

ambient sunlight conditions (covered with a 50% shade cloth) with a freshwater flow

through system to maintain a steady water temperature (± 3oC of ambient). Control and

nutrient treatments were designed such that the potential of both light and nutrients to

stimulate phytoplankton growth could be determined. The control bioassay treatment


functions as a light response treatment as suspended solids in the collected water settle

during the 7 day incubation.

At identical daily circadian times, all bioassay containers were removed, gently shaken

and a 20ml sub-sample was poured into 30ml clear glass cuvette that were then dark-

adapted for a period of 30 minutes. Cuvettes were then read in a Turner TD1000 design

fluorometer. In vivo fluorescence data from each treatment and site was plotted daily

over the course of the incubation. To quantify the response-potential of phytoplankton

from each site to simulated nutrient additions, a Phytoplankton Response Index (PRI)

was utilised as described in chapter 3. Light Phytoplankton Response Index (LPRI)

was derived from the following equation:




Nutrient Phytoplankton Response Index (NPRI) was derived from the following

equation:



FN and TN is time to maximum nutrient fluorescence. Units for PRI were fsu.d-1.

5.2.5 Data Analyses

Given the co-variation and complexity of the physical, chemical and biological data,

multivariate techniques were used to assess patterns in all physical/chemical

biological variables (except for mangrove data due to lack of distribution at all sites).

Chapter 5 101

Data were range standardised and the Bray–Curtis association measure was used to

produce a distance matrix. Non-metric multi-dimensional scaling (non-metric MDS)

was used to produce an ordination of the distance matrix, using PATN (Belbin 1993).

Analysis was performed in two dimensions with 10 random starts. Data from all six

sites for all sampling periods were used in the non-metric MDS analysis. To assess

which of the indices were influencing patterns in the non-metric MDS, principal axis

correlation was carried out. Principal axis correlation determined the direction and

correlation coefficient of the best fit of each variable used in the ordination to the

summary variables in ordination space (Belbin 1993). Variables with a correlation

coefficient of 0.7 or higher were considered in this study as having a significant effect

upon the ordination.

5.3 Results

5.3.1 Physical / Chemical Indicators of Water Quality

Largest range in temperature was approximately 9oC between wet (summer) and dry

(winter) surveys (Figure 5.3 a). Salinity was representative of rainfall conditions at

the time of sampling, lowest salinities measured in the wet surveys and higher

salinities recorded in the dry surveys (Figure 5.3 b). Salinity at the river mouth was

equivalent to oceanic salinity (~35 ‰) irrespective of season assessed. Total

suspended solids (TSS) were typically higher in the wet surveys in comparison to dry

surveys (Figure 5.3 c). Lowest TSS concentrations were recorded at opposing ends of

the river (0 and 25 km), highest values recorded in the middle section of the river.

Chlorophyll a concent rations were lowest at sites towards the river mouth

(~1.5 µg/L), with highest values similar, though recorded at different locations, for

wet and dry surveys (8.3 µg/L at 18 km and 11.9 µg/L at 25 km, respectively) (Figure


5.3 d). pH displayed similar trends as salinity, with highest values recorded at the

mouth, values decreasing with distance up river and slightly lower in the wet season

(Figure 5.3 e).

Nutrient concentrations varied considerably between wet and dry surveys. Highest

nutrient concent rations were measured in the wet surveys, many values in the dry

surveys approaching or below analytical detection limits. Nitrate was the dominant

nutrient species within in the river, the highest mean value detected in the wet at 18

km (25.0 µM) (Figure 5.3 g). All dissolved nutrients decreased at 25 km to

concentrations comparable to those recorded at the river mouth, except phosphate

which remained elevated. Total nutrient concentrations were also higher in the wet

surveys than dry surveys, highest concentrations again recorded at 18 km in the wet

(Figure 5.3 i & 5.3 j). Total nutrients decreased at 25 km, though this was not as

apparent as the decrease at this site in dissolved nutrients.

MDS analysis of all physical/chemical variables clearly identified the disparity

between wet and dry surveys (Figure 5.4). Lower river sites, however, were similar

between seasonal surveys as represented by overlap of these sites in the MDS. With

distance upriver, seasonal separation of water quality parameters for each site became

apparent. Principal axis correlations identified the wet surveys to be strongly driven

by salinity and nutrients, whereas the dry surveys were strongly influenced by the

salinity and chlorophyll a. Positions of sites within the groupings were different

between wet and dry and surveys also. In the dry surveys, there was a clear

distinction between 0-9 km and 9-25 km, indicating that sites within these ‘sub-

groupings’ were relatively similar in terms of physical/chemical parameters. This

Chapter 5 103

strong separation of lower and upper river sites evident in the dry surveys, was not

apparent in the wet surveys.

10

15

20

25

30

Tem

p. (

o c)

05

1015

2025

TSS

(mg/

L)

PO

43-(µ

M)

0

10

20

30

40

NO

32-(µ

M)

0 10 20 30Distance Upriver (km)

0

20

40

60

80

TN

(µM

)

TP

(µM

)


Wet Dry

NH

4+(µ

M)

6

7

8

9

10

pH

a

e f

g

i

h

c

0

5

10

15

Chl

a(µ

g/L

) d

0

10

20

30

40

Salin

ity

(‰)

b

j

0

10

20

30

40

0

1

2

3

0

1

2

3

10

15

20

25

30

Tem

p. (

o c)

05

1015

2025

TSS

(mg/

L)

PO

43-(µ

M)

0

10

20

30

40

NO

32-(µ

M)


0

20

40

60

80

TN

(µM

)

TP

(µM

)


Wet Dry

NH

4+(µ

M)

6

7

8

9

10

pH

a

e f

g

i

h

c

0

5

10

15

Chl

a(µ

g/L

) d

0

5

10

15

Chl

a(µ

g/L

) d

0

10

20

30

40

Salin

ity

(‰)

b

0

10

20

30

40

Salin

ity

(‰)

b

j

0

10

20

30

40

0

10

20

30

40

0

1

2

3

0

1

2

3

0

1

2

3

0

1

2

3

Figure 5.3 Physical/chemical variables measured along the Tweed River between wet and dry seasons. Parameters include a) temperature, b) salinity, c) total suspended solids (TSS), d) chlorophyll a, e) pH, f) ammonium (NH4

+), g) nitrate (NO3-), h) phosphate

(PO43-), i) total nitrogen (TN) and j) total phosphorus (TP).


TN(0.86)

13.5

0

25

18

9

4.5

0

2518 13.5

94.5

0

25

18

94.5

0

25 18

13.5

9

4.5

25

18

13.5

9

4.5

0

Salinity(0.90)

pH(0.85)

Chl a(0.86)

TP(0.89)

Dry 2000

Dry 1998

Wet 2000

Wet 1997

Wet 2001

Dry

Wet

13.5

Figure 5.4 Two dimensional non-metric MDS on physical/chemical data from all sites and all

surveys. Broken lines group data from wet and dry seasons. Significant principal axis correlation vectors (with correlation coefficients in brackets) are shown, each point represents one site at the indicated sampling time. The stress level of 0.064 was acceptable (Clarke & Warwick 1994).

5.3.2 Biological Indicators of Water Quality

5.3.2.1 Macroalgae and Mangrove Indicators

δ15N signatures and tissue nitrogen contents (%N) of incubated macroalgae displayed

similar spatial trends in the lower half of the river, both indices increasing from the

river mouth (~ 3.5 ‰ and 1.6 %, respectively) and reaching a plateau at 9-13.5 km

(~4.9 ‰ and 1.9 %, respectively) (Figure 5.5 a). δ15N signatures then decreased at 18

km to values equivalent to those measured at the river mouth, whereas %N remained

elevated for the rest of the river. Mangroves displayed stronger spatial contrasts in

leaf δ15N signatures and tissue nitrogen contents than the incubated macroalgae

(Figure 5.6 a). Highest δ15N signatures were recorded in the urban region of the river

Chapter 5 105

with a peak at 9 km adjacent to the sewage outfall (~ 7 ‰). Lowest values were

recorded in the agricultural region of the river, values remaining relatively constant at

all sites in this region (~ 5 ‰). Mangrove leaf %N increased with distance upriver

from ~ 1.5% at 4.5 km in the urban region to 2.7% at 25km in the agricultural region.

%Nδ15N%Nδ15N

%Nδ15N

%N

Mac

roal

gae

δ15N

(‰)

2

3

4

5

6

1.0

1.4

1.8

2.2b

2

3

4

5

6

1.0

1.4

1.8

2.2

%N

Mac

roal

gae

δ15N

(‰

)

Initial δ15N

Initial %N

a

0 5 10 15 20 25 30

Distance Upriver (km)

Wet DrySeason

Urban AgriculturalUrban AgriculturalSTP

Figure 5.5 Mean macroalgae δ15N signatures and tissue nitrogen contents : a) along the Tweed River and b) between wet and dry seasons. Error bars represent standard error. Initial δ15N signatures and %N values of macroalgae prior to incubation are represented and the transition between urban and agricultural regions is highlighted in (a). STP represents location of sewage treatment plant outfall in the Tweed River.

Seasonal variation of δ15N and %N between wet and dry surveys was evident in the

incubated macroalgae but not in mangrove leaf tissue. Macroalgae displayed elevated

mean δ15N signatures and %N in wet season surveys in comparison to dry season


surveys (Figure 5.5 b). Mangroves did not display this seasonal variation in δ15N and

%N (Figure 5.6 b).

2

4

6

8

0 5 10 15 20 25 30


1

2

3

%N

Man

grov

e δ15

N (‰

)

Urban Agricultural

a

Wet Dry2

4

6

8

1

2

3

%N

Man

grov

e δ15

N (

‰)

Season

b%Nδ15N

%Nδ15N%Nδ15N

STP

Figure 5.6 Mean mangrove δ15N signatures and tissue nitrogen contents: a) along the Tweed River and b) between wet and dry seasons. Error bars represent standard error. Transition between urban and agricultural regions is highlighted in (a). STP represents location of sewage treatment plant outfall in the Tweed River.

Macroalgae %N and δ15N had significant correlations with water column DIN

concentrations (p<0.001) (Figure 5.7 a & b). Macroalgae %N displayed a positive

linear correlation with DIN, whereas macroalgae δ15N displayed a polynomial

relationship with DIN concentrations. δ15N values did not increase at DIN

concentrations greater than ~10 µM, and showed evidence of a decrease at the highest

DIN concentration (~45 µM). Below 10µM, there appeared to be a strong positive

Chapter 5 107

correlation between δ15Ν and DIN concentrations. Correlations between %N and

δ15N of mangroves did not correlate with water column DIN concentrations (R = 0.13

& 0.24, respectively).

0

1

2

3

0

2

4

6

0 10 20 30 40 50

R = 0.77

R = 0.59

Mac

roal

gae

%Ν

Mac

roal

gae

δ15N

(‰)

DIN (µM)

a

b

Figure 5.7 Correlations of (a) macroalgal tissue nitrogen content (%N) and (b) δ15N signatures with dissolved inorganic nitrogen (DIN) – all data included.

5.3.2.2 Phytoplankton Indicators

Phytoplankton bioassay responses indicated different regions of the river to be

responsive to nutrients and light (Figure 5.7 a). Largest mean phytoplankton

responses to light were in the middle region of the river and were approximately 2


times greater during wet than dry season surveys (Figure 5.7 b). Lowest light

responses were recorded at 0 km (river mouth) and 25 km. Phytoplankton responses

to added nutrients displayed a different trend than light responses and showed marked

differences between wet and dry surveys (Figure 5.7 a & b). Largest nutrient

responses were recorded at 9 km opposite the sewage outfall. Comparable nutrient

responses at sites above the sewage outfall were intermediate between the highest

responses at 9 km and responses recorded at the river mouth.

0

2

4

6

8

10

12

14

0 5 10 15 20 25 30


Lig

ht P

RI

(fsu

.d-1

)

0

20

40

60

80

Nut

rien

t PR

I (f

su.d

-1)

Light PRI Nutrient PRI

a

0

3

6

9

Wet DrySeason

0

10

20

30

40

50Nutrient PRILight PRI

Lig

ht P

RI

(fsu

.d-1

)

Nut

rien

t PR

I (f

su.d

-1)

b

Figure 5.8 Mean phytoplankton light and nutrient responses (PRI) represented a) spatially and b)

temporally in the Tweed River.

5.3.3 Analysis of Bioindicator Data

Multi-dimensional scaling was employed to visualise variability between seasons and

sites based on bioindicator data (Figure 5.9), as performed for the water quality data.

Again, clear distinctions were identified between wet and dry surveys with a region of

Chapter 5 109

overlap between seasons, however the stress level of 0.31 indicates the strength of this

ordination in two dimensions is relatively poor, suggesting caution in interpretation

(Clarke & Warwick 1994). Unlike the physical/chemical MDS, the region of overlap

includes not only lower regions sites, but also the upper region site (25 km),

highlighting the similarity between the end member sites in terms of biotic responses.

Biotic variables which explain most of this pattern were light phytoplankton response

indices (LPRI) and tissue nitrogen contents (%N) of macroalgae.

9

0

25

1813.5

9

4.5

25

18

13.5

94.5

4.5 25

1813.5

94.50

2518

13.5

13.5

4.5

0

25

1813.5

4.5

0

%N(0.80)

LPRI(0.72)

Dry

Wet

Dry 2000

Dry1998

Wet 2000

Wet 1997

Wet 2001

Dry 2000

Dry1998

Wet 2000

Wet 1997

Wet 2001

Figure 5.9 Two dimensional non-metric MDS on bioindicator data from all sites and all surveys. Broken lines group data from wet and dry seasons. Significant principal axis correlation vectors (with correlation coefficients in brackets) are shown, each point represents one site at the indicated sampling time. The stress level was 0.31.

A difference between MDS analyses for physical/chemical and bioindicator data was

the position of sites within the groupings. In the physical/chemical MDS, sites were

positioned somewhat linearly, with 0 km and 25 km being at opposite ends of the

grouping spread for each of the surveys. The bioindicator MDS, however, did not


follow this trend as 0 km and 25 km were grouped close together at one end of the

spread with middle river sites at the other end of the spread.

5.4 Discussion

Indicators used in this study identified every region of the river to be influenced by

nutrient inputs. The combination of physical/chemical parameters and biological

indicators proved successful in defining zones of urban sewage and agricultural

nutrient influences in the river and how these varied seasonally.

The seasonal variation of water quality observed in the Tweed River is not unlike

most rivers in sub-tropical/tropical Australia. Rivers in this part of the world are

typically characterised by highly variable flows, commonly receiving 80-90% of their

total yearly nutrient loading during flood events when they are rapidly flushed (Eyre

1995). This was evident in the mid-upper regions of the Tweed River with strong

contrasts in physical, chemical and biological indicators between wet and dry season

surveys. The quality of water in the lower regions of the river, however, was

maintained predominantly by oceanic flushing, as indicated by the similarity of

indicators in this region irrespective of season. The degree of oceanic influence on

water quality at sites further upriver was governed by rainfall and subsequent run-off

and leaching of diffuse pollutants, which tend to accumulate in previous dry seasons

(Eyre & Balls 1999). Hence, the elevated nutrients, suspended sediment loads and

bioindicator responses observed in the wet surveys in contrast to the dry surveys.

The predominance of nitrate in the river, particularly in the wet surveys, is also

concurrent with trends observed in other sub-tropical/tropical estuaries in wet and/or

Chapter 5 111

monsoonal periods (Eyre 1993, Eyre 1994). A number of studies world-wide have

discussed links between the application of N-fertilisers on agricultural areas and

increases in nitrate concentrations in adjacent waterways (Addiscott et al 1991,

Edwards et al 1990, Singh & Sekhon 1979). Nitrogen fertilisers used for agricultural

purposes, usually contain nitrogen in one of three forms, ammonium, nitrate or urea.

The activities of soilborne bacteria, usually mean that whatever form of fertiliser is

applied (ammonium, nitrate or urea) the end product in the soil is nitrate (Addiscott et

al 1991). Problems associated with these conversions, is that nitrates are

conspicuously water soluble, making them particularly susceptible to leaching

following rainfall (Addiscott et al 1991). The reverse is true for ammonium, which is

a positively charged ion and therefore strongly attracted to fine negatively charged

sediment particles (e.g. clay), thereby rarely being washed out of soils (Addiscott et al

1991). This would explain the trends observed in the Tweed River of elevated nitrate

concentrations in comparison to ammonium concentrations, particularly evident in the

wet season in the agriculturally dominated section of the river.

Elevated mean δ15N signatures of incubated macroalgae in wet season surveys, was

attributed to this increased microbial processing of nitrogen in catchment soils and

river sediments in the upper river regions. Faster microbial processing of the lighter

isotope (14N) than the heavier nitrogen isotope (15N) results in products enriched in

14N and leaves residues that are enriched in 15N (Mariotti et al 1981). As discussed

previously, nitrogen fertiliser applications are generally microbially converted to

nitrates in the soil which can be denitrified or taken up by crops. Either process will

select the lighter isotope (14N) resulting in 15N enrichment of the remaining nitrate

pool. The rates of bacterial nitrogen conversion (nitrification/denitrification), and


therefore 15N enrichment of the remaining nitrogen pool, are largely controlled by

temperature (Mariotti et al 1988). The effects of temperature on δ15N signatures are

evident in cooler ecosystems where low δ15N signatures, indicative of fertiliser, have

been found in surface waters after periods of heavy rainfall (Mariotti & Letolle 1977,

McClelland et al 1997), whereas elevated δ15N signatures have been observed in

warmer ecosystems subjected to fertiliser inputs (Fry et al 2001). Temperatures were

within the optimum range for nitrifying and denitrifying bacteria (Helder & de Vries

1983), during wet season surveys in this study, potentially explaining the mean

increase of δ15N signatures observed in macroalgae incubated in the agricultural

region of the river, in comparison to δ15N signatures observed in the dry season

surveys.

The lack of seasonal variation in mangrove δ15N, as opposed to incubated macroalgae,

can be explained by the life-history of mangroves and the sediment nitrogen source

that they utilise. Mangroves have slower productivity rates than macroalgae

(Campbell 1996) and act as longer temporal integrators of external conditions. The

high and low nature of diffuse nitrogen inputs into the upper Tweed River will

therefore result in a mangrove δ15N signature that is an ‘average’ of the variation,

whereas the continuous discharge of sewage nitrogen (enriched in 15N) at site 3,

results in a continuous elevated δ15N signature in mangroves and macroalgae adjacent

to this sewage outfall. Lower δ15N values recorded at the sewage discharge site (~6

‰), in comparison with values observed by Costanzo et al. (2001) in Moreton Bay -

Australia (~10 ‰), is likely due to strong tidal flushing of low δ15N seawater and

lower fractionation of nitrogen in the smaller sewage treatment discharging in this

region.

Chapter 5 113

The combination of mangrove and macroalgae tissue δ15N and %N proved a

successful combination for discerning sewage and agricultural inputs with different

δ15N signatures. This demonstrated that δ15N is not a stand-alone measure in systems

receiving multiple nitrogen inputs, as lower δ15N signatures in the agricultural region

of the river did not represent a lower nitrogen influence on biotic processes.

The polynomial relationship found between macroalgae δ15N and DIN in this study is

concurrent with other research where isotopic fractionation during algal DIN uptake

increases at higher DIN concentrations resulting in low δ15N values (Altabet 2001,

Fogel & Cifuentes 1993, Waser et al 1999). DIN in those studies was predominantly

NH4+, whereas NO3

- was the dominant source in this study suggesting that nitrogen

species is irrelevant in controlling fractionation of DIN uptake in macroalgae at high

concentrations. Mangrove δ15N, however, displayed no correlation with water column

DIN, which either supports the hypothesis that δ15N is not dependent on DIN

concentrations, rather on DIN source, or more likely that water column DIN

concentrations do not correlate with sediment nutrients in this system. This

assumption is supported by the lack of correlation between mangrove %N and DIN

concentrations, a relationship that was strong in the macroalgae, which utilises

nutrients directly from the water column.

Large seasonal variations were evident in phytoplankton responses to increased light

and nutrient availability, with minimal responses in the cooler dry season surveys than

the warmer wet season surveys. Due to ample light and nutrient supply to nutrient

bioassays, it was assumed that the large variation in seasonal temperatures was the

dominant factor controlling phytoplankton responses to nutrients in this region.


Temperature sets the theoretical maximum growth rate of phytoplankton, and together

with light and nutrient concentrations, regulates daily phytoplankton growth rates

(Eppley 1972). Phytoplankton growth relationships with seasonal variations in

temperature have been shown in other estuaries in the region (Chapter 3, O'Donohue

& Dennison 1997) and overseas (Eppley 1972, Fisher et al 1999, Pennock 1987).

Phytoplankton responses to increased light availability are dependent on external

nutrient concentrations, particularly nitrate, in order to support phytoplankton growth

as described in Chapter 3. The large variation in water column nitrate concentrations

observed seasonally in the Tweed River is therefore likely to be a dominant control of

phytoplankton responses to increased light availability. Largest light responses

measured in the middle region of the Tweed River coincided with highest nitrate

concentrations, thereby supporting this relationship.

The mean increase in water quality at the upper most site of the river in wet season

surveys is likely a result of a combination of physical and biological factors. Due to

the presence of a weir, which effectively divides the river, catchment inputs to this

upper most site are limited, as opposed to the middle region if the river which receives

the combined input of the upper Tweed River and the Rous River. Tidal excursions,

although present at site 6, were smaller than excursions in the lower parts of the river,

resulting in less tidal resuspension of fine particulate matter. Increased clarity at this

site was most likely responsible for the elevated chlorophyll a concentrations in this

region.

Chapter 5 115

5.5 Conclusion

This study found that the Tweed River was sensitive to different anthropogenic

activities in its catchment and that each activity had a unique influence on receiving

water quality. Subsequently, nutrient influences could not be assessed with a single

indicator and required a combined approach in order to better characterise nutrient

influences in the system. The use of physical, chemical and biological indicators

successfully delineated zones of sewage and agricultural nutrient influences within the

Tweed River and identified no region of the river to be immune to nutrient inputs.

Nutrient influences on water quality displayed strong seasonal variations, with greater

influences in the wet season where indicators throughout the estuary reflected the high

catchment nutrient loadings typically associated with this season.

CChhaapptteerr 66 Discussion

Indicators for assessing and monitoring nutrient influences in coastal waters were

developed in a semi-enclosed coastal embayment receiving large sewage inputs

(Chapter 2 & 3). Indicators comprised of δ15N signatures of marine flora and

phytoplankton bioassays, combined with standard physical/chemical parameters. The

potential of these indicators to provide useful information on the influence of shrimp

farm effluent in a tidal mangrove creek (Chapter 4), and agricultural inputs into a

coastal river estuary (Chapter 5), were also investigated. The combination of these

techniques proved successful in discerning spatial and temporal influences of the

various nutrient inputs to coastal marine systems. This chapter aims to provide a

synthesis of the results and findings from the three case studies investigated in this

thesis, thereby providing a forum for the different indicators to be compared and

contrasted. Finally, situations where these techniques and findings have been adopted

will be discussed.

6.1 Conceptual Overview of Thesis Findings

The previous chapters clearly show the utility of δ15N measurements of marine flora

and phytoplankton bioassays in assessing the influence of nutrient inputs to coastal

waters. To identify overall trends found and synthesise divergent results into a single

depiction for each study region, several conceptual diagrams will be presented and

briefly described. These diagrams depict the major environmental features evident in

each study region and attempt to encapsulate information attained from the different

systems studied.

118 Discussion

6.1.1 Moreton Bay – sewage effluent dominated

Moreton Bay is a sub-tropical, shallow coastal embayment on the east coast of

Australia. The bays drainage catchment to the west is largely dominated by urban

development and receives large point-source inputs of secondary treated sewage

effluent. The eastern side receives relatively low-nutrient oceanic water and few

anthropogenic inputs. Consequent ly, strong eutrophication gradients exist from

western to eastern sides of the bay (Figure 6.1). Physical/chemical parameters, that

displayed greatest change along this west-east gradient, included nitrates (NO3-)

chlorophyll a, and secchi depth (Figure 6.1 A). The trends observed in these variables

did not differ appreciably between summer and winter/spring seasons, potentially due

to extended dry periods prior to sampling surveys. δ15N signatures were found to be

elevated in naturally occurring macroalgae, seagrasses and mangroves on the western

shore close to sewage effluent outfalls, values approaching δ15N signatures typical of

treated sewage effluent (~ 10 ‰) (Figure 6.1 B). δ15N signatures of flora growing on

the eastern shores were typically < 3‰. The zone of sewage influence into the

western bay, determined by incubating macroalgae in the water column, identified

distinct sewage plumes emanating from the central western shore. This technique

highlighted distinct variations in the geographical extent of sewage plumes between

seasons, with increased δ15N signatures detected up to 10km in summer and 5km in

winter from the sewage source. Response of phytoplankton to increased light

availability was greater in the rivers than the bay and did not vary between summer

and winter (Figure 6.1 C). Reponses to increased nutrient availability displayed

similar spatial trends as light responses (higher in rivers than bay), though these did

vary between seasons (large responses in summer) primarily due to changes in

temperature.

Chapter 6 119

Comparison

Winter / Spring

Similar trend as above but lower responses

Similar as aboveWest-east gradient

Defined zone of sewage influence – 5 km

Same as aboveWest-east gradient• Similar as above

Temp. (23 oC)

West-east gradient• NO3

-

(~ 10 – 0.5 µM)•Chlorophyll(~ 3 - 1µg/L)

Temp. (26 oC)

Phys / Chem

West-east gradient(~10–1 fsu.d-1)

Higher in rivers than bay

Phytoplankton Bioassays



West-east gradient(~ 10 – 3 ‰)



Confined to western Bay

Summer

• Semi-enclosed bay

• 2 – Dimensional

• Urban Catchment

• Tidal

• Large sewage inputs

• Point Source

• Continuous discharge

δ15N SignaturesComparison

Winter / Spring

Similar trend as above but lower responses

Similar as aboveWest-east gradient


Same as aboveWest-east gradient• Similar as above

Temp. (23 oC)

West-east gradient• NO3

-

(~ 10 – 0.5 µM)•Chlorophyll(~ 3 - 1µg/L)

Temp. (26 oC)

Phys / Chem









Confined to western Bay

Summer

• Semi-enclosed bay

• 2 – Dimensional

• Urban Catchment

• Tidal

• Large sewage inputs

• Point Source

• Continuous discharge

δ15N SignaturesMoreton BayNitrate

Chlorophyll

Secchi Depth

Temperature

Mangrove δ15N

Macroalgae δ15N

Incubated Macroalgae

δ15N

Light Response

Nutrient Response

Seagrass δ15N

Sewage Inputs

A B C

Figure 6.1 Diagrammatic representations of findings in Moreton Bay during summer and winter/spring assessments.

120 Discussion

6.1.2 Cardwell – Shrimp Effluent

The shrimp farm located at Cardwell discharges effluent into a 3km long tidal

mangrove creek. Apart from the shrimp farm, this creek drains a small relatively

undisturbed catchment. Physical/chemical parameters identified distinct changes in

the receiving creek with respect to farm operations (Figure 6.2 A). Elevated water

column NH4+ and chlorophyll a concentrations were measured when the farm was in

operation, in contrast to when the farm was inactive. At all times, physical/chemical

parameters at the mouth of the effluent creek, were equivalent to control values,

indicating effluent was contained within the effluent-receiving creek. However,

elevated δ15N signatures of mangroves and macroalgae indicated a broader influence

of shrimp farm effluent, extending to the lower regions of the farms intake creek

(Figure 6.2 B). Mangroves still displayed elevated δ15N signatures when the farm was

inactive due to their longer temporal integration, thereby providing a ‘memory’ of

historical conditions. Elevated concentrations of chl a, when the farm was in

operation, were measured at upstream sites close to the location of farm effluent

discharge. Bioassays indicated that phytoplankton at these sites did not respond to

further nutrient additions, however downstream sites showed large growth responses

to nutrient additions (Figure 6.2 C). The opposite was evident with phytoplankton

responses to increased light availability.

Chapter 6 121

Comparison

Farm Inactive Ascending with distance away from source(~ 0–40 fsu.d-1)

Small peak in middle of creek(~ 0–10 fsu.d -1)

No δ15N increase(~ 3 ‰)

High to lowSame as above‘memory’

Contained within creek

Low • NH4

+

(~ 1.5 µM)

• Chlorophyll(~ 1.0 µg/L)


High to low• NH4

+

(~ 40 – 1µM)

• Chlorophyll (~10 – 1µg/L)

Phys / Chem

Peak in mid-creek(~ 0-20 fsu.d -1)


Peak in mid-creek(~ 0–40 fsu.d-1)

Increase δ15N throughout creek (~ 5 ‰)

Elevated

Variable spatial

Decreasing δ15N from source(~ 8 – 3 ‰)

Farm Active• Mangrove lined creek

• 1 – Dimension

• Small undisturbed catchment

• Minor flushing

• Shrimp Farming

• Point Source

• Pulsed and seasonal ( ie when ponds inoperation)

δ1 5N SignaturesComparison

Farm Inactive Ascending with distance away from source(~ 0–40 fsu.d-1)

Small peak in middle of creek(~ 0–10 fsu.d -1)

No δ15N increase(~ 3 ‰)

High to lowSame as above‘memory’


Low • NH4

+

(~ 1.5 µM)

• Chlorophyll(~ 1.0 µg/L)


High to low• NH4

+

(~ 40 – 1µM)

• Chlorophyll (~10 – 1µg/L)

Phys / Chem

Peak in mid-creek(~ 0-20 fsu.d -1)


Peak in mid-creek(~ 0–40 fsu.d-1)

Increase δ15N throughout creek (~ 5 ‰)

Elevated

Variable spatial

Decreasing δ15N from source(~ 8 – 3 ‰)

Farm Active• Mangrove lined creek

• 1 – Dimension

• Small undisturbed catchment

• Minor flushing

• Shrimp Farming

• Point Source

• Pulsed and seasonal ( ie when ponds inoperation)

δ1 5N SignaturesCardwellAmmonium

Chlorophyll

Secchi Depth

Temperature

Mangrove δ15N


δ15N

Light Response

Nutrient Response

Shrimp Effluent

A B C

Figure 6.2 Diagrammatic representation of findings in the shrimp effluent receiving creek at Cardwell during periods when the farm was active and inactive.

122 Discussion

6.1.3 Tweed River – Agriculture Dominated

The Tweed River meanders across a flood plain dominated by sugar cane farming in

the mid-upper regions and urban development in the lower region. The mouth of the

river is stabilised and routinely dredged for navigational purposes (Plate 6.3). Tidal

influence penetrates to most of the system below the weir which is located

approximately 40km upstream. Strong spatial (within river) and temporal (seasonal)

trends in water quality and bioindicator responses were observed. Poorest water

quality was observed in the mid-upper river region in the wet season due to large

diffuse inputs. Poor physical/chemical water quality was indicated by elevated NO3-,

total suspended solid and chlorophyll concentrations. The elevated concentrations of

NO3- in the river in the wet season were attributed to microbial nitrification in the

catchment soils and river sediments. The combination of mangrove and macroalgal

tissue δ15N and %N discerned sewage and agricultural influences within the river.

Elevated δ15N and %N represented sewage inputs in the lower river, whereas low

δ15N and elevated %N was indicative of agricultural inputs in the mid-upper river.

Mangrove δ15N and %N did not display seasonal variation due to the longer

integration time of mangroves, whereas macroalgae did reflect the seasonal variation

in water quality, lower values in the cooler dry season than the warmer wet season.

Phytoplankton responses to increased light and nutrients were largely restricted to the

wet seasons most likely due to increased nutrients and temperature, respectively.

There appeared to be a ‘trade-off’ between nutrient and light responses, with light

responses peaking in the middle region of the river and nutrient responses peaking on

either side of this light response.

Chapter 6 123

Similar trend as above but lower mean values

Increase in lower river to plateau for rest of river(~ 1.6 – 2.0 %)

Same as above

Increase upriver from mouth(~ 4 – 7 %)

Tissue Nitrogen Content (%N)Comparison

Dry Season No responseMinor responseSimilar trend as above but lower mean values

Same as aboveWater quality low and uniform throughout.• NO3

- (~ 5 µM)• Temp. 18 oC

• ElevatedChl a upriver(~ 10µg/L)

Poorest water quality mid -river• NO3

- (~30µM)• Chlorophyll a

(~ 10µg/L)• Temp. 26 oC

Phys / Chem

Elevated mid -river (~ 10 fsu.d-1)


Elevated either side of light response in mid-upper river(~50 fsu.d-1)

δ15N defined urban (~ 5 ‰) vs. agricultural (~3.5 ‰)


Wet Season• River – ocean discharge

• 1-dimensional

• Sparse riparian vegetation

• Well flushed

• Agricultural dominated upper river

• Urban dominated lower river

• Diffuse inputs upper

• Seasonal

• W et driven

δ15N Signatures

Similar trend as above but lower mean values

Increase in lower river to plateau for rest of river(~ 1.6 – 2.0 %)

Same as above

Increase upriver from mouth(~ 4 – 7 %)

Tissue Nitrogen Content (%N)Comparison

Dry Season No responseMinor responseSimilar trend as above but lower mean values

Same as aboveWater quality low and uniform throughout.• NO3

- (~ 5 µM)• Temp. 18 oC

• ElevatedChl a upriver(~ 10µg/L)

Poorest water quality mid -river• NO3

- (~30µM)• Chlorophyll a

(~ 10µg/L)• Temp. 26 oC

Phys / Chem

Elevated mid -river (~ 10 fsu.d-1)


Elevated either side of light response in mid-upper river(~50 fsu.d-1)



Wet Season• River – ocean discharge

• 1-dimensional

• Sparse riparian vegetation

• Well flushed

• Agricultural dominated upper river

• Urban dominated lower river

• Diffuse inputs upper

• Seasonal

• W et driven

δ15N SignaturesTweed RiverNitrate

Chlorophyll

Secchi Depth

Temperature

Mangrove δ15N


δ15N

Light Response

Nutrient Response

Sewage Inputs

Agricultural Inputs

Mangrove %N

Incubated Macroalgae %N

A B C D

Figure 6.3 Diagrammatic representation of findings in the Tweed River during the wet and dry seasons.

124 Discussion

6.2 Applicability of Physical/Chemical and Biological Indicators

for Assessing Nutrients in Estuarine and Coastal Waters

The challenge of this thesis research was to develop indicators that impart integrated

information on the source and influence of nutrients in estuarine and coastal waters.

The approaches taken to address this issue were tested in the aforementioned case

studies, thereby subjecting them to a diversity of systems and inputs. The relative

ability of the various indicators to relate nutrient conditions in each of the study

regions was evident in the synthesis provided in the preceding section. How these

indicators are generically applicable will be the forum of discussion in this section.

The principal difficulty in assessing water quality is the variability of the water

column in space and time. There are great changes in estuarine and coastal water

quality at any one fixed point, notable with daily ebb and flow of tides, seasonal

inputs of freshwater and intermittent discharges of pollutants (Wilson 1988, Wolfe &

Kjerfve 1986). Historically, variability in water quality has often confounded

assessment and monitoring programs, based on instantaneous ‘snap-shot’ measures

typical of physical/chemical sampling (Maher et al 1994). Approaches have been

developed to capture aspects of variability using physical/chemical parameters, by

conducting spatially intensive (Eyre & Pepperell 1999) or temporally intensive

sampling strategies (Jarvie 2001). However, these approaches are costly, labour

intensive and due to logistical reasons, rarely address both spatial and temporal

variabilities together. Indeed, temporal collection of physical/chemical parameters

may not always be possible, as was evident in the Cardwell case study, where boat

access to the effluent creek was limited to high tide due to the shallow nature of the

creek and surrounding waters.

Chapter 6 125

Having described the drawbacks of using physical/chemical parameters for water

quality assessment, their value is not to be dismissed in a broader assessment

framework as they do represent many components of the studied system. These

shortcomings relate primarily to the sole use of these parameters as water quality

indicators. Research in this thesis successfully used physical/chemical measures for

characterising aspects of water quality and as a basis to learn more information on the

studied systems. The quick and relatively cheap information obtained from

physical/chemical sampling were useful for inspecting a study region prior to

commencement of an intended sampling program. Trends observed in

physical/chemical data (e.g. salinity, nutrients, secchi depth, and chlorophyll) in this

initial inspection were used to aid design of the subsequent water quality assessment

that incorporated both physical/chemical measurements and biological indicators.

Studying the responses of biota, or biotic communities, as a means of assessing water

quality is a powerful tool, as a) it is the impacts of water quality on aquatic biota that

managers are often trying to assess, b) it avoids the necessity of interpreting water

quality values in terms of biotic impacts, and c) biotic communities will reflect

integrated ‘assessment’ of water quality parameters that may i) act synergistically,

where the combined impact is more severe then any one impact alone, ii) have

impacts that vary sporadically and outside an established sampling schedule, and iii)

be below the detection limits of standard sampling and analysis capabilities.

Marine flora were chosen as biological indicators as they are efficient at reflecting

past exposure to nutrients (Costanzo et al 2000, Dennison et al 1987, Fong et al 1994,

Horrocks et al 1995, Jones et al 1996), and so by sampling them at one point in time,

it is possible to infer information on the average nutrient availability present in the

126 Discussion

studied system. The variation of marine flora used within this thesis (phytoplankton,

macroalgae and mangroves) each provided a different viewpoint on nutrient dynamics

in the different systems studied due to their different life stages and positions within

the environment.

Phytoplankton responsiveness to increased nutrient and light availability was

indicative of the systems potential water quality status, not attainable from the

physical/chemical parameters measured in this thesis (Chapter 3). Key points to note

in the use of this indicator were the combined interpretation of phytoplankton

responses and ambient physical/chemical parameters. For instance, the low response

of phytoplankton to nutrient additions in eastern Moreton Bay (Chapter 3) and the

upper region of the effluent creek at Cardwell (Chapter 4), do suggest that these

environments will not largely responds to increased nutrient inputs. However, the

reason for these scenarios was very different when taking into account

physical/chemical data; the upper region of the effluent creek at Cardwell being

nutrient replete with phytoplankton already in high concentrations, whereas in eastern

Moreton Bay, low bioassay responses were due to low initial standing stocks of

phytoplankton that were likely not accustomed to elevated nutrient conditions

[O'Donohue, 1997 #210].

Macroalgae and mangroves were used to assess the presence and extent of nutrient

inputs in marine systems. Macroalgae incubated in the water column were useful in

characterising nutrient conditions in the water column over a known timeframe. This

thereby integrated variability encountered within days (e.g. tidal variations, episodic

rainfall). In addition this technique allowed measurements to be made at any desired

Chapter 6 127

location, as naturally occurring marine flora was not always present in regions of

interest. This was particularly evident in Moreton Bay which was assessed in two-

dimensions unlike the linear one-dimensional approaches used to assess water quality

in the Tweed River and at Cardwell where flora could be collected along the banks of

these systems. The slower growth of mangroves provided longer temporal integration

in the order of months and/or years as demonstrated by the ‘memory’ effect displayed

in mangroves in Cardwell when the shrimp farm was inactive and in the dry season in

the Tweed River when diffuse inputs were minimal. Also mangroves provided

information on the sediment nutrient status, which also reflects an integrated

assessment of nutrient variability (Voss & Struck 1997).

The difference of δ15N signatures associated with various nutrient sources

successfully defined and delineated different nitrogen sources. It was clear that δ15N

signatures of flora were not always similarly reflected by similar inputs (i.e. sewage

inputs to Moreton Bay and the Tweed River did not result in matching δ15N signatures

of incubated macroalgae). A suite of biotic and abiotic factors can influence nitrogen

inputs and thereby the relative signature measured in marine flora. These factors

include microbial processing (e.g. nitrification, denitrification), volatilisation, dilution

and even mixing of different δ15N sources. The lower δ15N signatures (~5 ‰)

measured in macroalgae incubated adjacent to the sewage outfall in the Tweed River,

compared to macroalgae incubated in western Moreton Bay (~10‰), is likely a

combination of tidal flushing in that part of the river and interaction of low δ15N

signatures from the upstream agricultural region. The combination of tissue δ15N and

%N proved a successful combination for discerning multiple nitrogen inputs. This

was demonstrated in the Tweed River where elevated δ15N and %N represented

128 Discussion

sewage inputs, whereas low δ15N and elevated %N was indicative of agricultural

inputs.

6.3 Application of Findings

6.3.1Communication

The success of the methods developed in this thesis, is evident by the management

actions that have occurred from the presented data and the incorporation of these

techniques into regional environmental programs (Dennison & Abal 1999). The

critical component of this transition from science to management was the concurrent

communication of results and findings to management bodies with ongoing research.

This enabled both parties to develop questions and objectives and to envision the final

outcome. The benefit of this approach was that research was continually evolved to

address real-time problems encountered in each of the study regions. Of course,

research conducted in this thesis was not applicable to every aspect of management

concern, though a combined involvement between parties generally assisted

interpretation of issues and collaboration with other specialists.

Simultaneous involvement of research and management was achieved through a

variety of communicative forms, including public meetings, workshops and

presentations of print and digital media. Apart form immediate contact with

management bodies (e.g. council meetings), an effective form of communication was

colour newsletters presented in a fashion that was communicable to all levels of

involvement, from the general public to management bodies. The success of this form

of communication was evident in the Tweed River study, where the local council

supplied each household in the district with copies of these newsletters, a total of

Chapter 6 129

~30,000+ homes (Appendix I). Initial newsletters aimed to describe the nature of the

studied environment, highlighting predominant features and pressures on the system

and outlining research to be conducted. Subsequent newsletters presented results,

interpretations and actions to be taken to compensate degraded systems (e.g. Moreton

Bay Study – Appendix II). This form of communication provided a conduit for

information to flow between end-users of the respective environment.

6.3.2 Continued monitoring

In addition to developing indicators aimed at assessing nutrient influences in estuarine

and coastal waters, much of the data collected in this thesis has acted as baseline data

for future comparisons in these systems. Having been involved with management

bodies throughout the research process has lead to techniques developed in this thesis

being adopted in ongoing monitoring programs. For instance, the relevance of

sewage mapping using δ15N signatures of incubated macroalgae provides a technique

that will assess the success, or failure, of large financial investments in sewage

upgrades in Moreton Bay. Theoretically, following nutrient reduction emissions from

sewage treatment plants, the δ15N influence of sewage plumes in the bay should begin

to retreat. This form of evidence is a powerful advertisement for councils aiming to

improve the environment, satisfy government regulations and justify funding of such

projects.

130 Discussion

6.5 Conclusion

Over 2000 scientific articles, associated with marine eutrophication, have been

published in the last 25-30 years (>1 article/week) (Biol. Abstracts). Considering that

the concept “marine eutrophication” was virtually unheard of 30 years ago (Nixon

1995), highlights the recent increase in occurrence, awareness and research of this

phenomenon. This thesis research, albeit a small contribution to the mass of interest

in this field, has provided useful techniques for assessing and monitoring biologically

available nutrients in coastal waters. These techniques were found to be most

effective when used in combination with each other, evident by their success in

describing water quality in very different contexts. Strategies employed to test these

indicators have yielded functional information on the influence of various nutrient

inputs, thereby providing resource managers with baseline criteria to gauge future

monitoring efforts. Australia’s population growth rate (>100,000 persons per year

over 1990-2000) means that increasing population of coastal areas will result in

increased pressures on near shore ecosystems. It is hoped that the indicators for

assessing and monitoring coastal waters developed herein may be used to mitigate

risk of current and future population pressures to coastal marine ecosystems.

References

Addiscott TM, Whitmore AP, Powlson DS. 1991. Farming, fertilisers and the nitrate

problem. Wallingford, U. K.: CAB International

Aidar E, Sigaud Kutner Teresa CS, Nishihara L, Schinke Katya P, Braga Maria

Cristina C, et al. 1996. Marine phytoplankton assays: Effects of detergents.

Marine Environmental Research 43: 55-68

Alongi, D. M., D. J. Johnston, et al. 2000. Carbon and nitrogen budgets in shrimp

ponds of extensive mixed shrimp-mangrove forestry farms in the Mekong

delta, Vietnam. Aquaculture Research 31: 387-399.

Alongi, D. M., F. Tirendi, Trott, L. A. 1999. Rates and pathways of benthic

mineralization in extensive shrimp ponds of the Mekong delta, Vietnam.

Aquaculture 175: 269-292.

Altabet MA. 2001. Nitrogen isotopic evidence for micronutrient control of fractional

NO3- utilisation in the equatorial Pacific. Limnology and Oceanography 46:

368-80

Belbin L. 1993. PATN Patter Analysis Package: Users Guide: CSIRO Division of

Wildlife and Ecology

Berounsky VM, Nixon SW. 1990. Temperature and the annual cycle of nitrification in

waters of Narragansett Bay (Rhode Island USA). Limnology And

Oceanography 35: 1610-7

Bonetto C, Zalocar Y, Planas D, Pedrozo F. 1991. Responses of phytoplankton to

experimental nutrient enrichment in the Paraguay, Bermejo and Upper Parana

rivers. Tropical Ecology 32: 47-64

Borum J, Sand Jensen K. 1996. Is total primary production in shallow coastal marine

waters stimulated by nitrogen loading? Oikos 76: 406-10

Boyd CE. 1995. Bottom Soils, Sediment and Pond Aquaculture. New York: Chapman

& Hall

Briggs MRP, Funge SSJ. 1994. A nutrient budget of some intensive marine shrimp

ponds in Thailand. Aquaculture and Fisheries Management 25: 789-811

Burford MA, Glibert PM. 1999. Short-term nitrogen uptake and regeneration in early

and late growth phase shrimp ponds. Aquaculture Research. March 30: 215-27

Cabana G, Rasmussen J. 1996. Comparison of aquatic food chains using nitrogen

isotopes. Proceedings of the National Academy of Sciences of the United

States of America 93: 10844-7

Cabana G, Rasmussen J. B. 1994. Modelling food chain structure and contaminant

bioaccumulation using stable nitrogen isotopes. Nature London 372: 255-7

Campbell NA. 1996. Biology. Menlo Park, California: The Benjamin/Cummings

Publishing Company, Inc. 1206 pp.

Carignan R, Planas D. 1994. Recognition of nutrient and light limitation in turbid

mixed layers: Three approaches compared in the Parana floodplain

(Argentina). Limnology and Oceanography 39: 580-96

Carrick HJ, Schelske CL, Aldridge FJ, Coveney MF. 1993. Assessment of

phytoplankton nutrient limitation in productive waters: Application of dilution

bioassays. Canadian Journal of Fisheries and Aquatic Sciences 50: 2208-21

Clarke KR, Warwick RM. 1994. Change in Marine Communities: An Approach to

Statistical Analysis and Interpretation. Plymouth, U. K.: Natural Environment

Research Council

Clesceri LS, Greenberg AE, Trussell RR. 1998. Standard Methods for the

Examination of Water and Waste Water: APHA-AWWA-WPCF

133

Cole JJ, Peierls BL, Caraco N, Pace ML. 1993. Nitrogen loading of rivers as a human

driven process. In Humans as components of ecosystems: the ecology of subtle

human effects and populated areas, ed. MJ Mc Donnell, STA Picket, pp. 138-

54. New York: Springer Verlag

Cosser P, ed. 1997. Nutrients in marine and estuarine environments. Canberra,

Australia: Department of Environment

Costanzo SD, O DMJ, Dennison WC. 2000. Gracilaria edulis (Rhodophyta) as a

biological indicator of pulsed nutrients in oligotrophic waters. Journal of

Phycology. August 36: 680-5

Costanzo SD, O'Donohue MJ, Dennison WC, Loneragan NR, Thomas MT. 2001. A

new approach for detecting and mapping sewage impacts. Marine Pollution

Bulletin 42: 149-56

Cowell BC, Dawes CJ. 1991. Nutrient enrichment experiments in three central Florida

(USA) lakes of different trophic states. Hydrobiologia 220: 217-32

Cressie NAC. 1993. Statistics for Spatial Data: Revised Edition. New York: Wiley

Cullen P. 1982. Managing the Australian coastal zone. In Australia's Maritime

Horizons in the 1980s, occasional papers in maritime affairs: 1, pp. 82.

Canberra, Australia: Australian Centre for Maritime Studies

D'Elia CF. 1987. Too Much of a Good Thing: Nutrient Enrichment of the Chesapeake

Bay. Environment 29: 6-11

D'Elia CF, Sanders, J. G, and Boynton, W. R. 1985. Nutrient enrichment studies of a

coastal plain estuary: Phytoplankton growth in large-scale, continuous

cultures. Canadian Journal of Fisheries and Aquatic Science 4: 397-406

Delwiche CC, Bryan BB. 1976. Denitrification. Annual Review of Microbiology 30:

241-62

Dennison WC, Abal EG. 1999. Moreton Bay Study: A scientific basis for the Healthy

Waterways Campaign. Brisbane: South East Queensland Regional Water

Quality Management Strategy. 245 pp.

Dennison WC, Aller RC, Alberte RS. 1987. Sediment ammonium availability and

eelgrass (Zostera marina) growth. Marine Biology 94: 469-78

Downing JA. 1997. Marine nitrogen: Phosphorus stoichiometry and the global N:P

cycle. Biogeochemistry 37: 237-52

Duke NC, Bunt JS, Williams WT. 1984. Observations on the floral and vegetative

phenologies of North-eastern Australian mangroves. Australian Journal of

Botany 32: 87-9

Edwards AC, Pugh K, Wright G, Sinclair AH, Reaves GA. 1990. Nitrate status of two

major rivers in NE Scotland with respect to land use and fertiliser additions.

Chemistry and Ecology 4: 97-107

Eppley RW. 1972. Temperature and phytoplankton growth in the sea. Fisheries

Bulletin 70: 1063-85

Eyre B. 1993. Nutrients in the sediments of a tropical north-eastern Australian

estuary, catchment and nearshore coastal zone. Australian Journal of Marine

and Freshwater Research 44: 845-66

Eyre B. 1994. Nutrient biogeochemistry in the tropical Moresby River estuary system

north Queensland, Australia. Estuarine Coastal and Shelf Science 39: 15-31

Eyre B. 1995. A first-order nutrient budget for the tropical Moresby Estuary and

catchment, North Queensland, Australia. Journal of Coastal Research 11:

717-32

Eyre B. 1998. Transport, retention and transformation of material in Australian

estuaries. Estuaries 21: 540-51

135

Eyre B, Balls P. 1999. A comparative study of nutrient behaviour along the salinity

gradient of tropical and temperate estuaries. Estuaries 22: 313-26

Eyre BD, Pepperell P. 1999. A spatially intensive approach to water quality

monitoring in the Rous catchment, New South Wales, Australia. Journal of

environmental management 56: 97-118

FAO. 2000. The state of world fisheries and aquaculture, Food and Agriculture

Organization, Rome, Italy

Fisher TR, Gustafson AB, Sellner K, Lacouture R, Haas LW, et al. 1999. Spatial and

temporal variation of resource limitation in Chesapeake Bay. Marine Biology

Berlin. May, 1999; 133: 763-78.

Focht DD, Verstraete W. 1977. Biochemical ecology of nitrification and

denitrification. Advanced Microbial Ecology 1: 135-214

Fogel ML, Cifuentes LA. 1993. Isotopic fractionation during primary production. In

Organic Chemistry, ed. M Engel, SA Macko, pp. 73-98. New York: Plenum

Press

Fong P, Donohoe R, Zedler JB. 1994. Nutrient concentration in tissue of the

macroalga Enteromorpha as a function of nutrient history: An experimental

evaluation using field microcosms. Marine Ecology Progress Series 106: 273-

81

Fong P, Zedler JB. 1993. Temperature and light effects on the seasonal succession of

algal communities in shallow coastal lagoons. Journal of Experimental Marine

Biology and Ecology 171: 259-72

Fry B, Bern AL, Ross MS, Meeder JF. 2000. δ15N studies of nitrogen use by the red

mangrove, Rhizophora mangle L. in south Florida. Estuarine Coastal and

Shelf Science 50: 291-6

Fry B, Gace A, McClelland JW. 2001. Chemical Indicators of Anthropogenic

Nitrogen Loading in West Coast NERR (National Estuarine Research Reserve)

Estuaries, The Cooperative Institute for Coastal and Estuarine Environmental

Technology

Fukami K, Nishijima T, Ishida Y. 1997. Stimulative and inhibitory effects of bacteria

on the growth of microalgae. Hydrobiologia 358: 185-91

Funderbunk SL, Jordan SJ, Mihursky JA, Riley D. 1991. Habitat requirements for

Chesapeake Bay living resources. Solomons, MD: Chesapeake Research

Consortium

Funge-Smith SJ, Briggs MRP. 1998. Nutrient budgets in intensive shrimp ponds:

Implications for sustainability. Aquaculture 164: 117-33

Gerhart DW, Likens GE. 1975. Enrichment experiments for determining nutrient

limitation: four methods compared. Limnology and Oceanography 20: 649-53

Gowen RJ, Rosenthal H, Maekinen T, Ezzi I. 1990. Environmental impact of

Aquaculture activities. In Business joins Science, ed. D Depauw, pp. 257:

Spec. Publ. Eur. Aquacult. Soc. no. 12

Grant J, Hatcher A, Scott DB, Pocklington P, Schafer CT, Winters GV. 1995. A

multidisciplinary approach to evaluating impacts of shellfish aquaculture on

benthic communities. Estuaries 18: 124-44

Grice AM, Loneragan NR, Dennison WC. 1996. Light intensity and the interactions

between physiology, morphology and stable isotope ratios in five species of

seagrass. Journal of Experimental Marine Biology and Ecology 195: 91-110

Hart, B. Draft. ANZECC Guidelines. National Water Quality Management Strategy.

Agriculture and Resource Management Council of Australia and New

Zealand.

137

Hansson S, Hobbie John E, Elmgren R, Larsson U, Fry B, Johansson S. 1997. The

stable nitrogen isotope ratio as a marker of food-web interactions and fish

migration. Ecology Washington D C 78: 2249-57

Healey FP. 1979. Short-term responses of nutrient-deficient algae to nutrient addition.

Journal of Phycology 15: 289-99

Heaton THE. 1986. Isotopic studies of nitrogen pollution in the hydrosphere and

atmosphere: a review. Chemical geology 59: 87-102

Hecky RE, Kilham P. 1988. Nutrient Limitation of Phytoplankton in Fresh-Water and

Marine Environments - a Review of Recent-Evidence on the Effects of

Enrichment. Limnology and Oceanography 33: 796-822

Heggie DT, Skyring GW, Berelson WM, Longmore AR, Nicholson GJ. 1999.

Sediment-water interaction in Australian coastal environments: Implications

for water and sediment quality. AGSO Journal of Australian Geology and

Geophysics 17: 159-73

Helder W, de Vries RTP. 1983. Estuarine nitrate maxima and nitrifying bacteria

(Ems-Dollard Estuary). Netherlands Journal of Sea Research 17: 1-18

Hensey MP. 1991. Environmental monitoring for fish farms. In Aquaculture and the

Environment, ed. N De Pauw, J Joyce, pp. 536. Oxford, England: Special

Publication of the European Aquaculture Society

Henze M. 1995. Wastewater treatment : biological and chemical processes. Berlin:

Springer-Verlag. 383 pp.

Hobbie JE, Fry B. 1990. Sewage derived 15N in the Baltic traced in Fucus. EOS : 190

Horrocks JL, Stewart GR, Dennison WC. 1995. Tissue nutrient content of Gracilaria

spp. (Rhodophyta) and water quality along an estuarine gradient. Marine and

Freshwater Research 46: 975-83

Howarth RH. 1988. Nutrient limitation on net primary production in marine

ecosystems. Annual Review of Ecology and Systematics 19: 89-110

Howarth RW, Billen G, Swaney D, Townsend A, Jaworski N, et al. 1996. Regional

nitrogen budgets and riverine N&P fluxes for the drainages to the North

Atlantic Ocean: Natural and human influences. Biogeochemistry 35: 75-139

Jackson C, Preston NP, Thompson PA, Burford M. In prep. Nitrogen budget of an

intensive shrimp farm.

Jarvie HP, Neal C, Smart R, Owen R, Fraser D, et al. 2001. Use of continuous water

quality records for hydrograph separation and to assess short-term variability

and extremes in acidity and dissolved carbon dioxide for the River Dee,

Scotland. Science of the Total Environment. 265: 85-98

Jeffrey DW, Brennan MT, Jennings E, Madden B, Wilson JG. 1995. Nutrient sources

for in-shore nuisance macroalgae: The Dublin Bay case. Ophelia 42: 147-61

Jensen MH, Lomstein E, Sorensen J. 1990. Benthic NH4+ and NO3

- flux following

sedimentation of a spring phytoplankton bloom in Aarhus Bight, Denmark.

Marine Ecology Progress Series 61: 87-96

Jones AB, Dennison WC, Stewart GR. 1996. Macroalgal responses to nitrogen source

and availability: Amino acid metabolic profiling as a bioindicator using

Gracilaria edulis (Rhodophyta). Journal of Phycology 32: 757-66

Jones AB, Dudley BJ, Dennison WC. 1998. Factors limiting phytoplankton biomass

in the Brisbane River and Moreton Bay. In Moreton Bay and Catchment, ed.

IR Tibbets, NJ Hall, WC Dennison, pp. 179-86. Brisbane: School of Marine

Science, The University of Queensland

Jones AB, O'Donohue MJ, Udy J, Dennison WC. 2001. Assessing ecological impacts

of shrimp and sewage effluent: biological indicators with standard water

quality analyses. Estuarine, Coastal and Shelf Science 52: 91-109

139

Junk G, Svec HJ. 1958. The absolute abundance of the nitrogen isotopes in the

atmosphere and compressed gases from various sources. Geochimica Et

Cosmochimica Acta: 234-43

Kaplan W, Valiela I, Teal JM. 1979. Denitrification in a saltmarsh ecosystem.

Limnology and Oceanography 24: 726-34

Lake PS, Barmutta LA, Boulton AJ, Campbell IC, St. Clair RM. 1985. Australian

streams and Northern Hemisphere ecology: comparisons and problems.

Proceedings of the Ecological Society of Australia 14: 61-82

Lapointe BE, Clark MW. 1992. Nutrient Inputs from the Watershed and Coastal

Eutrophication in the Florida Keys. Estuaries 15: 465-76

Lavery PS, Lukatelich RJ, McComb AJ. 1991. Changes in the biomass and species

composition of macroalgae in a eutrophic estuary. Estuarine Coastal And Shelf

Science 33: 1-22

Lavery PS, McComb AJ. 1991. Macroalgal-sediment nutrient interactions and their

importance to macroalgal nutrition in a eutrophic estuary. Estuarine, Coastal

and Shelf Science 32: 281-95

Lean DRSaP, F. R. 1981. Photosynthetic response of lake plankton to nutrient

enrichment: a test for nutrient limitation. Limnology and Oceanography 26:

1001-19

Lee V, Olsen S. 1985. Eutrophication and Management Initiatives for the Control of

Nutrient Inputs to Rhode-Island Coastal Lagoons. Estuaries 8: 191-202

Lin CK. 1989. Prawn culture in Taiwan. What went wrong? World Aquaculture 20:

19-20

Lin CK, Ruamthaveesub P, Wanuchsoontorn P. 1993. Integrated culture of the green

mussel (Perna viridis) in wastewater from an intensive shrimp pond: concept

and practice. World Aquaculture 24: 68-73

Lindau CW, Delaune RD, Patrick WHJ, Lambremont EN. 1989. Assessment of stable

nitrogen isotopes in fingerprinting surface water inorganic nitrogen sources.

Water Air and Soil Pollution 48: 489-96

Lobban CS, Harrison PJ. 1994. Seaweed Ecology and Physiology. Cambridge:

Cambridge University Press

Longmore AR, Heggie DT, Flint R, Cowdell R, Skyring GW. 1999. Impact of runoff

on nutrient patterns in northern Port Phillip Bay, Victoria. AGSO Journal of

Australian Geology and Geophysics 17: 203-10

Madansky A. 1988. Prescriptions for Working Statisticians. Berlin: Springer-Verlag.

34-6 pp.

Maher WA, Cullen PW, Norris RH. 1994. Framework for designing sampling

programs. Environmental Monitoring and Assessment 30: 139-62

Malakoff D. 1998. Death by suffocation in the Gulf of Mexico. Science 281: 190-2

Mariotti A. 1983. Atmospheric nitrogen as a reliable standard for natural 15N

abundance measurements. Nature: 685-7

Mariotti A, Germon JC, Hugert P, kaiser P, Letolle A, et al. 1981. Experimental

determination of nitrogen kinetic isotope fractionation: some principles;

illustration for the denitrification and nitrification processes. Plant and Soil 62:

413-30

Mariotti A, Landreau A, Simon B. 1988. Nitrogen-15 isotope biogeochemistry and

natural denitrification process in groundwater: Application to the chalk aquifer

of northern France. Geochimica Et Cosmochimica Acta 52: 1869-78

141

Mariotti A, Letolle R. 1977. Applicatio del'etude isotopique de l'azote en hydrologie

et hydrogeologie - Analyse des resultats obtenus sur un exemple precis: le

Basin de Melarchez (Seine-et-Marne, France). Journal of Hydrology 33: 157-

72

MBCWQMST. 1999. The Crew Members Guide to the Health of our Waterways.

Brisbane: Moreton Bay Catchment Water Quality Management Strategy

Team. 97 pp.

McClelland JW, Valiela I, Michener RH. 1997. Nitrogen-stable isotope signatures in

estuarine food webs: A record of increasing urbanization in coastal

watersheds. Limnology and Oceanography. July 42: 930-7

Moriarty DJW. 1997. The role of microorganisms in aquaculture ponds. Aquaculture

151: 333-49

Naylor RL, Goldburg RJ, Mooney H, Beveridge M, Clay J, et al. 1998. Nature's

subsidies to shrimp and salmon farming. Science 282: 883-4

Naylor RL, Goldburg RJ, Primavera JH, Kautsky N, Beveridge MCM, et al. 2000.

Effect of aquaculture on world fish supplies. Nature London 405: 1017-24

Nixon SW. 1986. Nutrients and the productivity of estuarine and coastal marine

ecosystems. Journal of the Limnological Society of South Africa 12: 43-71

Nixon SW. 1995. Coastal marine eutrophication: A definition, social causes, and

future concerns. Ophelia 41: 199-219

Nixon SW, Pilson MEQ. 1983. Nitrogen in Estuarine and Coastal Marine Systems. In

Nitrogen in the Marine Environment, ed. EJ Carpenter, DG Capone, pp. 565-

648. New York: Academic Press

Norris RH, Norris KR. 1995. The need for biological assessment of water quality:

Australian perspective. Australian Journal of Ecology 20: 1-6

NRC. 1993a. Managing Wastewater in Coastal Urban Areas. Washington, DC.:

National Academy Press

NSWSPCC. 1985. Priority issues involved in the diffuse pollution of waterways

especially by agricultural chemicals. New South Wales, State Pollution

Control Commission, Sydney 06

O'Donohue MJ, Dennison WC. 1997. Phytoplankton productivity response to nutrient

concentrations, light availability and temperature along an Australian estuarine

gradient. Estuaries 20: 521-33

O'Donohue MJ, Glibert PM, Dennison WC. 2000. Utilization of nitrogen and carbon

by phytoplankton in Moreton Bay, Australia. Marine and Freshwater

Research 51: 703-12

O'Donohue MJ, Heil CA, Lowe S, Horrocks JL, Abal EG, Costanzo SD and Dennison

WC. 1997. Assessing the impact of a flood event on Moreton Bay using

marine plants as bioindicators of water quality. In Moreton Bay and

Catchments, ed. IR Tibbetts, NJ Hall, WC Dennison, pp. 585-98. Brisbane:


Paerl HW, Bowles ND. 1987. Dilution bioassays: Their application to assessments of

nutrient limitation in hypereutrophic waters. Hydrobiologia 146: 265-74

Parsons TR, Maita Y, Lalli CM. 1984. A manual of Chemical and Biological Methods

for Seawater Analysis. Toronto: Pergamon Press. 173 pp.

Peckol P, Demeo Anderson B, Rivers J, Valiela I, Maldonado M, Yates J. 1994.

Growth, nutrient uptake capacities and tissue constituents of the macroalgae

Cladophora vagabunda and Gracilaria tikvahiae related to site-specific

nitrogen loading rates. Marine Biology Berlin 121: 175-85

143

Pennock JR. 1987. Temporal and spatial variability in phytoplankton ammonium and

nitrate uptake in the Delaware Estuary (USA). Estuarine Coastal And Shelf

Science 24: 841-58

Peterson BJ, Fry B. 1987. Stable Isotope in Ecosystem Studies. Annual Review of

Ecology and Systematics: 293-320

Preston NP, Jackson C, Thompson PA, Austin M, Burford M. 2000. Prawn farm

effluent: composition, origin and treatment. Fishing Research and

Development Corporation Final Report 95/162, FRDC, Canberra

Pruder GD. 1992. Marine shrimp effluent: characterisation and environmental impact.

In Proceedings of the Special Session on Shrimp Farming, ed. J Wyban, pp.

187-94. Baton Rouge: World Aquaculture Society

Reddy KR, Patrick WH. 1976. Effect of frequent changes in aerobic and anaerobic

conditions on redox potential and nitrogen loss in a flooded soil. Soil Biology

and Biochemistry 8: 491-5

Redfield AC, Ketchum BH, Richards FA. 1963. the influence of organisms on the

composition of seawater. In The Sea, ed. MN Hill. New York: Wiley

Interscience

Ryther JHaD, W. M. 1971. Nitrogen, Phosphorus, and Eutrophication in the Coastal

Marine Environment. Science 171: 1008-13

Samocha TM, Lawrence AL. 1997. Shrimp farms' effluent waters, environmental

impact and potential treatment methods. In Interactions between cultured

species and naturally occurring species in the environment, ed. BJ Keller, pp.

35-8

Sansanyuth P, Phadungchep A, Ngammkontha S, Ngdngam S, Sukasem P, et al. 1996.

Shrimp pond effluent: Pollution problems and treatment by constructed

wetlands. Water Science and Technology 34: 93-8

Sharp JH. 1983. The distributions of inorganic nitrogen and dissolved and particulate

organic nitrogen. In Nitrogen in the Marine Environment, ed. EJ Carpenter,

DG Capone. New York: Academic Press

Singh B, Sekhon GS. 1979. Nitrate pollution of groundwater from farm use of

nitrogen fertilisers - a review. Agriculture and the Environment 4: 207-25

Smayda TJ. 1990. Novel and nuisance phytoplankton blooms in the sea: Evidence for

a global epidemic. In Toxic Marine Phytoplankton, ed. E Graneli, B

Sundstrom, L Elder, DM Anderson, pp. 29-40. New York: Elsevier

Smith Evans M, Dawes A. 1997. Early experiences in monitoring the effects of Hong

Kong's new generation of sewage outfalls on the marine environment. Marine

Pollution Bulletin. July Dec 33: 317-27

Smith SM, Hitchcock GL. 1994. Nutrient enrichments and phytoplankton growth in

the surface waters of the Louisiana bight. Estuaries 17: 740-53

Smith VH, Tilman GD, Nekola JC. 1999. Eutrophication: Impacts of excess nutrient

inputs on freshwater, marine, and terrestrial ecosystems. Environmental

Pollution 100: 179-96

Sweeney RE, Kalil, E. K. and Kaplan, I. R. 1980a. Characterisation of domestic and

industrial sewage in southern California sediments using nitrogen, carbon,

sulphur and uranium tracers. Marine Environmental Research 3: 225-43

Sweeney RE, Liu KK, Kaplan IR. 1978. Oceanic nitrogen isotopes and their uses in

determining the sources of sedimentary nitrogen. In Stable Isotopes in the

Earth Sciences. Bulletin No. 20, pp. 9-26. Wellington: New Zealand

Department of Science and Industrial Research

Sweeney RE, Kaplan IR. 1980. Tracing flocculent industrial and domestic sewage

transport on San Pedro Shelf, southern California, by nitrogen and sulphur

isotope ratios. Marine Environmental Research 3: 215-24

145

Teichert CDR, Rouse DB, Potts A, Boyd CE. 1999. Treatment of harvest discharge

from intensive shrimp ponds by settling. Aquacultural Engineering 19: 147-61

Thompson PA, Hosja W. 1996. Nutrient limitation of phytoplankton in the upper

Swan River Estuary, Western Australia. Marine and Freshwater Research 47:

659-67

Troell M, Ronnback P, Halling C, Kautsky N, Buschmann A. 1999. Ecological

engineering in aquaculture: Use of seaweeds for removing nutrients from

intensive mariculture. Journal of Applied Phycology 11: 89-97

Udy JW, Dennison WC. 1997a. Growth and physiological responses of three seagrass

species to elevated sediment nutrients in Moreton Bay, Australia. Journal of

Experimental Marine Biology and Ecology. Oct. 217: 253-77

Udy JW, Dennison WC. 1997b. Physiological responses of seagrasses used to identify

anthropogenic nutrient inputs. Marine and Freshwater Research 48: 605-14

Underwood AJ. 1981. Techniques of analysis of variance in experimental marine

biology and ecology. Oceanography and Marine Biology: An Annual Review

19: 513-605

Uusitalo R, Yli Halla M, Turtola E. 2000. Suspended soil as a source of potentially

bioavailable phosphorus in surface runoff waters from clay soils. Water

Research. 34: 2477-82

Valiela I. 1995. Marine Ecological Processes. New York, USA: Springer Verlag

Van Donk E, Veen A, Ringelberg J. 1988. Natural community bioassays to determine

the abiotic factors that control phytoplankton growth and succession.

Freshwater Biology 20: 199-210

Viessman Jr. W, Hammer MJ. 1998. Water Supply and Pollution Control. Reading,

MA: Addison-Wesley

Vitousek PM. 1994. Beyond global warming: Ecology and global change. Ecology

Tempe 75: 1861-76

Vitousek PM, Aber JD, Howarth RH, Likens GE, Matson PA, et al. 1997. Human

alteration of the global nitrogen cycle: Source and consequences. Ecological

Applications 7: 737-50

Vitousek PM, Howarth RW. 1991. Nitrogen Limitation on Land and in the Sea - How

Can It Occur. Biogeochemistry 13: 87-115

Vitousek PM, Mooney, H. A., Lubchenco, J. and Melillo, J. M. 1997. Human

Domination of Earth's Ecosystems. Science 277: 494-9

Voss M, Struck U. 1997. Stable nitrogen and carbon isotopes as indicator of

eutrophication of the Oder river (Baltic sea). Marine Chemistry 59: 35-49

Wainright SC, Haney JC, Kerr C, Go lovkin AN, Flint MV. 1998. Utilization of

nitrogen derived from seabird guano by terrestrial and marine plants at St.

Paul, Pribilof Islands, Bering Sea, Alaska. Marine Biology Berlin. April 131:

63-71

Waser NA, Yu ZM, Yin KD, Nielsen B, Harrison PJ, et al. 1999. Nitrogen isotopic fractionation during a simulated diatom spring bloom: importance of N-starvation in controlling fractionation. Marine Ecology-Progress Series 179: 291-6

Wilson JG. 1988. The Biology of Estuarine Management. London: Croom Helm Ltd.

204 pp.

Wilson JG. 1994. The role of bioindicators in estuarine management. Estuaries 17:

94-101

Wolanski E, Spagnol S, Thomas S, Moore K, Alongi DM, et al. 2000. Modelling and

visualizing the fate of shrimp pond effluent in a mangrove-fringed tidal creek.

Estuarine Coastal and Shelf Science. Jan. 50: 85-97

147

Wolfe DA, Kjerfve B. 1986. Estuarine Variability: an overview. In Estuarine

Variability, ed. DA Wolfe, pp. 3-17. Orlando, Florida: Academic Press

Wolter C, Minow J, Vilcinskas A, Grosch UA. 2000. Long-term effects of human

influence on fish community structure and fisheries in Berlin waters: An urban

water system. Fisheries Management and Ecology. Feb. 7: 97-104

Wu RSS. 1999. Eutrophication, water borne pathogens and xenobiotic compounds:

Environmental risks and challenges. Marine Pollution Bulletin 39: 11-22

Zann LP. 1995. Our sea, our future: major findings of the State of the Marine

environment Report for Australia, Great Barrier Reef marine Park Authority,

Townsville, Australia

Zar JH. 1984. Biostatistical analysis: Prentice-Hall International, Inc.

development of indicators for assessing and …development of indicators for assessing and...

Documents