Exporting FMSY to other stocks in two parts

“Exporting” FMSY to other stocks in two parts: • Regression model with life-history covariates • Theoretical and empirical linkages to life-history theory Henrik Gislason, Michael Melnychuk, John Pope, Erla Sturludóttir, Henrik Sparholt, Gunnar Stefansson

Is FMSY predictable? Many possible influences hypothesized: • • • Body length, weight Age at maturity and age at selection Trophic level and index of cannibalism Growth parameters Natural mortality • • Nested taxonomic levels Ecoregion Habitat type Environmental temperature

Filtering & variable selection approach • Assess collinearity among numerical predictor variables, reducing list to: L∞ * K ; M ; Age 50% maturity ; trophic level ; preferred temperature • Compare models with differing random effects: species ; taxonomic group ; aggregated ecoregion • Compare models with differing fixed effects • Compare influences on FMSY for different estimation models: assessment ; Schaefer ; Fox ; P-T meta-analysis ; aggregate

Distributions of FMSY estimates

Influences of numerical predictors on log(FMSY) Schaefer Trophic L L∞ * K M T˚ pref Age 50% mat Fox L∞ * K M Trophic L T˚ pref Age 50% mat P-T meta-analysis M L∞ * K Trophic L T˚ pref Age 50% mat Aggregate FMSY M L∞ * K Trophic L T˚ pref Age 50% mat

Species random intercept offsets from overall FMSY Schaefer Fox

Species random intercept offsets from overall FMSY P-T meta-analysis Aggregate FMSY

Proportion of variation explained by model conditional R 2 Schaefer 0. 64 Fox 0. 58 P-T meta-analysis, taxa pooled 0. 59 P-T meta-analysis, by taxa 0. 53 P-T empirical φ, by taxa 0. 53 Aggregate FMSY 0. 37