development of indicators for assessing and …development of indicators for assessing and...
TRANSCRIPT
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
The University of Queensland
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
4.2.1 Site Location 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
5.2.1 Site Location 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).
32 Mapping Sewage Impacts
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
34 Mapping Sewage Impacts
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
36 Mapping Sewage Impacts
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).
38 Mapping Sewage Impacts
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.
40 Mapping Sewage Impacts
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
42 Mapping Sewage Impacts
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.
46 Phytoplankton Bioassays
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).
48 Phytoplankton Bioassays
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.
50 Phytoplankton Bioassays
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.
52 Phytoplankton Bioassays
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
derived from the following equation:
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
54 Phytoplankton Bioassays
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 Phytoplankton Bioassays
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)
56 Phytoplankton Bioassays
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.
58 Phytoplankton 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
60 Phytoplankton Bioassays
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).
3.4.2 Phytoplankton Bioassays
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
62 Phytoplankton Bioassays
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.
64 Phytoplankton Bioassays
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).
66 Phytoplankton Bioassays
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.
70 Assessing Influences of 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.
72 Assessing Influences of Shrimp Farm Effluent
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.
74 Assessing Influences of Shrimp Farm Effluent
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:
Light PRI = (FC – FO) / TC
Chapter 4 75
where FC is maximum control fluorescence, FO is initial fluorescence and TC is time to
maximum control fluorescence.
Nutrient stimulated responses were inferred growth of phytoplankton above that of the
control phytoplankton growth. Nutrient Phytoplankton Response Index (NPRI) was
derived from the following equation:
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
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
76 Assessing Influences of Shrimp Farm Effluent
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).
78 Assessing Influences of Shrimp Farm Effluent
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 Biological Indicators
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
80 Assessing Influences of Shrimp Farm Effluent
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
4.3.2.1 Phytoplankton Bioassays
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).
82 Assessing Influences of Shrimp Farm Effluent
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.
84 Assessing Influences of Shrimp Farm Effluent
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 Biological Indicators
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
86 Assessing Influences of Shrimp Farm Effluent
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.
4.4.2.2 Phytoplankton Bioassays
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,
88 Assessing Influences of Shrimp Farm Effluent
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
94 Nutrient Influences in a Sub-Tropical River Estuary
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.
96 Nutrient Influences in a Sub-Tropical River Estuary
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
98 Nutrient Influences in a Sub-Tropical River Estuary
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 Biological Indicators
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.
5.2.4.2 Phytoplankton Bioassays
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
100 Nutrient Influences in a Sub-Tropical River Estuary
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:
Light PRI = (FC – FO) / TC
where FC is maximum control fluorescence, FO is initial fluorescence and TC is time to
maximum control fluorescence.
Nutrient Phytoplankton Response Index (NPRI) was derived from the following
equation:
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.
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
102 Nutrient Influences in a Sub-Tropical River Estuary
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
)
0 10 20 30Distance Upriver (km)
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 10 20 30Distance Upriver (km)
0
20
40
60
80
TN
(µM
)
TP
(µM
)
0 10 20 30Distance Upriver (km)
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).
104 Nutrient Influences in a Sub-Tropical River Estuary
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
106 Nutrient Influences in a Sub-Tropical River Estuary
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
Distance Upriver (km)
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
108 Nutrient Influences in a Sub-Tropical River Estuary
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
Distance Upriver (km)
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
110 Nutrient Influences in a Sub-Tropical River Estuary
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
112 Nutrient Influences in a Sub-Tropical River Estuary
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.
114 Nutrient Influences in a Sub-Tropical River Estuary
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(~30–20 fsu.d-1)
Higher in rivers than bay
West-east gradient(~ 10 – 3 ‰)
Defined zone of sewage influence – 10 km
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
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(~30–20 fsu.d-1)
Higher in rivers than bay
West-east gradient(~ 10 – 3 ‰)
Defined zone of sewage influence – 10 km
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 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)
Contained within creek
High to low• NH4
+
(~ 40 – 1µM)
• Chlorophyll (~10 – 1µg/L)
Phys / Chem
Peak in mid-creek(~ 0-20 fsu.d -1)
Phytoplankton Bioassays
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’
Contained within creek
Low • NH4
+
(~ 1.5 µM)
• Chlorophyll(~ 1.0 µg/L)
Contained within creek
High to low• NH4
+
(~ 40 – 1µM)
• Chlorophyll (~10 – 1µg/L)
Phys / Chem
Peak in mid-creek(~ 0-20 fsu.d -1)
Phytoplankton Bioassays
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
Incubated Macroalgae
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)
Phytoplankton Bioassays
Elevated either side of light response in mid-upper river(~50 fsu.d-1)
δ15N defined urban (~ 5 ‰) vs. agricultural (~3.5 ‰)
δ15N defined urban (~ 7 ‰) vs. agricultural (~4.5 ‰)
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)
Phytoplankton Bioassays
Elevated either side of light response in mid-upper river(~50 fsu.d-1)
δ15N defined urban (~ 5 ‰) vs. agricultural (~3.5 ‰)
δ15N defined urban (~ 7 ‰) vs. agricultural (~4.5 ‰)
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 SignaturesTweed RiverNitrate
Chlorophyll
Secchi Depth
Temperature
Mangrove δ15N
Incubated Macroalgae
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:
The University of Queensland
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.