River ScienceSeptember 2000

Page 1

Issue 1, February 2000

The Microcystis bloom inthe Swan River in February2000 (Photo courtesyDennis Sarson).

A week later, rain-laden clouds left over from

Cyclone Steve came down from the north anddeposited over 12 000 million tonnes, i.e. 12 000

gigalitres (GL) of water on the Avon River catch-

ment, an area the size of Tasmania. This resulted inover 270 GL entering the estuary (Figure 1).

This was enough water to fill the Swan-Canning

rivers and estuary five and half times over. The soil

was already saturated from the earlier January rain,so there was a rapid and huge runoff particularly

from the normally dry Lockhart subcatchment and

Unseasonable freshwater flow

The cause of the unusual and massive toxic blue-

green algae bloom in the Swan and lower Canning

rivers in February 2000, can be largely explained bytwo unseasonably large rain events that occurred in

January 2000.

The rain that fell in early January gave much of the

hinterland of the Swan-Avon River system a good

soaking. This resulted in the catchment being unableto absorb much more rain.

Conditions that made the river green and potentially toxic

CONTENTS

Conditions that made theriver green and potentiallytoxic .............................. 1

Factors favouring growthof the toxicblue-green algaeMicrocystis ................... 4

A Microcystispopulation explosion ..... 5

Efforts to control theMicrocystis bloom ......... 6

An action plan for therivers’ future .................. 8

Acknowledgments ........ 8

‘Summer surprise’The Swan River blue-green algal bloom

February 2000

‘Summer surprise’The Swan River blue-green algal bloom

February 2000

Issue 2, September 2000

TH

E

GOVERNMENTO

F

WE

S T E R N A U S T R

AL

I A

2

River ScienceSeptember 2000

Page 2

Yenyenning Lake system, about 200 km south east

of Perth, in the Avon catchment. It was the first time

waterways in the Lockhart subcatchment had flowedsignificantly in the summer for forty years. Even

during the winter, this area usually does not receive

sufficient rain to contribute significant runoff to theSwan-Avon River system.

The source of nutrientsfor algal growth

The huge torrent of freshwater came down the Swan-

Avon in two pulses, bringing with it high levels ofnutrients, mainly nitrogen (N) and phosphorus (P).

The nutrients came from an accumulation of animalmanures, plant material, soil particles with nutrients

attached and fertilisers washed into the creeks and

rivers. Nitrogen and phosphorus are the two essentialingredients required for algal growth. Nutrient levels

were as high as 7 mg/L for total nitrogen and

between 0.2 to 0.3 mg/L for total phosphorus. Thenitrogen levels were seven times the limit considered

healthy for freshwater entering estuaries and two to

three times the limit considered healthy forphosphorus.

This nutrient rich freshwater, which carried over

800 tonnes of nitrogen and 35 tonnes of phosphorus,flowed into the wide, shallow estuary of the Swan

Yenyenning Lakes flooding.Flooding at Beverley, Avon River.

Figure 3. Satellite image of the catchment of the Swan-Avon River in late January 2000, showing waterways inthe Lockhart subcatchment and Yenyenning Lakes inblue, which are full and flowing. The Swan Estuary is tothe upper left of the image.

Figure 2. Rainfall (mm) inthe Swan-Avon Rivercatchment for January2000 compared to theaverage rainfall inJanuary in brackets.

Figure 4. Total Nitrogen concentrations measuredwhere the Avon River enters the Swan-Canning Estuaryat Walyunga Canyon in the Darling Range.

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

Tota

l Nitr

ogen

(mg/

L)

Aug

87

Jul 9

1

Jun

95

Jun

99

Feb

88

Jan

92

Dec

95

Dec

99

Aug

88

Jul 9

2

Jun

96

Jun

00

Feb

89

Jan

93

Dec

96

Jul 8

9

Jul 9

3

Jun

97

Jan

90

Jan

94

Dec

97

Jul 9

0

Jul 9

4

Jun

98

Jan

91

Jan

95

Dec

98

Figure 1. Swan-AvonRiver flow (gigalitres perday) measured atWalyunga Canyon – startof the Swan-CanningEstuary, showing normalsummer and winter flowand the unseasonal flowthat occurred in January-February 2000. 0

5

10

15

20

25

Jan Feb March April May June July Aug Sept Oct Nov Dec

1994199519961997199819992000 (incomplete

record)

Flow

(GL/

day)

River ScienceSeptember 2000

Page 3

Sources of algae are plentiful

The Swan River has hundreds of naturally occurring

microscopic algae species (known collectively asphytoplankton) most of which are not harmful to

humans or animals. Indeed many algae species are

important sources of food for estuarine invertebrates,such as worms, prawns and mussels. The phyto-

plankton also supports the zooplankton (microscopic

animals) which in turn are food for many fish species.In addition to the algae already in the river, reservoirs

of all types of algae are present in the drains, ponds,

lakes and wetlands connected to the Swan-Canningsystem. These can be flushed into the river by rain.

By considering the timing of detection and progress

downstream it is believed that the strain ofMicrocystis that bloomed in February 2000 was

most likely to have originated from stagnant pools

in the Avon Catchment.

River where the water was relatively clear, calm

and warm. This combined with lots of sunlight,

clear skies and long day length, provided idealconditions for photosynthesis and algal growth.

Normally the increased nutrient levels caused by

freshwater runoff from the Avon catchment occurin winter, when conditions for algal growth are not

favourable; water temperatures are low, day length

is short and the river is more turbulent due to stormsand strong flow.

In February 2000, dissolved inorganic nitrogen,

which is instantly available nutrition for algalgrowth, reached 0.5 to 1.2 mg/L in the water between

Success Hill and Blackwall Reach. This was five to

12 times higher than normal for summer conditionsin the estuary.

The river’s algae go through typical seasonal bloom

cycles. Under normal summer river conditions

various algae bloom and die. The extent of anyparticular blooms depends largely on the availability

of the nutrients nitrogen (N) and phosphorus (P)

and the salinity of the water. Different algae preferdifferent conditions so a number of factors will

determine which algae will be the most successful

and bloom at any particular time.

The Microcystis bloom was only a temporary

aberration of the normal seasonal cycle shown in

Figure 6. In a normal year beneficial spring bloomsoccur in response to nutrients washed in by the late

winter and early spring rains.

When the spring bloom collapses due to exhaustionof nitrogen supply the algae die and fall to the river

bottom. Summer blooms of mainly dinoflagellates

use N and P derived from the sediment and urbandrains.

With the onset of autumn rains nutrients are once

more washed into the river system and more algalblooms occur. In winter, low water temperature and

short days (less sunlight for photosynthesis)

significantly reduce algal growth.

Figure 5. Median dissolved inorganic nitrogen (DIN)concentrations during the Microcystis bloom. Medianvalues are compared with the bloom values (February2000) and other values usually observed in spring andsummer.

Distance from estuary mouth at Fremante Harbour (km)

Conc

entra

tion

DIN

(mg/

L)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

7 12 18 23 28 30 35 37 39

SummerSpringFebruary 2000

Figure 6. The relative abundance of phytoplankton and the scale of the blooms in theupper Swan River by the seasons of the year and the timing of the salt wedge. TheANZECC guideline for blooms indicates the relative magnitude of cell numbers occurringin the upper Swan. The freshwater deluge that stimulated the toxic Microcystisaeruginosa bloom in February 2000 is clearly seen in the summer.

River ScienceSeptember 2000

Page 4

Factors favouring growth of the toxicblue-green algae MicrocystisSalinity and the normal riversituationDuring winter, the Swan River upstream of theCauseway is flushed with freshwater. Downstreamof the Causeway, the water becomes stratified withfresh to brackish water over deeper saline waters.During the spring and summer, the freshwater flowslows and almost stops. This allows seawater(salinity 35 parts per thousand) from the IndianOcean and lower estuary to move into the upperestuary in what is known as a ‘salt wedge’. Seawateris denser than freshwater, so initially the wedgemoves upstream underneath the freshwater. As thesummer progresses, tidal exchange with seawaterworks to extend wedge conditions and then mixingoccurs throughout the water column. This producessalinities from 25 parts per thousand in the extremeupper regions to 35 parts per thousand in the lowerregions of the Swan Estuary, typical summerconditions for the Swan.

a number of algae already present in the system butin relatively low abundance.

Why Microcystis?One of these algae – a blue-green – Microcystisaeruginosa, scientifically known as a cyanobacteriaand referred to as Microcystis for short, doesespecially well in freshwater and low salinityconditions. These were the very conditions thatoccurred in the Swan River in February 2000.

Microcystis has several characteristics that give itan advantage over its algal competitors who alsolike low salinity conditions (algal species are alwayscompeting for nutrients, sunlight and space). Themost important of these are:

• It has an extremely buoyant nature. Microcystis

has a very effective gas vacuole system that helpsit to float to the very top of the water and receive

strong sunlight.

• This buoyancy, which allows it to photosynthesiseefficiently, results in the production of large

amounts of sugar. After a few hours of photo-

synthesis in the sun, Microcystis cells get soheavy with sugar they sink below the surface

where they consume the sugar they have produced.

Once they exhaust their sugar supply they float tothe surface again, ready to use the next day’s first

sunlight. If conditions are calm, this floating

characteristic gives Microcystis an instantadvantage over other algae.

A combination of Microcystis’ naturalcharacteristics, the unseasonable flood of freshwaterand nutrients that entered the system and the warm,sunny summer conditions allowed it to growspectacularly and produce large colonies of brightblue green cells and extensive scums.

Microscope photographof Microcystisaeruginosa.

Figure 7. Surface and bottom salinity at Melville Waterin the middle region of the Swan River during the bloomand afterwards.

Salinities in Melville Water for December 1999-June 2000

Salin

ity(p

pt)

Surface SalinityBottom Salinity

0

5

10

15

20

25

30

35

40

Dec January February March April May June

Salinity and the unusual summerrain conditionsThe unusual freshwater flush that occurred inFebruary lowered the salinity level in the Swan to 5to 10 parts per thousand – it was fresh in the upperreaches and only slightly saline in the middle andlower estuary. Water in the middle and lower estuarywas highly stratified with near freshwater on thesurface and more salty water on the bottom. Thislevel of salinity (fresh to low) and the huge input ofnutrients (N and P) in runoff (mainly from theLockhart River subcatchment, see Figure 3) favoured

River ScienceSeptember 2000

Page 5

For over five years the Swan River Trust has regu-larly measured the concentration of various algae,i.e. numbers per millilitre (mL), at nine sites aroundthe river.

Soon after the freshwater flowed into the estuarinesystem in February 2000, water samples showedconcentrations of Microcystis beginning to riserapidly. Under favourable conditions their growthis exponential, with the number of cells doubling atleast every day. Health warnings were posted assoon as the concentration of Microcystis reached20 000 cells per mL, the health limit consideredsafe for the recreational use of the river. In somebeach and sheltered locations Microcystisconcentrations on the surface reached 130 millioncells per mL, producing bright green scums.However, in most of the open more turbulent areasof the river, cell counts were below the health alertlevel where whole water column samples producedcounts varying between 5000 and 14 000 cellsper mL.

At the peak of the bloom, the public was warned notto have contact with the water and both the riversand estuary were closed for 12 days. The SwanRiver was closed between Success Hill andFremantle Port and the Canning River between theCanning and Mt Henry Bridges. Scums andaccumulations of algae were worst on the northernshores of the estuary, especially Claisebrook Cove,Crawley-Matilda Bay and Freshwater Bay. Scumswere also present but occurred less frequently onsome south shore beaches, such as Deepwater Bay.Fortunately they did not last long.

Hazard versus risk of the toxicbloomThe levels of microcystin, the major toxin producedby Microcystis, varied considerably during the bloomand exceeded the World Health Organisation’srecommended level (0.5 µg/L) during one period,when concentrations reached 8 µg/L. Bio-assay testsindicated that as little as 20 to 25 mg/kg of mammalweight of the algae was fatal for mice. Mice areoften used as an indicator mammal species to warnhealth authorities that a substance may be toxic tohumans. As a comparison DDT can kill mice at80 mg/kg.

While the toxicity of this strain of Microcystis washigh and therefore the hazard it represented washigh, the risk to humans was actually low. This isbecause the chance of contact or ingestion of a toxicquantity of Microcystis or tainted water was

considered low. Only if recreational activities wereundertaken in the sheltered bays where toxic scumswere located would the risk increase. Given thenumber of recreational activities for this time ofyear, authorities felt they could not chance anyonein the public becoming seriously ill.

A Microcystis population explosion

Toxic algae sign with City of Perth in the background

Figure 8. The area of the river and estuary where the Microcystis aeruginosa scumswere most prevalent during the bloom in February 2000. A toxic blue-green Anabaenabloom in the Canning River above the Kent Street Weir had just collapsed.

River ScienceSeptember 2000

Page 6

Controlling the bloomThe Swan River Trust and Water and RiversCommission tried a number of methods to reducethe Microcystis bloom. They included attempts toincrease the salinity of the surface layer of waterbeyond the tolerance of the algae, using a slurry ofclay particles mixed with a flocculating materialapplied in a fine spray on the water to try and sinkthe algae, and, skimming off high concentrations ofMicrocystis using oil spill equipment. Major publichealth risk was in areas of scum formation so theSwan River Trust and Water and Rivers Commissionfocussed control efforts on removing accumulationsof scum that had aggregated in various shelteredparts of the river. This helped prevent extensivescums of Microcystis forming on nearby shorelines

and popular recreational beaches where it couldhave remained a health hazard for weeks as the algalcells broke down releasing their toxins. It is estimatedthat over 900 tonnes of Microcystis was removedfrom the river and safely disposed of using sewagetreatment. Treatment of broad reaches of the riverwas not successful by any method. However, trialsusing fine clays mixed with the flocculating materiallooked promising and if blooms occur again, it maybe used at an earlier stage.

Management of the bloomThe response to the bloom was coordinated by theWater and Rivers Commission and the Swan RiverTrust. In the first few days when water samplesindicated rapidly growing colonies of toxic

Efforts to control the Microcystis bloom

Pumping saltwater from depth and spraying iton Microcystis concentrations to disperseaccumulations was also unsuccessful during thebloom in February 2000.

Spraying concentratedbrine solution on Micro-cystis concentrations todisperse accumulationswas unsuccessful duringthe bloom in February2000.

Skimming successfullyremoved concentratedMicrocystis scums fromthe Swan River during thebloom in February 2000.

River ScienceSeptember 2000

Page 7

Microcystis, an intensive sampling program wasimmediately established throughout the Swan-Canning rivers and estuary to monitor the algae’sgrowth, location and toxicity. For example, 20 lowerestuary and 35 to 41 beach sites were monitoredseveral times a week during the bloom. A crisismanagement team with active communication withother agencies held daily meetings to organisecontrol efforts, water sampling and testing and beachclean up efforts. The team also provided informationto the media, river users and the general communityon the current status of the bloom and its publichealth implications.

The collapse of the bloom

Natural salinity increases in the estuary caused theMicrocystis bloom to collapse. When the fresh-water flow began to subside towards the middle ofFebruary the normal summer seawater intrusionfrom the Indian Ocean into the Swan Estuary beganto re-establish itself. As the seawater progressed upthe estuary it raised the salinity above the critical10 parts per thousand level which was considered tobe the upper tolerance limit for this strain of Micro-cystis. The die off was first noticeable in the lowerestuary and quickly spread to the middle and upperestuary as the seawater mixed with freshwater. Bythe end of February the few remaining isolatedpockets of Microcystis in the upper reaches of theSwan River had also collapsed.

Using oil spill booms inMatilda Bay to concentrateand skim scums ofMicrocystis.

Figure 9. Sample sites,transect routes and beachobservation sites used tomonitor progress anddecline of the Microcystisbloom, February 2000.

PERTH

FremantleFremantle

MidlandMidland

PERTHWATER

MELVILLEWATER

RI

VER

CANNI N

G

IND IA N

OCEAN

N

Kilometres

0 5

Yule Brook

Canning

River

Swan

River

Microcystis

Beach observation sitesBoat observation sitesRoutine sampling sitesTransect pathway forsalinity and oxygen profiles

Response

Stirl

ing

Brid

ge

Wat

er P

olic

e(P

ier 2

1)

Blac

kwal

l Rea

ch

Fres

hwat

er B

ay

Poin

t Wal

ter

Luck

y Ba

y

JoJo

´s

Way

len

Bay

Mel

ville

Wat

er

Nar

row

s

Mat

ilda

Bay

Dots indicate actual sampling points

0 20 40 60 80 100

120

140

160

180

Swan River dissolved oxygen (% saturation)

14 February 2000

0 5 10 15 20 25Distance from Fremantle (km)

Dept

h fro

m s

urfa

ce(m

)

-10

-8

-5

-4

-2

0

100100

6060

6060

2020

2020

120120

Figure 10. Dissolvedoxygen profiles (water depthby site up the river to theNarrows) during the Micro-cystis bloom.

River ScienceSeptember 2000

Page 8

W RATER AND IVERSCOMMISSION

Restoring our natural treasure

The Western Australian Government, LocalGovernments and the Commonwealth Governmentare committed to implementing a program to reducethe amount of nutrients that find their way into theSwan-Canning system. By lowering nutrient levelsin the river, algal growth will be significantlyreduced. This will be accomplished throughintegrated catchment management and restoration,improved planning and land-use management,modifying river conditions to reduce algal bloomsfed by nutrients from the sediments, monitoring thehealth of the river and filling in critical gaps in ourknowledge that could help us better manage andprevent blooms.

There are two major parts to the integrated catchmentmanagement program. The first is the Swan-CanningCleanup Program (SCCP) Action Plan. The planaims to reduce nutrient inputs entering the Swanand Canning Rivers from urban development(domestic and industrial) and rural and semi-ruralland uses. The plan focuses on the Swan-Canningcoastal catchment (approximately 2 100 km2) whichis bound to the east by the Darling Range. Theintroduction of water sensitive design requirementsin regional and town planning and developmentschemes, improving land use activities, reducingdiffuse nutrient pollution and fostering communityled catchment restoration will be important aspectsof this plan.

The SCCP Action Plan is not designed to addressthe nutrient problems associated with inlandcatchments east of the Darling Range. These areascontributed the nutrients which caused the toxicalgal bloom in the Swan-Canning Rivers in February2000. This was a one in twenty year summer rainevent and a one in forty year summer flood event. Itis difficult to reduce nutrient losses from the landunder these unusual natural conditions. Long termefforts though, would reduce nutrient accumulationsor the potential for nutrients to be so easily washedinto the waterways. The Action Plan also does notdeal with the normal annual winter runoff that bringsa flush of nutrients into the river system from theAvon catchments east of the Darling Range. Theseinland catchments come under another program, theSwan-Avon Integrated Catchment ManagementProgram.

This second part of the Integrated CatchmentManagement Program involves the much larger

Avon catchment further upstream and east of theDarling Range (approximately 121 000 km2). Effortsto reduce nutrient inputs from this larger area arebeing addressed through initiatives that focus oncatchment restoration, revegetation, erosion, betterland use and reductions in, and more efficient useof, fertilisers. The West Australian Salinity ActionPlan is highly relevant to this program and directlyaffects this work.

The nutrient reduction programs that have been setin place are long term. Significant results may notbe seen for 10 to 20 years, but they are essential ifwe are to continue to improve the health of one ofour greatest natural assets – the Swan-Canning riversystem. By reducing nutrient inputs the chances ofour beautiful river being closed again because of atoxic bloom will be much less likely in the future.

Acknowledgments

This series is an initiative of the Aquatic ScienceBranch of the Water and Rivers Commission withfunding from the Swan-Canning Cleanup Program.This issue of River Science was written by TomRose, Rhys Brown and Malcolm Robb.

An action plan for the rivers’ future

For more informationSwan River TrustLevel 3 Hyatt Centre 87 Adelaide Terrace

East Perth Western Australia 6004

Telephone (08) 9278 0400

Facsimile (08) 9278 0401

Website www.wrc.wa.gov.au/srt

Tell us what you think of our publications at

www.wrc.wa.gov.au/public/feedback

ISSN 1443-4539

Printed on environmentally friendly paper

September 2000

Clay and a flocculating compound being sprayed toclump and sink Microcystis cells was partially successfulbut needed to be done earlier in the bloom.