Transcript Slide 1

Climate Variability and
Paralytic Shellfish Toxins in
Puget Sound Shellfish
Stephanie K. Moore
School of Oceanography
University of Washington
Box 355351, Seattle, WA 98195
Nathan J. Mantua, Vera L. Trainer and Barbara M. Hickey
Alexandrium catenella
Photographs: Anthony Odell
Paralytic Shellfish Toxins (PSTs)
• Suite of toxins produced by species of
Alexandrium
• Saxitoxin (potent neurotoxin)
• Can accumulate in filter-feeding shellfish
– Paralytic Shellfish Poisoning
• Death in 2 hours
Butter Clam
Closures only in
Northwest basin
and Strait
First Central and
Whidbey basin
closures
Closures in
most of Sound
First South
basin closures
Trainer et al. 2003
Project Goals
• To better understand the role of climate
forcing on Harmful Algal Blooms in Puget
Sound…
Climate variability and Puget
Sound PSTs
Puget Sound PSTs
Puget Sound
Oceanography
“Weather”
ENSO
PDO
Global Warming
• Days to Weeks
• Years
• Decades
• Centuries?
• < 2 Weeks
• 1 Year
• 2 Years
• Decades?
Project Goals
• To better understand the role of climate forcing
on Harmful Algal Blooms in Puget Sound
– Describe space-time patterns of seasonal,
interannual, and interdecadal variability in Puget
Sound oceanographic properties
– Evaluate the nature of local, regional, and large-scale
forcings on Puget Sound
– Identify environmental conditions conducive to HAB
development
– Determine the capacity for prediction of HAB risks
Oceanographic data
• WA DoE Marine Waters Monitoring Program monthly
temperature and salinity profiles from 16 stations for
1993-2002
• Biannual (summer/winter) UW PRISM transects from 40
stations from 1998-2003
• DFO monthly average Race Rocks SST and SSS from
1937-2002
• USGS stream gage data, 1950-present
• PFEL monthly mean upwelling index at 48N, 1950-2002
• US Climate Division air temperature and precipitation,
1950-2002
• Large scale climate indices: NPI, PDO, Nino3.4, 19502002
Climatological mean temperature profiles
Surface intensified warming in Whidbey,
Hood Canal and South basins
Climatological mean salinity profiles
Winter freshening
Winter/Spring freshening
Summer freshening
Monthly mean
river discharge
1993-2002
• Total runoff into Puget
Sound peaks
biannually in Winter
and May-June
• Fraser River has
minimum discharge in
Winter, peaks in
May-June-July
Climatological mean density profiles
Year-round stratification
Summer stratification
PRISM cruise temperature transects:
6 yr means for 1998-2003
PRISM cruise salinity transects:
6 yr means for 1998-2003
Interannual variability
• EOF/PCA analysis of
DoE MWM profiles
• Normalized
anomalies of
temperature, salinity,
density, and the
depth and strength of
stratification
• Leading EOF/PC
captures a substantial
fraction of the total
variance in profiles for all
16 stations
• Temperature 42%
• Salinity 58%
• Density 56%
• Coherence not as
strong for stratification
• Maximum Buoyancy
Frequency 22%
• PC’s highlight strong
interannual variability
during the 1993-2002
period of record
Profiles during
extreme periods
Vertical profiles of
temperature, salinity
and density at station
CMB003 for months
when PC1T, PC1S
and PC1D,
respectively, showed
extreme positive (top)
and negative (bottom)
anomalies
Loading patterns
• The loading vectors for
the leading PC of
temperature, salinity
(shown right) and
density generally show
strong positive values
from surface to depth
• Therefore, the leading
PC’s represent depth
and area averaged
variations in
oceanographic
properties at all 16
station in Puget Sound
We use Race Rocks SST and SSS as
a proxy for PC1T, PC1S, and PC1D to
increase our sample sizes
Correlations with
PCs for 1993-2002
JFM
AMJ
JAS
OND
Race Rocks SST
with PC1T
.868***
.748***
.456
.748***
Race Rocks SSS
with PC1D
.733
.832
.804**
.817***
Race Rocks SSS
with PC1S
.733
.823*
.886***
.827***
Statistical significance at 90%, 95%, and 99% confidence intervals are
indicated by *, **, and ***, respectively.
Correlation analysis
Race Rocks SST and SSS with local and
large scale indices, 1951-2002; degrees
of freedom are noted with subscripts
• Puget Sound streamflow is
significantly correlated with
SSS during winter, spring and
summer
• SST is strongly correlated with
regional air temperature during
all seasons, but correlations
with large-scale climate indices
are also significant during
winter
Lag correlations: large scale indices
and Race Rocks SST and SSS
1950-2002
Longer term context 1937 to 2002
Annual mean SST and SSS phase diagram for Race
Rocks lighthouse
The 1993-2002 MWM data we’ve examined comes
from a period that was substantially warmer than prior
decades
Race Rocks annual mean SST
1921-2005
Figure courtesy of Dr. Phil Mote, UW Climate Impacts Group
Summary: Part I
• Analysis of monthly T and S profiles reveals basin-wide coherence
at seasonal to interannual timescales
• We use longer term time series of SST and SSS from Race Rocks
to evaluate correlations between Puget Sound properties and local
to large-scale forcing parameters
– Local air temperature is highly correlated with SST, while freshwater
inflow is highly correlated with SSS
– Large-scale influences (indicated by Aleutian Low, ENSO and PDO
indices) are most prominent in winter, but winter SST effects persist for
the following 2-3 seasons, and this indicates relatively good prospects
for predicting Puget Sound temperature variations one to a few seasons
into the future
• There are strong warming trends in Race Rocks SST in the period
since 1970, so the recent period of MWM and PRISM sampling is
unique at least for the past ~85 years
Alexandrium catenella
Photographs: Anthony Odell
Shellfish toxicity data
ADVANTAGES
• Toxin levels in shellfish provide
proxy for bloom dynamics
• Integrates fine-scale spatial
and temporal variability in
water column toxicity
WDoH Biotoxin Program
Other shellfish 20%
LIMITATIONS
• Sampling concentrated in
summer and during high PST
episodes
• Few long term records at same
site
• Various shellfish species
– e.g. Pacific Oysters, Butter
Clams, Blue Mussels
Littleneck clams 16%
Blue mussels 31%
Butter clams 18%
Pacific oysters 17%
Must match toxin uptake and
depuration rates of shellfish
with water column dynamics!
Shellfish toxicity and bloom dynamics
decoupled when:
1. Toxin retention is high and detoxification rates are long
e.g. Sequim Bay butter clams
2. Shellfish species sensitive to PSTs
e.g. Filucy Bay pacific oysters
Selection of blue
mussel PST data for
trend analysis
• Because the more you
look the more you see…
• Only sites with regular,
long term observations
– Must have one or more
observations per month for
at least 70% months
– Leaves 23 sites over the
time period 1990 to 2002
Puget Sound
The more you look the more you see…
Climatology of toxicity 1990 to 2002
Spatial and
temporal
variability
in blue
mussel
PSTs
NB: 20,000 µg
STXeq 100 g-1
November 2002!
PCA blue mussel PSTs
“Hot spot” indicator sites
Annual indices of bloom behavior
No significant trends with time
No. of toxic sites 1990 to 2002
Closures only in
Northwest basin
and Strait
First Central and
Whidbey basin
closures
Closures in
most of Sound
First South
basin closures
Trainer et al. 2003
Correlations of annual indices with
environmental properties
Env.
properties
HOT SPOT SITES (1990 to 2002)
Day of First
No. Days
Day of Max
-.0111
-.4911
-.396
-.286
-.036
.5311
-.3611
-.3811
.0211
.2111
.0211
.623
PRCP
-.3911
.1911
-.0111
-.4711
-.4311
-.4011
-.243
STRM
-.1711
.3911
.1611
.0211
.0211
-.593
TIDE
<.011
-.021
.141
-.051
-.141
-.041
-.231
UPWL
.014
.324
.474
.114
-.024
.134
-.173
WIND
-.034
.204
.254
.034
.024
.084
-.463
NPI
-.3511
.0811
.2111
-.2211
-.3111
-.1811
-.303
ENSO
.3011
-.1311
-.2111
-.2911
-.2011
-.2111
.483
PDO
.036
.146
.036
-.236
.016
.066
.403
SST
-.1811
SSS
-.056
AIR
First to Last
.0811
.3211
Log10
(Max+1)
Log10
(Mean+1)
Log10
(Cum+1)
-.0511
-.1111
-.093
-.406
-.436
<.013
No significant trends with environment at p ≤ 0.05
Finer temporal resolution…
6 biggest events identified by “hot spots”
Hindcasting of mean toxicity for all
of Puget Sound 1990 to 2002
Timescales of El Niño effects
ENSO teleconnections to
Puget Sound are most
apparent from Nov-Apr…
Exceptional PST events tend
to occur in the fall
Capacity for prediction
• Low streamflow, weak surface winds and
low tidal variability precede toxic events
– Typically occurs in early fall following the peak
of seasonal warming of surface waters
• Forecasting PSP risk is limited to
days/weeks
• No evidence for increased risk during El
Niño years
Summary: Part II
• Shellfish species selection for trend analysis of
PSTs is important
• Toxicity in Puget Sound is highly variable in
space and time
• Annual indices of bloom behavior from 1990 to
2002 show
– No significant trends with time
– No relationships with environmental properties
• A combination of high frequency weather events
determine bloom occurrence
Acknowledgments
• Mitsuhiro Kawase, Jan Newton, Jonathan
Kellogg and Mark Warner (UW)
• Skip Albertson (DoE)
• Sheryl Day (NOAA NWFSC)
• Funding provided by NOAA's West Coast
Center for Oceans and Human Health
Meaningful relationship with
environment?
Mean annual toxicity versus PC1T, PC1N and PC1S
Basin-wide annual mean temperature
and maximum buoyancy frequency
Basin-wide annual mean
temperature and salinity
PRISM cruise density transects:
6 yr means for 1998-2003
Climate variability
Pacific Decadal
Oscillation
20-30 years
El Niño/Southern
Oscillation
6-18 months
Warm phase PDO
Warm phase ENSO
(El Niño)
Source: Climate Impacts Group, University of Washington
What does this mean for the PNW?
El Niño and warm phase PDO generally
affect:
– Weather patterns
• Warmer and drier winters
• ↓ Precipitation/runoff/stream flow
– Structure of ocean
• ↓ Upwelling
• Warmer surface waters in ocean
• Less nutrients
Decadal variability of butter clam toxicity
Butter clam toxicity and no. warm
SST days 1957 to 1991