WKLife

WKLife

ToR

  • Establish relationships between life-histories and biological reference points in order to develop robust proxies for \(F_{MSY}\) and \(F_{lim}\).

Long-term goals

  • Develop robust catch rules
  • Condition Operating Models on
    • fishing history, depletion levels, selectivity assumptions, natural mortality.
  • Diagnostics
  • Consider nature of Time-series

mydas

  • Develop Proxy \(MSY\) Reference Points across the spectrum of data-limited stocks.
  • Diagnostics for screening reference points, assessment methods, HCRs and MPs
  • Management Strategy Evaluation to develop plans to manage risk given uncertainty

Screening Indicators

Indicators

  • \(L_{max5\%}\) mean length of largest 5%
  • \(L_{25\%}\) \(25^{th}\) percentile of length distribution
  • \(L_{c}\) Length at \(50\%\) of modal abundance
  • \(L_{mean}\) Mean length of individuals \(> L\)
  • \(L_{bar}\) Mean length
  • \(L_{max_{y}}\) Length class with maximum biomass in catch
  • \(L_{95\%}\) \(95^{th}\) percentile
  • \(P_{mega}\) Proportion of individuals above \(L_{opt} + 10\%\)

Reference points

  • \(L_{opt} = L_{\infty}\frac{3}{3+\frac{M}{K}}\), assuming \(M/K = 1.5\) gives \(\frac{2}{3}L_{\infty}\)
  • \(L_{F=M} = 0.75l_c+0.25l_{\infty}\)
  • \(f^{'} = \bar{L}_{y-1}/L_{mat}\)
  • \(f^{''} = min(1, \bar{L}_{y-1}/L_{mat})\)
  • \(f^{'''} = min(1, \bar{L}_{y-1}(1-d)/L_{mat})\) for d > 0

Robustness

Simulate stocks, different values of \(k\), \(h\), … and fuctional forms and compare proxies to \(MSY\) reference points

Life History Parameters

Life history parameters by species, species ordered by $k$

Life history parameters by species, species ordered by \(k\)

Simulations

Simulated time series

Simulated time series

Length distributions

Sampled length distributions for different periods.

Sampled length distributions for different periods.

Receiver Operating Characteristic

Receiver Operating Characteristic curves for recovery period, the points correpond to the ICES reference level (red) and the best descriminate threshold (black), ordered by k and percentile.

Receiver Operating Characteristic curves for recovery period, the points correpond to the ICES reference level (red) and the best descriminate threshold (black), ordered by k and percentile.

Area under the curve

Area under ROC curve, a good length based indicator should have a high area close to 1.

Area under ROC curve, a good length based indicator should have a high area close to 1.

Impact of uncertainty

Euclidean distance from TPR=1 and FPR=0.

Euclidean distance from TPR=1 and FPR=0.

Emprical Control Rules

Used Machine Learning

  • Perform random search then use
  • Support vector regression to model relationship between performance measures and control parameters
  • Genetic algorithms to find pareto Frontiers

Pareto Frontiers

Time series for an example simulation.

Time series for an example simulation.

Strict condition \(P(B>B_{lim}\)) > 0.95

Simulations for random combinations of control parameters (k1 and k2), colours showing probability of Blim > 95%.

Simulations for random combinations of control parameters (k1 and k2), colours showing probability of Blim > 95%.

Pareto Optimal Solutions

A Pareto plot for objectives corresponding to maximising yield  and minimising average annual variation in yield (AAV), non-optimal solutions and optimal solutions that lie along the Pareto front are indicated.

A Pareto plot for objectives corresponding to maximising yield and minimising average annual variation in yield (AAV), non-optimal solutions and optimal solutions that lie along the Pareto front are indicated.

Pareto Frontiers

A Pareto plot for objectives corresponding to maximising yield  and minimising average annual variation in yield (AAV), non-optimal solutions and optimal solutions that lie along the Pareto front are indicated.

A Pareto plot for objectives corresponding to maximising yield and minimising average annual variation in yield (AAV), non-optimal solutions and optimal solutions that lie along the Pareto front are indicated.

Calibration Regressions

Calibration regressions identifing the values of k1 and k2 to achieve a given Pareto optimal solution. Vertical lines correspond to $80% MSY$, AAV value derive from the Pareto optimum.

Calibration regressions identifing the values of k1 and k2 to achieve a given Pareto optimal solution. Vertical lines correspond to \(80% MSY\), AAV value derive from the Pareto optimum.

MSE

Summary statistics, for the MSEs run for the a  single set of control parameters, k1=0.6 and k2=0.6.

Summary statistics, for the MSEs run for the a single set of control parameters, k1=0.6 and k2=0.6.

Elasticity

Simulations

Kobe phase plot with trajectory.

Kobe phase plot with trajectory.

Biomass

Elasticity of $SSB/SSB_{MSY}$ as a function of fishing mortality.

Elasticity of \(SSB/SSB_{MSY}\) as a function of fishing mortality.

Harvest rate

Elasticity of $F/F_{MSY}$ as a function of fishing mortality.

Elasticity of \(F/F_{MSY}\) as a function of fishing mortality.