Comparative selectivity on length at maturity among Alaskan sockeye salmon fisheries Neala Kendall Tom Quinn School of Aquatic and Fishery Sciences University of Washington Seattle, WA, USA M. Bond

Need for evaluating harvest selection  Life history traits are changing over time (Darimont et al PNAS)

Need for quantifying harvest selection  Life history traits are changing over time (Darimont et al PNAS)  Importance of quantifying harvest selection (Carlson et al Ecology Letters, Edeline et al PNAS)

Need for quantifying harvest selection  Life history traits are changing over time (Darimont et al PNAS)  Importance of quantifying harvest selection (Carlson et al Ecology Letters, Edeline et al PNAS)  Evaluating implications of selection (Olsen et al Nature, Heino and God Ø 2002 Bulletin of Marine Science)

Difficult to quantify fishery selection Hutchings, Nature 2005 Fish caught Length Fishery selectivity curve ?

Alaska salmon: good model to study fishery selection  Semelparous  Anadromous  Length and age at maturity easy to measure  Know population size and structure  Long-term gillnet fisheries  Large, long term data set J. Carter

Research questions  Does fishery selection vary by fish length and sex? J. Carter

Research questions  Does fishery selection vary by fish length and sex?  Are fish length at age changes over time correlated with fishery selection? J. Carter

Alaska sockeye salmon fisheries Bristol Bay Upper Cook Inlet Alaska

Methods  Yearly historical fishery reconstruction  Characterize annual length distributions in catch, escapement (not caught), and total run  Estimate: 1) Length-specific vulnerability 2) Selection differentials

Length frequency histogram example Females Males

Vulnerability profiles by length Females Males

Selection on length: SSDs SSD Larger fish are escaping to spawn than are getting caught + - Smaller fish are escaping to spawn than are getting caught Standardized selection differential = length escapement – length total run std. deviation of length total run

J. Carter Female SSDs more consistently negative than male SSDs Female average Male average

Fishery selection on Bristol Bay sockeye salmon and length at age changes over time

Differential fishery selection on ocean age 2 fish Proportion frequency Standardized selection differential P < P = P < P = P < P = P = P = Nushagak Naknek-Kvichak Egegik Ugashik Togiak Nushagak Naknek-Kvichak

Average length of ocean age 2 fish has decreased over time Average length (mm) slope=-0.60 slope=-0.64 slope=-0.26 slope=-0.31 slope=-0.16 slope=-0.05 slope=-0.19 slope=-0.21 slope=-0.16 slope= Average length (mm) Egegik Ugashik Togiak Nushagak Naknek-Kvichak slope: P = slope: P = slope: P = slope: P = slope: P = 0.869

SSDs somewhat correlated with decreasing length at age over time

Conclusions  Are larger than average fish more vulnerable to being caught ? YES Fish caught ConclusionsConclusions

Males  Does the fishery harvest different lengths of males than females? YES Length Females Conclusions

 Are fish length at age changes over time correlated with fishery selection? YES Fishing districts that harvest larger fish show a greater decline in fish length at age over time Conclusions

 Mary Beth Loewen and Matt Foster, ADFG  Mark Willette and Terri Tobias, ADFG  Tim Baker, ADFG  Jeff Hard, NOAA Fisheries  Alaska Salmon Program, UW  Funding: School of Aquatic and Fishery Sciences Alaska Sustainable Salmon Fund National Science Foundation Gordon and Betty Moore Foundation Acknowledgements

J. Bennis J. Carter

Bristol Bay—100% gillnets

Chignik—100% purse seines

Alitak—mixed purse seines/gillnets

Vulnerability differs by length & sex example Females J. Carter Males

Females Vulnerability curves differ among years J. Bennis Length (mm) J. Carter

Vulnerability profiles Chignik female

Vulnerability profiles Chignik male

Vulnerability profiles Upper Cook Inlet female

Vulnerability profiles Upper Cook Inlet male

 Using estimated selectivities, model ideal length and age at maturity under different harvest scenarios (J Ø rgensen et al Evol. Apps.)  Calculate maturation reaction norms for spawning populations J. BennisJ. Carter Future work, extensions J. Bennis