CSIRO OCEANS & ATMOSPHERE

Peter Thompson, Todd O'Brien, Hans Paerl, Benjamin Peierls, Paul Harrison, Malcolm Robb.

Phytoplankton and hydrology: in a changing world

Motivations

• Cloern and Jassby (2010) asked “Why does phytoplankton biomass fluctuate mildly in some places and wildly in others?”

Australia

• “Low, erratic rainfall patterns over much of the country combined with small coastal catchments and high evaporation rates mean that annual discharges from Australian rivers are the lowest and most variable in the world” (McMahon, 1982).

PETER THOMPSON CSIRO.

Water and Global Change (EU FP6)

Australia: Low runoff and high inter annual variability

Swan Estuary

• The Swan Estuary has a catchment area of ~ 121,000 km2

PETER THOMPSON CSIRO.

• Eutrophic

• Productive

• Blooms • Karlodinium

veneficum

• Microcystis aeruginosa

Swan River sampling stations

Phytoplankton sampling (+ other water quality parameters) is ~ weekly since 1994 at 10 stations.

Indian Ocean

Climate

PETER THOMPSON CSIRO.

SW Australia: Long term drying trend

• Rainfall down ~ 100 mm relative to long term mean

• Rainfall has considerable inter annual variability but shows a significant long term decline.

PETER THOMPSON CSIRO.

Swan River: Cyanobacteria

Anomalously high mid summer rain encourages blue greens, mostly: Microcystis aeruginosa

Anabaena circinalis

SCOR WG 137 Diverse group from around the planet

Data managed by Todd O’Brien (NOAA)

Basic data analysis is available online using tools Todd has built.

A subset of the WG investigated hydrology. The diversity of sites, methods and data was both a challenge and a strength.

DPSIR model (1979)

• States (for fossil fuels) • CO2, pH, temperature, precipitation,

flow, residence time, salinity, nutrients

• Impacts (for precipitation) • Dilution

• Advection

• Stratification potentially leading to variation in turbulence,

• Mixed layer depth and therefore irradiance

• Growth

• Grazing

• Covariates: temperature and insolation

• Responses • biomass, taxa

11 |

Pressures

State

Impacts

Response

Driver

Human Population growth

Clearing Impoundment Eutrophication Fossil Fuels

Phytoplankton in the coastal zone

tim

e

IPCC 2014 Jun – Jul - Aug

Sites

• Other = regions where lag between precipitation and flow was expected to be very long in winter

• Precipitation data from NOAA’s Earth System Research Laboratory quality controlled precipitation data set based on 67,200 rainfall stations worldwide (Schneider et al., 2011).

Longitude

-150 -100 -50 0 50 100 150

Latit

ude

-40

-20

0

20

40

60

80

wet & wetterdry & dryerother

Time (month of year)

Jan

Feb

Mar

Apr

May

June July

Aug

Sep

t

Oct

Nov

Dec

Pre

cipi

tatio

n (m

m m

onth

-1)

0

50

100

150

200

250

Continent area Water body (number of sites

within)

Asia SE Hong Kong waters, (n = 10)

Australia SW Swan River estuary, (n = 5)

North

America

SW San Francisco Bay, (n = 7)

SE Neuse River and Pamlico

Sound, (n = 20)

NE Narragansett Bay, (n = 1)

NE Booth Bay, Maine, (n = 1)

NE Bay of Fundy, n = 5)

South

America

SE Patos Lagoon Estuary, (n = 1)

Europe N Skagerrak, Kattegat, (n=3)

NW North Sea, English Channel,

Irish Sea, (n = 8)

N Baltic Sea, (n = 37)

SW Guadiana Estuary, (n = 1)

SW Nervion River Estuary, (n = 2)

SW Bay of Biscay, (n = 5)

SW Mediterranean, (n = 5)

CSIRO.

Climate • Drying, yes….BUT this

often has a seasonal component • Strongest declines are:

– May (-0.8mm/y, p=0.02)

– June (-1.6mm/y, p=0.003)

– October (-0.38mm/y, p=0.02).

• Less in May, June and October = a longer dry season

January

time (year)

1940 1960 1980 2000 2020

0

20

40

60

80

100

120December

time (year)

1940 1960 1980 2000 2020

rain

fall (

mm

/mon

th)

0

20

40

60

80February

time (year)

1940 1960 1980 2000 2020

020406080

100120140160

July

time (year)

1940 1960 1980 2000 20200

100

200

300

400

500June

time (year)

1940 1960 1980 2000 2020

rain

fall (

mm

/mon

th)

0

100

200

300

400

500August

time (year)

1940 1960 1980 2000 20200

50

100

150

200

250

300

350

400

Oct

time (year)

1940 1960 1980 2000 2020

0

20

40

60

80

100

120

140Sept

time (year)

1940 1960 1980 2000 2020

rain

fall (

mm

/mon

th)

020406080

100120140160180

Nov

time (year)

1940 1960 1980 2000 2020

0

20

40

60

80

100

April

time (year)

1940 1960 1980 2000 2020

020406080

100120140160

March

time (year)

1940 1960 1980 2000 2020

rain

fall (

mm

/mon

th)

0

10

20

30

40

50

60

70May

time (year)

1940 1960 1980 2000 20200

50

100

150

200

250

Swan River

HAB and climate | Peter Thompson Hydrology and Phytoplankton 15 |

(A)Mayrainfall

1940 1960 1980 2000

Rai

nfal

l (m

m/m

onth

)

0

50

100

150

200

250

(B) June river flow

Time (year)

1980 1990 2000 2010

Riv

er fl

ow (m

3 /s)

103

104

105

106

(C)Swan EstuaryBlackwall Reach

0 50 100 150 200

Din

ofla

gella

tes

L-1

104

105

106

(D) Swan EstuaryArmstrong Spit

Rainfall (mm)

0 50 100 150 200D

inof

lage

llate

s L-1

104

105

106

107

• Lower precipitation in May and June

• Less river flow

• Fewer dinoflagellates at the most oceanic sites

Seasonal Patterns for dinoflagellates in the Swan Estuary

HAB and climate Hydrology and phytoplankton 16 |

Autumn Winter Spring

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

Swan River Estuary - S01

Blackwall Reach (Australia)

0.1758

p>0.10

-0.4540

p>0.10

-0.3563

p>0.10

0.4578

p<0.10 0.6478

p<0.01

0.5425

p<0.05

-0.4499

p<0.10

0.0144

p>0.10

0.2368

p>0.10

0.4502

p<0.10

-0.5314

p<0.05

-0.4574

p<0.10

Swan River Estuary - S02

Armstrong Spit (Australia)

0.1269

p>0.10

-0.1701

p>0.10

-0.3705

p>0.10

0.2901

p>0.10 0.6645

p<0.005

0.5712

p<0.05

-0.4180

p>0.10

0.1517

p>0.10

0.2619

p>0.10

0.4415

p<0.10

-0.2148

p>0.10

-0.4324

p<0.10

Swan River Estuary - S03

Narrows Bridge (Australia)

0.2169

p>0.10

0.1782

p>0.10

-0.2511

p>0.10

0.5257

p<0.05

0.3398

p>0.10

0.3677

p>0.10

-0.4763

p<0.10

0.2013

p>0.10

-0.1238

p>0.10

0.1360

p>0.10

-0.2840

p>0.10

-0.1213

p>0.10

Swan River Estuary - S04

Ron Courtney Island

(Australia)

-0.6306

p<0.10

-0.5524

p<0.10

-0.0883

p>0.10

-0.3465

p>0.10

-0.1696

p>0.10

-0.1622

p>0.10

-0.4776

p<0.10

0.0474

p>0.10

0.2833

p>0.10

-0.5128

p<0.05

-0.4302

p<0.10 -0.6477

p<0.01

Swan River Estuary - S05

Success Hill (Australia)

-0.5098

p>0.10

-0.5320

p<0.10

-0.1763

p>0.10

-0.3949

p>0.10

-0.1870

p>0.10

-0.2535

p>0.10

-0.2743

p>0.10

0.0533

p>0.10

-0.3302

p>0.10

-0.5514

p<0.05

-0.1061

p>0.10 -0.6865

p<0.005

1

Dinoflagellates were positively correlated with autumn and early winter precipitation at lower estuary sites. A drying climate is reducing these blooms. Is this true elsewhere?

Diatom example

• 25 sites in 4 regions

• 12 monthly time series

• ~ 300 correlations

• Probability of getting 40% positive slopes from 292 is very small.

(assumes a normal distribution and random observations)

Z Test for the Proportion

number of correlations 292

number of positive slopes 117

Sample Proportion 0.400685

Null Hypothesis p= 0.5

Standard Error 0.02926

a 0.05

Z Test Statistic -3.39419

Two-Tailed Test

Lower Critical Value -1.95996

Upper Critical value 1.959964

p-value 0.000688

Decision Reject

Timing of precipitation

HAB and climate | Peter Thompson Hydrology and Phytoplankton 18 |

All sites

Seasonwinter spring summer autumn

Prop

ortio

n of

pos

itive

tim

e se

ries

(%)

30

35

40

45

50

55

60

65 chladiatomsdinoschlorochryso eugleno r2 = 0.38

P=

0.0

2

P=

0.0

4

P=

0.0

00

9

P=

0.0

07

P=

0.0

06

P=

0.0

00

6

P=

0.0

3

P=

0.0

01

5

P=

0.0

01

5

• Generally there were positive responses during summer (P = 0.018)

• Winter

• Winter precipitation was negatively associated with chlorophyll a, diatoms and chrysophytes.

• For diatoms negative associations with precipitation were dominant in January & February.

• Spring

• Diatoms were negatively associated with precipitation in March and April while chlorophyte abundances increased with precipitation.

• Summer

• Chlorophyll a and Chlorophytes were positively associated with precipitation.

• Autumn

• Dinoflagellates were negatively associated with increased precipitation, similarly diatoms during October.

Wet and getting wetter

(A) Regions of increasing precipitation

chla diatoms dinos chlorophytes

prop

ortio

n of

pos

itive

tim

e se

ries

(%)

20

30

40

50

60

70

winter spring summer autumn

P=0.0007P

=0

.00

18

P=0.

0055

P=0.

021

P=0.

028

Chlorophyll a

responds positively to precipitation in autumn

Diatoms negative overall all seasons esp. spring

Dinos were mixed up in summer down in autumn

Regions of increasing precipitation

(B) diatoms by month

month of yeardec jan feb mar apr may jun jul aug sep oct nov

20

40

60

80

P=0.

03

P=0.

03

(A) Regions of decreasing precipitation

chla diatoms dinos chlorophytes

prop

ortio

n of

pos

itive

tim

e se

ries

(%)

30

40

50

60

70

80

90

winter spring summer autumn P <0.0008

P=0.

0001

P=0.

03

• Only chlorophytes showed a consistent response to more precipitation

• Over whole year (P=0.0008)

• Also during spring and summer

Drying regions

Drying regions: Chlorophytes by month

• 100% of sites showed a positive association of chlorophyte cell counts with precipitation in March.

(B) chlorophytes by month

month of yeardec jan feb mar apr may jun jul aug sep oct nov

20

40

60

80

100

P=0.

0005

P=0.

004

Using long term averaged values….

• Longest time series was 33 years of monthly sampling, n~ 396.

• A pattern of response for chlorophytes?

• (advection is unlikely to be the primary driver)

HAB and climate | Peter Thompson Hydrology and Phytoplankton 23 |

(A) Chlorophytes

salinity0 10 20 30

103

104

105

106

107

108

(C) Chrysophytes

Salinity0 10 20 30

103

104

105

106

107

2D Graph 1

0.5 1.0 1.5 2.0 2.5 3.0 3.5

Y D

ata

0.51.01.52.02.53.03.5

REPHY Parc Leucate (Mediterranean)REPHY Villefranche (Mediterranean)Thau Lagoon (Mediterranean)

(B) Dinoflagellates

0 10 20 30

Long

term

ave

rage

abu

ndan

ce (c

ells

L-1

)

102

103

104

105

106

107

108

North Sea (Baltic)Arhus Bugt (Baltic)Koge Bugt (Baltic)Hevring Bugt (Baltic)Ringkobing Fjord (Baltic)Nissum Fjord (Baltic)Nissum Bredning (Baltic)Logstor Bredning (Baltic)Skive Fjord (Baltic)Lister Dyb (Baltic)Alborg Bugt (Baltic)Anholt East (Baltic) Vejle Fjord (Baltic)Ven (Baltic)Arkona (Baltic)Mariager Fjord (Baltic)Horsens Fjord (Baltic)Roskilde Fjord (Baltic)Lillebaelt-South (Baltic)Lillebaelt-North (Baltic)Odense (Baltic)Gniben (Baltic)Storebaelt (Baltic)Bornholm Deep (Baltic)Swan River-Blackwall (Australia)Swan River-Armstrong (Australia)Swan RIver-Narrows (Australia)Swan RIver-Courtney (Australia)Swan River-Success (Australia)San Francisco-lower south (USA)San Francisco-mid south (USA)San Francisco-north south (USA)San Francisco-central bay (USA)San Francisco-San Pablo (USA)San Francisco-Suisun (USA)San Francisco-Sacramento (USA)Patos Lagoon (Brazil)REPHY Antifer (English Channel)REPHY At So (English Channel)REPHY Donville (English Channel)REPHY Pen (English Channel)REPHY Point SNR (English Channel)Bay of Fund-Brandy Cove (Canada)Bay of Fundy-Deadmans Harbour (Canada)Bay of Fundy-Lime Kiln Bay (Canada)Bay of Fundy-Passamaquoddy (Canada)Bay of Fundy-Wolves Is. (Canada)SMHI A17 (Sweden)SMHI Anholt East (Kattegat)SMHI Slaggo (Sweden)AZTI D2 (SE Bay of Biscay)Nervion River E1 (southern Bay of Biscay)Nevion River E2 (southern Bay of Biscay)REPHY Kervel (Bay of Biscay)REPHY Le Cornard (Bay of Biscay)REPHY Men de Roue (Bay of Biscay)REPHY Quest Loscolo (Bay of Biscay)REPHY Teychan Bis (Bay of Biscay)REPHY Diana Centre (Mediterranean)REPHY Lazaret (Mediterranean)

Some closing observations

• Generally abundance declines with salinity (nutrients?)

• Dinoflagellate abundance patterns were not consistent along different estuaries • Proximal cause is not precipitation or salinity

for dinos (stratification?)

• Way forward? • Dynamic mechanistic model?

• Improved statistical model (e.g. GAM)?

(A) Swan River Estuary

0 10 20 30102

103

104

105

(B) San Francisco Bay

0 10 20 30

Long

term

mea

n ce

ll den

sity

(cel

ls L

-1)

102

103

104

105

106

107

chlorophytes diatoms dinoflagellates

(C) Neuse River

0 10 20 30Long

term

mea

n pi

gmen

t (µg

L-1

)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

fucoxanthinperidininchlorophyllb

(D) All sites, r2 = 0.1, P = 0.001

Long term mean salinity0 10 20 30 40

Long

term

mea

n ch

la(u

g/L)

0

5

10

15

20

Conclusions

• Increasing winter precipitation is generally negative for phytoplankton biomass

• Winter and spring diatom blooms are susceptible to increased precipitation

• Drying ecosystems will experience less biomass and fewer chlorophytes

HAB and climate | Peter Thompson Hydrology and Phytoplankton 25 |

Pressures

State

Impacts

Response

Driver

Human Population growth

Clearing Impoundment Eutrophication Fossil Fuels

Phytoplankton in the coastal zone