P-129
A Method to Evaluate Genetic Baselines That Use Both Diploid and Haploid Data, Which Avoids Overly Optimistic Simulations
Michael Garvin
,
Fisheries, University of Alaska Fairbanks, Juneau, AK
Michele M. Masuda
,
Auke Bay Laboratories, National Marine Fisheries Service, Alaska Fisheries Science Center, Juneau, AK
Jerome Pella
,
Auke Bay Laboratories, National Marine Fisheries Service, Alaska Fisheries Science Center, Juneau, AK
Rachel Riley
,
Division of Fisheries, University of Alaska Fairbanks, Juneau, AK
Adam Fuller
,
Harry K. Dupree Stuttgart National Aquaculture Research Center, Stuttgart, AR
Vladimir Brykov
,
Zhirmunsky Institute of Marine Biology, Russian Academy of Sciences, Vladivostock, Russia
Anthony J. Gharrett
,
Division of Fisheries, University of Alaska Fairbanks, Juneau, AK
Accurate genetic baselines are essential to perform mixed stock analysis, which is a major component of salmon conservation and management efforts. Current methods to evaluate the accuracy and precision of genetic baselines are either overly-optimistic, or do not accept haploid data. We developed a cross-validation procedure that accepts both haploid and diploid data, avoids overly-optimistic simulations, and can use maximum likelihood or Bayesian methods to evaluate the precision and accuracy of any genetic baseline.