Climate change and reduced freshwater flows are predicted to significantly degrade ecosystem states in the Coorong according to a new ecosystem response model. The model found that under future climate scenarios up to 46% of site-years could be in degraded states characterized by high salinity and reduced biodiversity. However, relatively small increases in freshwater flows, as achieved through initiatives like The Living Murray, can help mitigate these impacts. While no alternatives completely replace freshwater from the Murray River, additional water is needed to prevent the potential devastation of the Coorong ecosystems under climate change.

1. HEADLINE TO BE PLACED IN
THIS SPACE
CLLAMM Futures: Modelling
ecosystem responses to future scenarios
July 21st, 2009
Peter Fairweather & Rebecca Lester
Flinders University

2. Aims
• To develop an ecosystem response model
to predict future condition in the Coorong
• To model various possible future
scenarios:
• how will climate change & management affect
the ecosystem states of the Coorong?
• Today:
• a brief overview of approach & some of the
key findings

3. An ecosystem response model for
the Coorong

4. Variables included
Biological Environmental
• Bird abundances (UoA, DEH) • Flow parameters (MDBC)
• Commercial fish CPUE • Volumes, rates
(SARDI) • Lags from previous years
• Fish abundances (SARDI) • Hydrology (CSIRO)
• Macrobenthos abundances • Level above AHD, Depths
(FU) • Tidal influence, Lags
• Macrophyte prevalence • Meteorological (BoM)
(UoA) • Water quality (CSIRO, DEH)
• Salinity, [Nutrient], Turbidity

5. Alternating data sets approach
Biological Environmental
1. Identify
clusters
2. Identify
differing
conditions
3. Confirm
distinct states
4. Evaluate new
cases
5. Characterise
states

6. Ecosystem States
of the Coorong True False
CLLAMMecology
Hypersaline basin Marine basin

7. What do these states look like?
Estuarine/marine Degraded marine
CLLAMMecology

8. Environmental characteristics
MM + NL SL
DM
EM M UM HH AH UH DH
Low n
Days since
flow
Flow volume
Salinity
Tidal
influence
[TKN] na
[P] na
Turbidity na
♦ Very low ♦ Low ♦ Med ♦ High ♦ Very high
na = no data available

9. Biological characteristics MM + NL SL
DM
EM M UM HH AH UH DH
Low n
Fishing birds
Shorebirds
Waterfowl
Estuarine fish
Marine fish
Benthic
inverts
na
Ruppia na na na
♦ Very low ♦ Low ♦ Med ♦ High ♦ Very high

10. How is this modelling different?
• Multivariate, not focussed upon a univariate
response
• Algorithm defined both states & thresholds
• Used these to define transitions for predictions
• Wanted to model perceived decline
• Designed to include transitional states
• Dataset did not allow equilibrium to be identified
∴ removed assumption of stability from model

11. Management implications of
ecosystem state approach
• Simplifies definition of ecosystem condition
per se
• Allows management at ecosystem scale
• not just indicator species or single parameters
• not based on preconceptions
• More flexible to manage for states than a
focus on a single species or parameter

12. Scenario analyses

13. Approach
MDB SY
Hydrodynami
c model
Ecosystem
state model
Water benefits
Evaluation of
Ramsar obligations
scenarios
Other measures of impact

14. Scenario modelling
• Climates from CSIRO Sustainable Yields for
MDB project, benchmark = 2030
• Flows from MDBA via Ian Webster’s
hydrodynamic model
• 20 scenarios done, e.g.:
• Baseline = historic climate + average in/outflows
• Median Future = moderate climate
• Dry Future = extreme climate
• Extreme + 40cm SLR = extreme climate + SLR
• Modelled 12 sites X 114 years = 1368 site-yr
CLLAMMecology

15. Changing ecosystem states
CLLAMMecology

16. How to read distribution of
ecosystem states (e.g. Baseline)
Sites on the y-axis
Each site-year is
colour coded
Key to colours
Years on the x-axis
Training
Blues & yellows data period
are “healthy”
while greens &
purples/reds are
“unhealthy” CLLAMMecology

17. What does an ugly future look like?
e.g. Dry Future
CLLAMMecology

18. Effect of climate & sea-level rise
Scenario
name
Key to colours used
= state name
How to
read
this
graph
Blues & yellows are
“healthy” while
greens & reds are
Percent of total “unhealthy”
site-years in each
CLLAMMecology
state

19. Effect of climate & extraction
levels

20. Effect of The Living Murray
initiative

21. Ordination of 20 scenarios
2D Stress: 0.04 Effect of
Median Future, +40 cm SLR
Alternative model
Median Future, +20 cm SLR
Sea-level rise
Dry Future, +40 cm SLR The Living Murray
Dry Future, +20 cm SLR Extraction level
Historic Natural
Climate change
Median Natural
Baseline
Dry Natural
Historic TLM off Dry TLM off
Dry Future
Median TLM off
Historic TLM on Baseline Dry TLM on
Median Future
Dry Future, -10 cm SLR
Median TLM on Median Future, -10 cm SLR
MM Dredging
Max USED Flows
Alt baseline

22. Summary of trends across scenarios
• Baseline = a mix of healthy & degraded states
• Future climates invoke more degraded states
e.g. Baseline = 6%, Median Future = 11%, Dry Future = 46%
• Sea-level rise alters mix but still ~45%
• Degraded states involve little flow, higher salinities &
a much reduced range of biota
• ‘a future heading south’ = conditions like the South
Lagoon become more common in both lagoons
CLLAMMecology

23. Key findings
• Climate change has potential to dramatically
affect ecosystem states in the Coorong
• e.g. more degraded states: Baseline = 6%,
Median Future = 11%,Dry Future = 46%
• Extraction of water is exacerbating the
problem
• Even small amounts of environmental water
can alleviate the worst effects of climate
change
• Nothing as good as water over the barrages

24. Key messages
• Climate change has the potential to
devastate the Coorong
• at current extraction levels
• But relatively small amounts of water will
mitigate the worst
• e.g. TLM & other similar initiatives
• Other interventions are less effective
• but may be necessary before flows return

25. Conclusions
• No substitutes for barrage flows
• Climate change does not have to destroy
Coorong ecosystems
• extraction levels play a much bigger role
• Additional fresh water is needed
• River Murray must be the major source

26. HEADLINE TO BE PLACED IN
Thank you
THIS SPACE
Acknowledgements
• CLLAMMecology Research Cluster &
CSIRO CLLAMM researchers (esp. for th
datasets)
• CSIRO Collaboration Fund
• DEH, DWLBC, SA MDB NRM Board
• FR3cE

27. Spare slides for questions

28. HEADLINE TO BE PLACED IN
Recent rainfall
THIS SPACE

29. Ecosystem states
• An explicit link sought between physical &
biological data
• driven by management requirements
• Includes transitional states
• no emphasis on stability
• during decline & (potentially) recovery
• Built as a state-&-transition model
• usually defined by expert opinion
• data-driven definition of each is also possible

30. Defining states & transitions
1. Identify 2. Identify 3. Confirm
clusters differing distinct states
•Group cases conditions •Tests relationship
•Become preliminary •Find environmental b/w biota &
states transitions environment
•Group average •Test model stability •Identifies distinct
cluster analysis •CART analysis states
•ANOSIM analysis
4. Evaluate new 5. Characterise
cases states Biological data
•Classifies test cases •Environmental &
•Tests predictive biological Environmental
capacity characteristics data
•CART & ANOSIM •Indicator species
analyses •SIMPER, BVStep,
other analyses

31. Long v Short time frames
Long term Short term
• Annual time step • Quarterly time step
• 10 sites, 1999 – 2007 • 12 sites, 2005 – 2007
• 3 distinct states (86%) • 6 distinct states (66%)
• Long-term analyses did not pick up recent
deteriorations in condition
• Long- & short- term models could be
combined (61%)
• One model with 8 distinct states

32. Key biota of marine grouping
• Estuarine/marine • Unhealthy marine
• High abundances of • Good numbers of yellow-
mulloway, flounder, black eyed mullet, hardyheads,
bream some marine/estuarine
• Good numbers of fishing species
birds, small waders & teal • Good numbers of fishing
• Diverse benthic fauna birds, moderate numbers
of shorebirds
• Marine • Diverse juvenile
• High abundances of black invertebrate fauna
bream & salmon
• High numbers of fishing
• Degraded marine
birds & small waders • Too few cases to
• High number of large categorise using SIMPER
polychaete worms

33. Key biota of hypersaline grouping
• Healthy hypersaline • Unhealthy hypersaline
• Diverse waders, many • Some tolerant marine fish
waterfowl, esp teal • High numbers of banded stilt,
• Few fish or invertebrate many shorebirds
data • Few invertebrates
• Average hypersaline • Degraded hypersaline
• Few fish present • Many hardyheads
• High numbers of waterfowl, • Some waterfowl, few other
some fishers birds
• Many chironomids, few • Almost no invertebrates
other inverts

34. Community composition – birds,
fish and macrophytes
Standardise Samples by Total
Resemblance: S17 Bray Curtis similarity
01 2D Stress: 0.17
Murray Mouth
01 North Lagoon
04
South Lagoon
06
06 01
02 02 00 00 07
01 00 00 00
04
02 02 00 01 07 04 07
03 03 0205 03 07
020502 07
06
04 00 05 04 0000 01 06
04 06 07
04 0006 03 02 05
07 01 03 03 07 04 04 05
02 0603 05 06 07 05
01 05 04 03 06
02 03 06
04 00 05
07 02 02 01 04
05 01 00 01 01
05
07
06
06
07

35. Hypersaline basin vectors
Improved in both Improved water levels but more days
without flow
Deviation in water level from
+
0
Baseline
-
Fewer days without flow but lower
Deterioration in both
levels
- 0 +
Deviation in days without barrage flow from Baseline
a) Historic Natural & Median Natural, b) Median Future, c) Dry Future & Median Future, +20cm
SLR, d) Dry Natural, e) Median Future, -10cm SLR, f) Median Future, +40cm SLR, g) Dry Future,
-10cm SLR, h) Dry Future, +20cm SLR, i) Dry Future, +40cm SLR, j) Historic TLM off, k) Historic
TLM on, l) Median TLM off, m) Median TLM on, n) Dry TLM off, o) Dry TLM on.

36. More key messages
• In the absence of barrage flows
• dredging is essential
• SLSRS would have a big short-term impact
• channel works + pumping = best option for South
Lagoon states
• additional South East water would have a
longer lasting impact
• But none are as effective as barrage flows