OVERVIEW

1
The Implications of Incomplete and Imperfect
Information for Multinational Management of
Transboundary Marine Fisheries A Cooperative
Research Project Robert McKelvey U. Montana
Kathleen Miller National Center for
Atmospheric Research Peter Golubtsov Moscow
State University, Russia
OVERVIEW
SPLIT STREAM MODEL RESULTS
APPLICATION MULTINATIONAL TUNA MANAGEMENT
EEZs in the WCPO
The Dynamic Imperfect Information Split Stream
Harvesting Model

• The Project Setting and Program
• Harvested marine fish stocks have been depleted
  worldwide.
• The Culprit Common Harvest of a breeding fish
  stock by non-cooperating fleets--The Tragedy of
  the Common
• The Remedy Central management by a
  governmental regulatory authority, to restrain
  and allocate harvests among fleets.
• Complications
• Transboundary fish stocks, which migrate
  across zones of national authority, requiring
  multilateral cooperative management.
• Stochastic stock demographics, due to
  oceanic environmental variability
• (e.g. El Nino episodes.)
• Inability to predict environmental regime
  shifts accurately, and in a timely way, so as to
  make current harvesting decisions optimally.
• The program Examine alternative management
  institutions through case studies and
  game-theoretic modeling.

• Tuna Fisheries Worldwide A Comparative Study
• Species of tuna are found is all of the worlds
  oceans, in a wide variety of oceanographic,
  ecological, socioeconomic, and political
  circumstances. The goal here is to understand
  the implications of these differing conditions.
• All tuna stocks are highly migratory, and thus
  their ranges may intersect the EEZs of many
  coastal nations. Furthermore, oceanic climates
  are often quite variable, so that the sites of
  high concentrations of a tuna stock may move
  great distances from year to year. Hence fish
  concentrations within any particular EEZ may
  fluctuate substantially.
• Tuna are valuable market fish, and while some
  high-value species have been over-harvested, most
  stocks are not yet severely depleted. Thus,
  fishing fleets from many distant-water nations
  are attracted to all major tuna fisheries.
• These circumstances complicate cooperative
  management, since there are a great many
  stakeholders, with differing rights and playing
  differing roles in the fishery.
• The strategic model is designed to explore the
  impact of all these circumstances on effective
  fishery management, and the interrelations among
  the stake-holders.

• Both the stock-recruitment parameter b and the
  stock-split parameter q are random and
  imperfectly observed.
• Each fleets objective is to maximize the
  discounted sum of its annual payoffs to the time
  horizon T.

Pacific Tuna Catch and the El Nino 3 Index
1995-1998
Influence of harvesting cost/price ratio for
different knowledge structures. Symmetric game.
Global Tuna Fisheries are Big Business

• The Stochastic Imperfect Information
  Split-stream Game
• A generic game, adjustable to describe
  various species-types, e.g. salmon, migrating
  along coastlines and small schooling species
  (sardines, anchovies) which are easily
  over-harvested and subject to large annual
  demographic variations.
• 
  
• A migratory stock splits into two sub-streams,
  each accessible to only one of two competing
  harvesting fleets. After harvest, the residual
  sub-streams reunite to form the brood-stock for
  the subsequent generation.
• The new generation grows to maturity and once
  again splits into sub-streams. The growth and
  subsequent split processes both are stochastic,
  and imperfectly observed by the fleets. The
  process repeats annually.
• Each fleet chooses its harvest policy to
  maximize the expected discounted sum of annual
  net payoffs over time, in response to the policy
  of its competitor. Thus it harvests to optimize
  the balance of immediate net return and its
  expected share of anticipated future returns.
• The model demonstrates that, in
  non-cooperative harvesting, enhancing the quality
  of the fleets information may actually be
  destructive to boththus contradicting the maxim
  that information transparency is always
  advantageous.
• (Details on the far-right panel.)

Thousand US
Cur full current knowledge, both fleets Min
minimum current knowledge (only probability
distribution) both Cur-Min a outcome for fleet
a when it has full knowledge and fleet b has only
minimal knowledge Cur-Min b corresponding
outcome for player b
Knowledge structure
Effects of an El Niño
All alternatives competitive unless marked Coop
cooperative Note Information Inversions i.e.
enhancing information yields inferior outcomes
Harvest of tuna by zones (1950-1997)
Application to small schooling pelagics(e.g.
sardines-anchovies)

• Multilateral Management of Highly Migratory
  Stocks through a Regional Fisheries Management
  Commission (RFMC)
• The RFMC Strategic Game Model assumes M
  countries with fleets and N coastal countries
  whose extended economic zones (EEZs) intersect
  the range of a highly migratory harvested
  fish-stock. Here M,N may be large, and while
  they may intersect, usually differ.
• All of these stake-holding countries are
  members of the RFMC. The Commission makes long
  term policy for the fishery, gradually lowering
  the overcapacity of the fleets and thereby
  raising the brood-stock level for sustainability
  and profitability. It does this by negotiating
  annually the distribution of harvest capacity
  among the fleets and/or vessel days of harvest
  among the coastal states.
• With these regulatory rules, it sets up
  annually a sequence of single-season competitive
  sub-games among the fleets and coastal states.
  Thus each coastal state sets its harvest-access
  royalties and allocates its vessel-days of
  harvest among the fleets, and each fleet
  allocates its vessel distribution among coastal
  sites.
• The Tropical-Pacific tuna case, displayed on
  the central panel, illustrates the management
  issues to be addressed through this modeling
  analysis. (Currently the strategic model is
  being fine-tuned.)

• Specifies symmetry in information but erratic
  temporal changes in stock growth and
  distribution. (Figures show mean values over
  time)
• Usually the growth rate is high, and stock-split
  favors the a-site, but rarely the growth rate is
  small, with stock almost exclusively at b-site.
  The tight schooling favors over-harvesting.
• The cooperative solution based on bargaining
  which is responsive to the asymmetries in the
  fleets competitive strengths.

Note dominance of Pacific Center Region (i.e.
Tropical Pacific) and the dominance of Pacific
Center West over Pacific Center East

• Management issues in the
• Tropical Western Pacific
• The new Western and Central Pacific Tuna
  Commission treaty came into effect in late 2004,
  with substantive debate in 2005. Unusually, the
  island states have been working together closely,
  especially the 8 haves (parties to the Nauru
  Agreement). Their interests deviate from those
  of the 7 sometimes haves at the opposite pole
  of the El Nino cycle, plus Australia and New
  Zealand.
• The harvesting nations do not cooperate, but
  instead compete with one another for access to
  the island states EEZs. Nevertheless, they
  retain the lions share of the returns from
  harvest.
• The island states hope that, by further
  cooperation, they will achieve a higher share of
  harvest returns. To this end, they insisted on a
  voting system within the RFMC which gives them,
  collectively, a veto over its decisions. They
  also have insisted that the regulatory controls
  allocate vessel-days of harvest to them, rather
  than allocating harvesting capacity to the
  fleets.
• Our RFMC model is designed to examine the
  effectiveness, and societal implications of these
  circumstances

Varying the Fleets Risk Attitude

• Tropical Tuna in the Western Pacific Specific
  Circumstances
• Stock ranges are huge, and contain an unusually
  large number of coastal states, mainly island
  states, within the stocks range. Further most
  tuna are harvested within their EEZs rather than
  on the high seas.
• However stock concentrations move about from year
  to year, across vast distances, following the
  loci of food concentrations, as determined by
  ocean currents and surface temperatures.
• These tuna stocks are the largest and most
  valuable in the world, attracting many
  distant-water fleets (DWFs) worldwide. Few
  coastal states have substantial fleets of their
  own, and the DWFs take 85 of the catch.
• Over time, stock-harvest has expanded
  dramatically, with total fleet size becoming
  excessive, so that stocks are increasingly
  vulnerable to over-harvest.

• Risk attitude risk averse for d lt 1, risk
  neutral at d 1 risk accepting for d gt 1
• Note small inversion on left (where fleets
  highly risk averse), but not on right (where
  fleets mildly risk averse or mildly risk
  accepting)
• As risk aversion increases (i.e. d ? 0) then
  competitive outcomes merge with cooperative
  outcomes.