566 



Abstract. ^The blood protein trans- 

 ferrin is an important source for iden- 

 tifying genetic variation in coho sal- 

 mon, Oncorhynchus kisutch, with 

 starch gel electrophoresis. We present 

 new data for transferrin allele frequen- 

 cies for 48 samples of coho salmon col- 

 lected at 34 different locations from 

 Oregon to southern British Columbia. 

 To analyze transferrin allele frequen- 

 cies from a larger geographic area, data 

 from various sources were compiled for 

 135 samples from sites from California 

 to southern British Columbia. In a sta- 

 tistical analysis of temporal variation 

 within locations, 26 of 63 pairwise com- 

 parisons (41%) showed significant dif- 

 ferences between samples (P<0.05). An 

 analysis of variance revealed signifi- 

 cant between-locale variability and no 

 significant within-locale variability due 

 to whether or not the brood years of the 

 samples were three years apart. Rela- 

 tive gene diversity values were 22.3% 

 between geographic regions, 2.1% be- 

 tween stocks within geographic regions, 

 3.1% between temporal comparisons, 

 and 72.5% for within-sample variation. 

 Total genetic diversity (H T ) was 0.586, 

 and the average genetic diversity 

 within populations (H s ) equaled 0.425. 

 The samples showed genetic variation 

 that was related to geographic location. 

 Applications of transferrin data to ge- 

 netic stock identification are discussed 

 in relation to these results. 



Transferrin polymorphism In coho 

 salmon, Oncorhynchus kisutch, and its 

 application to genetic stock 

 identification 



Donald M. Van Doornik 

 George B. Milner 



Manchester Field Station. Northwest Fisheries Science Center 



National Marine Fisheries Service, NOAA 



PO Box 130. Manchester. Washington 98353 



Gary A. Winans 



Northwest Fisheries Science Center 



National Marine Fisheries Service. NOAA 



2725 Montlake Blvd E. Seattle, Washington 981 12 



Manuscript received 25 April 1996. 

 Fishery Bulletin 94:566-5751 19961. 



The iron binding protein, transfer- 

 rin (Tf), is a highly polymorphic lo- 

 cus in coho salmon, Oncorhynchus 

 kisutch, that exhibits allelic fre- 

 quency differences between popula- 

 tions (Utter et al., 1970; Hjort and 

 Schreck, 1982; Olin, 1984; Bartley 

 et al., 1992). The transferrin locus 

 is a potentially valuable addition to 

 the set of polymorphic loci currently 

 used for genetic stock identification 

 (GSI) of coho salmon (Milner, 1993). 

 In the past, one drawback of includ- 

 ing transferrin in multilocus GSI 

 surveys was the necessity of acquir- 

 ing a serum sample to resolve al- 

 lelic variation. However, transferrin 

 allele frequencies can now be ob- 

 tained from heart tissue (Van 

 Doornik et al., 1995), therefore it is 

 feasible to incorporate variation at 

 this locus into GSI analyses involv- 

 ing coho salmon. 



This study was conducted as part 

 of an ongoing research effort to de- 

 velop a geographically extensive 

 baseline of allele frequencies for 

 coho salmon. Our objectives are to 

 increase the number of polymorphic 

 loci for which allele-frequency data 

 are obtainable, to verify the genetic 

 basis of previously documented and 



new allelic variation, and to expand 

 the geographic range of our baseline 

 data set. Given adequate geo- 

 graphic coverage and genetic differ- 

 ences among major contributing 

 stocks, allele-frequency data can be 

 used for stock identification of 

 mixed stock fisheries (Milner et al., 

 1985; Milner, 1993). 



The objective of this paper is to 

 examine the potential of transfer- 

 rin for use in genetic stock identifi- 

 cation of coho salmon. Specifically, 

 we 1) report new transferrin allele 

 frequencies for 48 samples collected 

 from 34 coho salmon stocks, 2) com- 

 pile allele-frequency data for trans- 

 ferrin data from other sources to 

 compare allele frequencies over a 

 larger geographic area, 3) analyze 

 the temporal stability of transfer- 

 rin allele frequencies, and 4) exam- 

 ine levels and patterns of variation 

 between samples using gene diver- 

 sity and genetic identity analyses. 



Materials and methods 



Blood or tissue samples, or both, 

 were collected, as described by Van 

 Doornik et al. (1995), for 48 coho 



