estimates of M from the third method to be less bi- 

 ased than those from the first two. 



Usually, capture heterogeneity leads to the more 

 catchable animals joining the marked population, 

 and we expect marked animals to be more catchable 

 than unmarked animals. Capture heterogeneity, 

 however, is only prevalent among decayed in- 

 dividuals who are all less catchable than fresh, un- 

 marked individuals. Thus, capture heterogeneity, by 

 placing the more catchable decayed individuals in 

 the marked population, results in the capture prob- 

 ability of marked animals being closer to the cap- 

 ture probability of unmarked animals. This reduces 

 the negative bias in population size (N), immigra- 

 tion (B), and escapement (E) estimates, which was 

 caused by age-dependent catchability. Again, the 

 third method, by removing decayed individuals and 

 decreasing the fraction of the population which is 

 decayed, will not be affected by capture heteroge- 

 neity as strongly as the other two methods. 



Manly and Parr estimators will have the same 

 ameliorating affects because of capture heteroge- 

 neity as their Jolly- Seber counterparts. Since the 

 estimate of catchability, p, should be accurate for 

 the more catchable animals, estimated survival 

 should be accurate for that group. Bias would result 

 from correlations between catchability and survival. 

 Also, since p is estimated for marked (and thus 

 decayed) individuals, using the more catchable 

 decayed individuals to estimate p brings the 

 estimated catchability closer to the actual catch- 

 ability of the unmarked individuals. Again, this 

 reduces the bias in N, B, and E which is caused by 

 age-dependent catchability. 



There are other approaches to estimating param- 

 eters from populations with age-dependent survival 

 and capture rates. By placing carcasses in two readi- 

 ly identifiable age classes, fresh (and thus <1 wk old) 

 or decayed (and thus older than 1 wk), Pollock's 

 (1981) modified Jolly-Seber analysis of the data could 

 have been made. Since this method requires recap- 

 tures of decayed individuals, it could not be used to 

 analyze data from previous surveys, and it would 

 require more sampling effort in future surveys than 

 the method 3 estimate. If different age classes have 

 sufficiently different capture or survival rates, then 

 this method will provide more accurate estimates. 

 If not, then it will yield the same estimate as the 

 third method, but would have higher variances, as 

 more parameters are estimated. 



FISHERY BULLETIN: VOL. 84, NO. 2 



ACKNOWLEDGMENTS 



We would like to thank L. B. Boydston of the 

 California Department of Fish and Game for making 

 us aware of this problem, for many helpful discus- 

 sions, and for assisting in data collection. S. Sykes 

 was supported by the California Department of Fish 

 and Game during the sampling. We are grateful for 

 the comments by K. Pollock, T. Schoener, and N. 

 Matloff on an earlier version of this manuscript. We 

 would also like to thank Ivan Paulsen for assistance 

 in data collection. 



LITERATURE CITED 



BUCKLAND, S. T. 



1980. A modified analysis of the Jolly-Seber capture-recapture 

 model. Biometrics 36:419-435. 



Carothers, A. D. 



1971. An examination and extension of Leslie's test of equal 

 catchability. Biometrics 27:615-630. 

 Darroch, J. N. 



196 1 . The two-sample capture-recapture census when tagging 

 and sampling are stratified. Biometrika 48:241-260. 

 Jolly, G. M. 



1982. Mark-recapture models with parameters constant in 

 time Biometrics 38:301-321. 

 Manly, B. F. J. 



1970. A simulation study of animal population estimation 

 using the capture-recapture method. J. Appl. Ecol. 7:13-39. 

 Manly, B. F. J., and M. J. Parr. 



1968. A new method of estimating population size, survivor- 

 ship, and birth rate from capture-recapture data Trans. Soc 

 Br. Entomol. 18:81-89. 

 Parker, R. R. 



1968. Marine mortality schedules of pink salmon of the Bella 

 Coola River, central British Columbia. Can. J. Fish. Res. 

 Board 25:757-794. 

 Pollock, K. H. 



1981. Capture-recapture models allowing for age-dependent 

 survival and capture rates. Biometrics 37:521-529. 



Schaefer, M. B. 



1951. Estimation of the size of animal populations by mark- 

 ing experiments. U.S. Fish Wildl. Serv., Fish. Bull. 52: 

 191-203. 



Seber, G. A. F. 



1982. The estimation of animal abundance and related param- 

 eters. MacMillan Publishing Co., Inc, N.Y. 654 p. 



Stauffer, G. 



1970. Estimates of population parameters of the 1965 and 

 1966 adult Chinook salmon runs in the Green-Duwamish 

 River. M.S. Thesis, Univ. Washington, Seattle, 155 p. 

 Sykes, S. D. 



1982. Multiple mark-recapture estimators of salmon spawn- 

 ing runs sizes. M.S. Thesis, Univ. California, Davis, 64 p. 



270 



