Use of biological life history, genetic, and vulnerability information to optimize the repertoire of

1
Use of biological life history, genetic, and
vulnerability information to optimize the
repertoire of oil combating technologies
ICES CM 2005/S12

• S. Kuikka, T. Lecklin, M. Rahikainen, K.
  Vainio-Mattila, T. Juntunen , T. Rosqvist, K.
  Jolma, A. Mäkinen,
• J. Ikävalko, J. Mattila, A. Aalto

OILECO-project, University of Helsinki
http//hykotka.helsinki.fi/oileco
2
Management questions to be answered

• How do we justify the requirements for
  substantial investments to prevent ecosystem
  impacts prior to accidents?
• How do we allocate the available oil combating
  resources between the potential oil spill areas
  to minimize the harm to ecosystem?
• ? ranking of decision options is a minimum
  requirement

3
Aims of OILECO

• To collect and compile information on the
  ecosystem values of the Gulf of Finland.
• To evaluate the sensitivity of the ecosystem
  components on oil spills.
• To produce supportive information for the
  decisions to safeguard the most valuable
  populations in the Gulf of Finland in the case of
  oil spill.
• To evaluate the justifications for the
  investments on preventive measures.

4
Complicated models and the level 3 modelling?
0.2
0.1
0.6
0.2
0.1
0.4
0.2
0.3
0.4
5
Gulf of Finland
6
The Gulf of Finland is highly sensitive to oil
spills - geographical and climatological
characteristics

• The Baltic Sea is a small on a global scale (370
  000 km²) and geologically young (8 000-10 000
  years)
• The Gulf of Finland is the eastern arm of the
  Baltic Sea 460 km long and 15-120 km wide
• Average temperature is 10 ºC - low temperature
  decreases biodegradation and evaporation rate and
  increases viscosity of oil
• GoF is frozen from December to March - ice cover
  decreases photodegradation and volatilisation of
  oil
• Salinity varies from 7 to 1 ppt - in brackish
  water natural dispersion of oil is slower than in
  sea water
• Swallow water and slow exchange decrease the
  dilution rate

7
The Gulf of Finland ecosystem is highly sensitive
to oil spills - special features of the wildlife

• Low biodiversity - a mixture of both marine and
  freshwater species living close to their
  physiological tolerance limits in the GoF
• The number of species is low but the abundance of
  a single species can be high
• Strong and short vertical and horizontal
  gradients occur - local genetic adaptations most
  likely
• A central route for migrating arctic birds,
  important spawning grounds for fish e.g. Baltic
  herring (Clupea harengus membras)
• Particularly Sensitive Sea Area (PSSA), Ramsar
  sites, Natura 2000 areas, Important Bird Areas
  (IBA), National Parks in Finland and in Estonia

8
Components of conceptual / logical model
Justifications for investments BEFORE the oil
spill

Conservation valuation specificity of genome
for Gulf of Finland and ecosystem role
Probability of oil spill
Recovery potential (recolonization and
regeneration)

Sensitivity and vulnerability potential impact
on populations and on genetic variability
Priorisation of oil combating measures AFTER
the oil spill

Conservation valuation specificity of genome
for Gulf of Finland and ecosystem role
Sensitivity and vulnerability potential impact
on populations and on genetic variability
Recovery potential (recolonization and
regeneration)
Mitigation potential (oil booms, dispersants
etc) and protect ability

9
1) Probability of an accident

• Maritime oil transport has increased during the
  last years and is estimated to grow even more
  strongly in future - from 40 million tons in 2000
  to 190 million tons in 2010
• Estimated occurrence of oil spill in the Baltic
  is 0,35 spills per 1000 journeys, that is, 14
  accidents per year including spills of bunker oil
• The most riskful area is between Helsinki and
  Tallinn (the dense passenger traffic)
• Improvements in maritime safety e.g. AIS
  (automatic identification system), GOFREP (ship
  reporting system)

10
Oil transportation in the Gulf of Finland in
years 1987-2003 and estimated development for
2004 and 2010 (Source http//www.vtt.fi/uutta/200
4/itameri/kaaviot.ppt)
11
2) Sensitivity of the species and populations
some observed oil-induced responses on
wildlife (modified from NAS (2003) Oil in the
Sea III Inputs, Fates and Effects)
decreased population size
no response
direct mortality
altered population structure
food supply reduced
indirect mortality
oil spill
foraging behaviour altered
lowered reproductive success
physiological stress
migration
habitat quality reduced
reduced habitat occupancy and use
12
2) Sensitivity of the species and populations

• An organism may be affected as a result of
  smothering and/or toxic effects of oil
• Early developmental stages seems to be especially
  vulnerable to hydrocarbon exposure
• Concentrations significantly lower than acutely
  toxic levels may alter the long-term survival of
  affected populations and thus alter population
  and community dynamic
• Indirect effects e.g. the changes in trophic
  cascades and habitat loss, can be as detrimental
  as direct effects

13
3) Conservation values of populations I -
genetic uniqueness

• Are there organisms with behavioural and
  physiological adaptations genetically unique to
  the GoF that might be permanently lost if a big
  oil accident happens?
• The nesting colonies of guillemot (Uria aalge)
  and razorbill (Alca torda) in Haverörn
• Grey seal (Halichoerus grypus) and Baltic ringed
  seal (Phoca hispida baltica) separated from the
  Atlantic populations 10 000 years ago
• An isolated population of vendace (Coregonus
  albula) in the eastern part of the GoF

Photo http//fulmar.free.fr
Photo http//fulmar.free.fr
Photo http//www.pinnipeds.org/species/species.ht
m
Photo http//www.pinnipeds.org/species/species.ht
m
14
3) Conservation values of populations II - the
ecosystem role of the key species

• The key species species that strongly influence
  community structure e.g. are principal elements
  in the food web or provide distinct habitats
• Low biodiversity Only few key species and
• no other species taking over their functions
• Blue mussel (Mytilus edulis) prey for many
• birds, seals and fish mussel beds provide
• substratum for e.g. barnacles
• Bladder wrack (Fucus vesiculosus)
• small invertebrates and young fish find their
  food
• and shelter among bladder wrack belts in rocky
  shores

Photo http//reports.eea.eu.int/report_2002_0524_
154909/en/page131.html
15
4) Recovery potential of populations

• Recovery potential of populations having high
  conservation value - probability estimates for
  recovery and time needed for recovery
• Reproduction the rate of population increase at
  low population densities is the driving force of
  recovery dynamics
• Recolonization - recovery may be enhanced by
  immigration from the adjacent populations
• Recovery times following spills can vary from a
  few days to more than ten years

16
5) Mitigation measures and protect ability

• Allocation of oil combating resources
• and use of appropriate techniques to
• protect most valuable populations
• Major categories of technologies
• Mechanical methods oil booms, skimmers, in-situ
  burning
• Chemical and biological methods dispersants,
  natural bioremediation
• Physical methods washing, raking, the voluntary
  oil brigades
• Tradeoffs between advantages and disadvantages of
  these oil spill mitigation measures to different
  biotopes are not well known

17
Modelling technique

• Luckily, there is very little data available ?
• ? other sources of information must be utilized
• Probabilistic Bayesian nets (like in Juntunen et
  al., 2005)
• Including the presented logic, but in a
  probabilistic format
• Degree of knowledge differs greatly in different
  parts of the model
• Some probabilistic dependencies can be estimated
  by Monte Carlo simulations, some from pure expert
  knowledge
• Later updating easy due to the Bayesian approach

18
Description of knowledge in a BN
model
19
Example graphical description of a Bayes net
0.2
0.1
0.6
0.2
0.1
0.4
0.2
0.4
0.3
20
The most simple model structure
Species to be fixed first gt probabilistic
outputs for the other variables
21
Future aims

• Use of several experts (model uncertainty
  dominates over parameter uncertainty)
• Development of decision model piece by piece
  OILECO is to produce the ecosystem piece
• Juntunen et al. (2005, same session) gives good
  estimates for the technical side
• Decision orientated modelling will likely help to
  focus the scientific activities (the concept of
  value-of-information)
• Interest to join to an EU project?