Transcript Delayed, chronic, and indirect effects of shoreline oiling/treatment

Exxon Valdez oil spill [EVOS] legacy:
Synthesis of long-term ecosystem responses
Riki Ott, Charles H. “Pete” Peterson
& Stanley “Jeep” Rice
Theme: Chronic effects of decade-long
contamination of key shoreline habitats
and indirect interactions are important
1994 EVOS oil, Prince William Sound, AK
Delayed, chronic, and indirect effects of
shoreline oiling/treatment
• Treat EVOS as an ecosystem
perturbation
• Capitalize on vast research
effort
• Synthesis focused on
shoreline habitats
• Contrast of NRDA based on
old ecotoxicity risk models
vs. field-based sampling
Assumptions about oil toxicology in 1989
•Alaska's Water Quality Standard for PAHs was 10 ppb and
provided conservative protection of natural resources
•Oil toxicity declined very rapidly in a matter of
days/weeks
•Acute toxicity tests of lab animals adequately predicted
risk
•Contact with feathers and fur was the only significant
route of injury to birds and mammals
•Oil spill impacts could be assessed on a species-by-species
basis with no regard for dependencies within the ecosystem
Contrasting terms
• Acute vs. Chronic exposure
(= short- vs. long-term)
• Lethal vs. Sublethal impacts
(= mortality vs. growth, reproduction, body condition)
• Immediate vs. Delayed response
(= rapid vs. postponed)
• Direct vs. Indirect effect
(= A  B vs. A  C  B)
• Trophic cascade vs. Biogenic habitat loss
(=
change in predator affects its prey, which affects its prey,
etc. vs. change in an organism that provides structural
living space for other organisms)
Benefits of EVOS field sampling approach
• Employs statistical sampling design
• Integrates responses across all
mechanisms
• Includes chronic effects on long
time scales
• Includes interactions of oil and
other stressors
• Includes indirect interactions from
trophic cascades, habitat
modifications, etc.
Synthesis of longterm ecosystem
responses
• Shows old assumptions of oil ecotoxicity to be
inadequate
• Weathered oil persists
• Weathered oil remains bioavailable in
important environments
• Weathered oil (multi-ring PAHs) induces
toxicity during chronic exposure
Water column PAH contamination
after EVOS
• Low (1-8 ppb) even
during weeks 1-5 in
1989
• Essentially below
detection by end of
summer 1989 using
traditional water sampling
• But filter-feeding mussels revealed exposure
and bioavailability even into summer 1992
via filtration of contaminated particulates
Persistence of oil
• Asphalts high on shore - biologically inactive
• Biologically available pockets in protected sites for >10
yrs
– Under armor of mussel beds with relatively
unweathered oil contaminating mussels at least into
summer 1994 at study’s end
– In groundwater of deltas of anadromous streams yrs
later
– In sediments among boulders on oiled beaches for yrs
• Transported on particles to shallow subtidal where elevated
PAHs persisted until 1995 at study’s end
Is the oil still
there?
Prince William
Sound
Alaska
.
en
iP
na
e
K
2001 Survey Results:
P
91 sites with
9,000 total pits
- 53 sites with oil
- 38 sites without oil
ka
as
l
A
la
su
n
i
en
Ko
k Is
dia
Gulf of
Alaska
100 Miles
N
Distribution of oil 12 years later
Upper
intertidal
Tidal
zone (m)
Surface oil
(# of pits)
Subsurface oil
(# of pits)
+ 4.8
+ 4.3
+ 3.3
+ 2.8
+ 2.3
+ 1.8
37
56
58
60
40
29
5
28
69
91
123
117
Biological
zone
(lower intertidal) < 1m
Oil below sampling grid = Yes
How far down = ?
Shoreline treatments
• Intense in 1989, 1990
summers with some
extending into 1991
• Invasive including wiping
surfaces, pressurized washes
with hot and cold water,
bioremediation, rock
washing, tilling, and berm
relocation
• Had major impacts on
shoreline habitat, plants and
animals
Direct effects on rocky intertidal species
• Fucus removal - high and mid
shore
• Limpet Tectura persona decline
high on shore
• Balanoid barnacle decline
• Blue mussel decline
• Periwinkle Littorina sitkana
decline
• Drilling predator Nucella
lamellosa decline
• Main cause - pressurized washing
Indirect effects on rocky intertidal
community
unoiled
• Modest bloom of ephemeral algae in
absence of Fucus competition and with
low grazing
0
• Absent nearby canopy, Fucus spore
arrival is limited and recruits desiccate
100 oiled & not cleaned
high on shore
• Opportunistic barnacle colonizes
heavily
0
• Fucus colonizing barnacle tests is
uprooted
100 oiled & cleaned
• Fucus expansion into low shore
inhibits red algae
100
0
1990 1991 1992
Possible indirect
effects on rocky
intertidal
• Potential induction of unstable cycle in
Fucus cover as single-aged colonist plants
senesce in synchrony 5 yrs later
• Reduction of biogenic habitat normally
provided by Fucus and blue mussels
impacts gastropods and smaller
invertebrates
Direct effects on shallow
sedimentary communities
•
•
•
•
Enhanced hydrocarbon-degrading bacteria
Reduced shoot densities of eelgrass until 1993
Reduced clam density (littlenecks, butters)
Sediment toxicity to amphipods and reduced
densities of some amphipod taxa
• Reduction of abundance of seastar
Dermasterias and helmet crab Telmessus
Indirect effects in shallow
sedimentary communities
• Increases in oligochaetes and surface depositfeeders (polychaetes) via enhanced microbial
food and reduced invert predation
• Continued depression in clam recruitment to
1998 via absence of fine sediments moved
down slope during pressurized washing
• Probable enhanced growth of demersal fishes
• Probable reduction in phytal eelgrass species
Direct effects in shallow subtidal
rocky community
• Size distributions of
dominant kelps imply
mortality of adults in 1989
followed by rapid
recolonization of juveniles
in 3 systems
• Reduction in abundance of
2 seastars Dermasterias
and Evasterias and helmet
crab Telmessus,
recovering by 1993
Avg. density in shallow bays
8
Telmessus cheiragonus
4
0
18
Dermasterias imbricata
9
0
1990
1991
1992
Indirect effects in
shallow subtidal rocky
community
• Enhanced abundances and sizes of green sea
urchins on northern Knight where sea otters
had not recovered by 1998, possibly inducing
future trophic cascades
• Possible release of predation on prey of
seastars and helmet crabs in kelp and
seagrass communities
Resident nearshore fishes
• Direct acute effects
– reduced abundance and biomass of intertidal fishes in
1990 with recovery well advanced by 1991
– parasite burden enhanced in some demersal fishes
• Direct effects of chronic exposure
– hemosiderosis in intertidal fishes 1990-93
– P450 detoxification system induced in masked
greenling - 1996
• Indirect effects
– in shallow subtidal, juvenile cod and Arctic shanny
increased by >100% at oiled sites perhaps in response
to enhanced mytilid prey
Intertidal and shallow-spawning fishes
- direct acute effects
• Premature hatch, larval
abnormalities and elevated
mortality in herring
• Egg mortality enhanced in pink
salmon
• Possible egg mortality in sand
lance and capelin, important
forage fishes that deposit eggs in
shallow sediments
Intertidal and shallow-spawning fishes
- effects of chronic exposure
• Pink salmon egg mortality
enhanced for at least 4 yrs due to
exposure to weathered oil in
streams.
• Possible genetic damage to pink
salmon affecting their rate of
straying and reproductive output
• Possible contribution of oiling
stress to disease induction that
led to 1993 herring crash, which
has persisted for at least 8 more
yrs
Summer foraging fishes in nearshore direct acute & chronic effects
• Pink salmon - reduced growth in 1989 with smaller
effect in 1990, likely lowering survival at sea
• Dolly Varden and cutthroat trout - reduced growth and
possibly survival 1-3 yrs after EVOS
• Chum (and other?) salmon likely experience analogous
effects of detoxification costs
• Induction of P450 detoxification system 1-3 yrs after
EVOS in pink salmon, Dolly Varden, soles, and others
Shorebirds and seaducks consumers of intertidal inverts
• Direct acute effects of EVOS
– large numbers of oiled dead birds
collected post spill
– shoreline surveys show detectable
declines in some taxa at oiled sites
compared with unoiled control sites
using a pre-spill survey in 1984-85
to control for spatial variability
Shorebirds - consumers of benthic invertebrates
• Chronic exposure effects
# of eggs laid in
2nd nesting after
loss of 1st nest
2
0
oiled
unoiled
– black oystercatcher - fed chicks more oiled
prey to achieve growth, fledged later, laid
fewer eggs in second brood upon
disturbance
– nesting of black oystercatchers stable on
unoiled Montague, while disrupted on
oiled Green Island, recovering by 1993
– black oystercatcher - counts declined
further from 1989-90 before recovering in
1991-93
Seaducks - consumers of benthic inverts
• Indirect and chronic exposure effects
– harlequin duck - lower adult female over-winter survival evident
in 1995-96, 1996-97, 1997-98
– harlequin duck - winter counts still declining in western (oiled)
compared to eastern PWS through 1997-98
– Barrow’s goldeneye - growing divergence in abundance between
oiled and unoiled survey data through 1998
– harlequin duck in 1998 and Barrow’s goldeneye in 1996-97 elevated P4501A
Avian consumers
of forage fish
• Direct acute effects
– oiled dead birds of many species collected on shore
– statistically detectable declines in 1989
• Indirect and chronic effects
– cormorants, mergansers, black-legged kittiwake, murres,
pigeon guillemot, and loons show depressed counts through at
least 1998 (1993 for loons)
– pigeon guillemot shows lower nest productivity in 1993-98
with reduced availability of high-lipid forage fish like herring
Avian scavengers and omnivores in
shoreline habitats
• Direct acute effects
– reduction of 10% in PWS bald eagles in 1989
– reduction of northwestern crows in 1989
– PWS mortality of eagle eggs exposed to oil in 1989
• Indirect and chronic effects
– northwestern crows more concentrated on oiled Kenai
shores in 1989
– over yrs PWS counts of bald eagles and glaucouswinged gulls are enhanced
c Bakhtin
Marine mammals of
shoreline habitats
• Direct acute effects of EVOS
– sea otter - over 1,000 oiled corpses recovered,
representing an estimated 2,800-5,000 mortalities
from encounter with oil
– harbor seal - direct mortality estimated at about
300 individuals in PWS alone
– Steller sea lion - likely mortalities from same
mechanism (inhalation of toxic fumes) that
affected seals
Marine mammals of
shoreline habitats
c Bennett
• Indirect and chronic exposure effects
– sea otter juveniles over-winter mortality higher in oiled
areas in 1990-91 and 1992-93
– sea otter carcass counts show high prime age mortality
for at least 5 yrs following EVOS
– sea otter counts not yet recovered in PWS, especially
northern Knight Island
– P450 detoxification system induced in sea otters in
1996-98
Marine mammals
of shoreline
habitats
• Indirect and chronic exposure effects (cont’d)
–harbor seal molt counts have not initiated convergence in
PWS through at least 1997
–river otters had larger home ranges in oiled areas, greater
abandonment of latrine sites 1990
–reduction of blood hemoglobin in river otters, diminishing
oxygen delivery, diving ability
–elevated haptoglobin counts in river otters through 1991
Indirect, chronic, delayed effects
• Abundant evidence of their existence from
controlled field sampling over yrs
• Loss of biogenic habitat
– Fucus, mussels, eelgrass
• Tend to move up the food chain
– organic enrichment of the food chain
– sublethal effects of ingesting contaminated
foods with reproductive and population
impacts
– loss of forage fish prey
Indirect, chronic, delayed effects
• Possible top-down effects
– increases in size and abundance of sea
urchins on northern Knight probably reflects
reduced predation by sea otters and may lead
to localized overgrazing of kelps and loss of
fish habitat
– reduction of harbor seals and sea lions may
be causing transient killer whales to switch to
sea otters with dramatic ecosystem
implications
Ecotoxicity requires a context of the web
of interacting species
•No species is independent of others
–habitat
–prey
–predator
•Ecosystem engineers - species that provide important
structural habitat (kelps, seagrasses)
•Keystone species – those with disproportionate controlling
influence on community composition (sea otters, herring)
Changing paradigms of oil impacts to
shoreline communities
• Old dogma - Short generation times of plants and
invertebrates and rapid weathering of oil on shore leads to
rapid recovery
• New recognition after EVOS
– strong interspecific interactions create cascades of
delayed effects over many years - keystone species and
habitat engineers
– preemption of space can inhibit recovery
– sensitive taxa (crustaceans) and oil persistence in
protected sediments slow recovery for yrs
Shifting paradigms in ecotoxicology
• Old 1970’s approach - based on lab
bioassays of acute mortality of individual
species to short-lived water-soluble
fraction of oil (mostly BTEX plus
napthalene)
• New 1990’s realizations – persistent biologically available 3-5 ring
PAHs from oil in protected habitats is toxic
with chronic impacts for yrs
– strong interspecific interactions, including
top-down trophic cascades, biogenic
habitat provision, and competition, induce
indirect and delayed effects for yrs
Limitations of old dogma underlying
ecotoxicity risk assessment
• Based solely on short-term
acute toxicity in lab
• Typically assesses only one
mechanism (eg, exposure to
WSF – water-soluble
fraction)
• Treats species as
independent, not linked
through food web or habitat
responses
Limitations of old dogma underlying
ecotoxicity risk assessment
•Extrapolates from few lab-rat
species to species taxonomically
similar but potentially different in
ecology and physiology
•Includes no effects of chronic
exposure, delayed impacts, or
interactions among species
•Includes no sub-lethal impacts on
growth, development, or
reproduction – all of which can
translate to population
consequences
Problems with field-based empirical
assessment approach
• Expensive to conduct – who should pay?
• Study design begets conclusions – who should design
the study?
• Typically lack before data – but rigorous study designs
are still possible
• Time-consuming to capture indirect, chronic, and
delayed effects – but necessary
• High uncertainty because complex mix of mechanisms
not all elaborated – who should bear the burden of
proof?
Implications for
future oil spills
• Omission of indirect, chronic, delayed effects
in ecotoxicity risk models amounts to a large
understatement of oil spill impacts
• Predictive ability of such indirect effects by
ecological science lies far in the future,
although some strong interactions can be
confidently predicted