DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY 22 October 2014 ECOSTAT meeting in London
WFD phytoplankton indicators of ecological status
Jacob Carstensen ECOSTAT, London, 22 October 2014 With contribution from SCOR WG137 and support from by:
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Outline for this presentation › Indicators for phytoplankton composition
› Some theory › Recent results from SCOR WG137 and other studies › Linking phytoplankton composition to nutrient pressure › Conclusion on suitability of indicators for phytoplankton
composition
› Indicators for phytoplankton blooms › Some theory › Linking phytoplankton composition to nutrient pressure › Conclusions of suitability of bloom indicators
› Monitoring requirements for chla indicators › Responses to comments from the review panel
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
The coastal zone is a dynamic and productive ecosystem with strong gradients
Carstensen, Cloern & Paerl (in prep.)
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Dominant life forms Diatoms, e.g. Skeletonema costatum Dinoflagellates, e.g. Peridiniella catenata
Cyanobacteria, e.g. Planktothrix agardhii Chlorophytes, e.g. Chlamydomonas spp.
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014 SCOR137: An
unprecedented large data set
85 sites and 30040 samples in total
Area # of stations
Period # of sample
s Chesapeake Bay 21 1984-2009 7198 Danish waters 18 1979-2013 7317 Dutch waters 14 1990-2011 6300 Finnish waters 11 1966-2010 4488 German Wadden Sea 1 1999-2012 1002 Gulf of Riga 4 1976-2012 899 Neuse River Estuary 4 2000-2013 615 Patos Lagoon Estuary 1 1993-2011 225 San Francisco Bay 4 1992-2013 656 Swedish waters 7 1983-2012 1340
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Diversity of coastal ecosystems - physics
0 10 20 30 40 50 60 70 80 900
10
20
30
40
Water depth (m)
Frequency
n=85A)
0 0.5 1 1.5 2 2.50
20
40
60
Tidal amplitude (m)
Freq
uency
n=85B)
0 5 10 15 20 25 30 350
5
10
15
20
Salinity
Freq
uency
n=85C)
4 6 8 10 12 14 16 18 20 220
10
20
30
Temperature (°C)
Freq
uency
n=85D)
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
…. and chemistry
0 1 2 3 4 5 6 7 8 90
10
20
30
Secchi depth (m)
Freq
uency
n=73E)
0 2 4 6 8 10 12 14 160
10
20
30
Chlorophyll a (µg L-‐1)
Freq
uency
n=82F)
0 20 40 60 80 1000
5
10
15
20
25
Total Nitrogen (µmol L-‐1)
Freq
uency
n=79G)
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 50
10
20
30
Total Phosphorus (µmol L-‐1)
Freq
uency
n=79H)
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Factors regulating the proportion of chlorophytes
-‐8-‐7-‐6-‐5-‐4-‐3-‐2-‐10
0 5 10 15 20 25 30 35
Logit(prop
. of chlorop
hytes)
Salinity
Raw
Adjusted
5%
1%
0.1%
-‐8-‐7-‐6-‐5-‐4-‐3-‐2-‐10
-‐2 -‐1 0 1 2 3
Logit(prop
. of chlorop
hytes)
Log(Total phosphorus) (µmol L-‐1)
Raw
Adjusted
5%
1%
0.1%
Carstensen, Cloern & Paerl (in prep.)
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Factors regulating the proportion of cyanobacteria
-‐10
-‐8
-‐6
-‐4
-‐2
0
0 5 10 15 20 25 30 35
Logit(prop
. of cyano
bact.)
Salinity
Raw
Adjusted5%
1%
0.1%
-‐10
-‐8
-‐6
-‐4
-‐2
0
0 20 40 60 80 100
Logit(prop
. of cyano
bact.)
TN:TP molar ratio
Raw
Adjusted5%
1%
0.1%
Carstensen, Cloern & Paerl (in prep.)
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Factors regulating the proportion of diatoms
-‐3
-‐2
-‐1
0
1
2
3
4
2.5 3 3.5 4 4.5 5 5.5 6
Logit(prop
. of d
iatoms)
Log(Total Nitrogen) (µmol L-‐1)
Raw
Adjusted90%
75%
50%
25%
10%
-‐3
-‐2
-‐1
0
1
2
3
4
5 10 15 20
Logit(prop
. of d
iatoms)
Temperature (°C)
Raw
Adjusted90%
75%
50%
25%
10%
Carstensen, Cloern & Paerl (in prep.)
-‐3
-‐2
-‐1
0
1
2
3
4
0 5 10 15 20 25 30 35
Logit(prop
. of d
iatoms)
Salinity
Raw
Adjusted90%
75%
50%
25%
10%
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Factors regulating the proportion of dinoflagellates
-‐6
-‐5
-‐4
-‐3
-‐2
-‐1
0
1
0 5 10 15 20 25 30 35
Logit(prop
. dino
flagellatres)
Salinity
Raw
Adjusted
25%
10%5%
1%
-‐6
-‐5
-‐4
-‐3
-‐2
-‐1
0
1
0.5 1 1.5 2 2.5 3 3.5
Logit(prop
. dino
flagellatres)
Stratification pattern
Raw
Adjusted25%
10%5%
1%
Mixed Intermittent StratifiedCarstensen, Cloern & Paerl (in prep.)
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Considering interannual variations - spring
Jurgensone et al. (2011) Estuaries and Coasts
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Considering interannual variations - summer
Jurgensone et al. (2011) Estuaries and Coasts
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
What is controlling the phytoplankton composition? › A general biomass increase is the most
prominent response to nutrient enrichment 1. Phenology and physical factors (salinity,
temperature and stratification) are most important
2. Changing nutrient ratios can lead to shifts between diatoms, cyanobacteria and other species
3. Diatoms may proliferate increasingly with nutrient enrichment
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Is an increase in the proportion of diatoms necessarily bad quality?
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Can we develop operational indicators? › 76 taxa investigated › ~50% responded to
changing nutrient level › 4 taxa changed their
probability of presence by >30%
› 1 taxon allowed for distinguishing a 20% change in nutrients (N=50) Carstensen & Heiskanen (2007)
Marine Ecology Progress Series
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Conclusion on indicators of phytoplankton community › The structure of the phytoplankton community
is extremely variable and mostly governed by physical processes rather than changing nutrient concentrations
› There is evidence that the phytoplankton community changes with nutrient concentrations, but these changes occur with large orders of magnitude
› Operational indicators that can discriminate between realistic pressure levels are not available at present …… and may not be possible at all!
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Phytoplankton bloom indicators
Carstensen, Henriksen & Heiskanen (2007) Limnology & Oceanography
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
There are many sites in the Baltic Sea where bloom frequency can be calculated
0
200
400
600
800
1000
1200
1400
0 30 60 90 120 150 180 210 240 270 300 330 360
Carbon
biomass (mg m
-‐3)
Julian day
Blooms
Non-‐blooms
B Baltic SeaTvärminne0
50
100
150
200
250
0 30 60 90 120 150 180 210 240 270 300 330 360
Carbon
biomass (mg m
-‐3)
Julian day
Blooms
Non-‐bloomsAGulf of BothniaQuark
Carstensen, Cloern & Klais (Subm.) Estuarine Coastal Shelf Science
There is no specific requirement on the sampling frequency, but monthly sampling would probably be a good idea.
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
0%
10%
20%
30%
40%
0.1 1 10
Bloo
m freq
uency
Total Phosphorus (µM)
BS CB DSE GBNRE RMS SFB WS
H p(linear)=0.0094p(spline)=1.0000
0%
10%
20%
30%
40%
0.01 0.1 1 10
Bloo
m freq
uency
Tidal amplitude (m)
BS CB DSE GBNRE RMS SFB WS
F p(linear)=0.0229p(spline)=1.0000
Factors governing the bloom frequency
0%
10%
20%
30%
40%
Bloo
m freq
uency
Mixed Intermittent Stratified
BSCBDSEGBNRERMSSFBWS
D F2,83=2.48; p=0.0897
0%
10%
20%
30%
40%
10 100
Bloo
m freq
uency
Total Nitrogen (µM)
BS CB DSE GBNRE RMS SFB WS
G p(linear)=0.0163p(spline)=1.0000
Carstensen, Cloern & Klais (Subm.) Estuarine Coastal Shelf Science
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Conclusion on indicators of phytoplankton blooms › The likelihood of phytoplankton blooms
increases with nutrient enrichment, but the triggering mechanism is physical – this adds to variability
› Bloom indicators can be calculated for many monitoring sites in the Baltic Sea, especially because time series are already long in many places
› Operational bloom indicators may potentially be developed, but the sensitivity of these in a realistic nutrient management scenario may be too coarse!
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Uncertainty aspects of phytoplankton indicators
Chla G-M = 2.0 µg/l Chla True = 1.6 µg/l
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Responses to review panel › Q1: Frequency of monitoring data for bloom
indicator (Carstensen et al. 2007) › The indicator can be calculated for many different
monitoring sites around the Baltic and it is generally applicable
› Q1: Bloom indicators from EUTRO-OPER based on Alg@line data › Data cover the open Baltic Sea and indicators are
developed for the open Baltic Sea › Q1: Bloom indicators based on remote
sensing › Remote sensing data in the coastal zone is
sensitive to shallowness, resuspended particles, CDOM etc.
› Indicators developed apply to the open Baltic Sea
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Responses to review panel › Q3: Causal linkages between eutrophication
and phytoplankton blooms/composition › Yes, these causal linkages exist, but they are not
operational on a realistic eutrophication range › Q3: Use of composite indicators
› Only useful when there are clear synergies in the pressure-response relationships
› Transparency of indicators is important, particularly for the intercalibration
› The suggested indicators have not been documented to be operational on a realistic eutrophication range
› Q3: Connection to HELCOM › There are links between CORESETII and GIG
experts
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Responses to review panel › Q3: Use of the large phytoplankton data set
from the Baltic Sea › These data have been analyses in many research
publications, but there is a gap between research ideas and operational tools needed for WFD implementation
› The GIG phytoplankton group has been ambitious › AQ1: Carvalho et al. (2013) recommended at
least 3 samples per year in at least 3 years › This is too simple! As shown, this depends on the
required sensitivity of the indicator as well as confidence and power levels.
DEPARTMENT OF BIOSCIENCE AARHUS UNIVERSITY ECOSTAT meeting in London 22 October 2014
Responses to review panel › AQ2: Believe that a multimetric indicator will
reduce uncertainty › Both yes and no. However, this has never been
documented – just postulated. › AQ3: Use of novel technologies
› Agree. Novel techniques do offer new possibilities, but the techniques need to mature and their reliability needs to be documented before operational indicators can be developed on such data
› There are no large data sets yet on many of the novel techniques that could potentially be used, e.g. pigment analyses, FlowCAMs, etc.
”For an operational WFD implementation, it is important that we use well-proven and sensitive phytoplankton indicators”