2.0 



1.0 _ 



+ 



e 

 + © 



© 

 © 



KEY 



<•> = Nuttal's cockle ( Cfinocardium nuttalii) 

 a = Pacific razor clam (Siliqua patula ) 

 + = other mollusks 



2 



1 1 1 — l — III) 



3 4 5 10 



- 1 1 1 1 — I — l l l l — 



20 30 40 50 100 



MAXIMUM AGE, YEAR 



FIGURE 2.— Plot of instantaneous mortality rate (yr ') against maximum observed age (yrl, both on logarithmic scales, for 28 stocks of 



mollusks. 



is the only information available; e.g., McBride and 

 Brown (1980) summarized the life history param- 

 eters of the major fish stocks in the western North 

 Atlantic. For the following species McBride and 

 Brown gave estimates of maximum age but not of 

 natural mortality rates. The estimates provided here 

 were calculated by the regression method. 



Species 



Ocean pout, Macrozoarces americanus 



Scup, Stenotomus chrysops 



Black sea bass, Centropristis striatus 



White hake, Urophycis tenuis 



Bull shark, Carcharhinus leueas 1 



Dusky shark, Carcharhinus obscurus' 



Angler, Lophius americanus 



Tilefish, Lopholatiius chamaeleonticeps 



'Maximum ages for the sharks were taken from the growth study by Hoe- 

 nig (1979). The estimates are believed to be conservative. 



The major limitation of the technique is that the 

 sample size is not taken into consideration. The max- 



imum age observed depends on the number of 

 animals in the sample since rare, old animals are 

 more likely to be found in large samples. However, 

 once a sample of, say, 200 animals has been ex- 

 amined, the maximum age tends to increase slowly 

 with increasing sample size. The nature of the re- 

 lationship between sample size and maximum age is 

 examined in Appendix A. Because the sample size is 

 not taken into consideration, it is not possible to at- 

 tach confidence bounds to the estimates or to test 

 hypotheses. 



Another limitation is that the age structure will 

 change slowly following a decrease in the mortality 

 rate. Hence, the maximum age will remain depressed 

 for several years resulting in an overestimate. 



This regression technique appears to have con- 

 siderable predictive power for estimating mortality. 

 It is useful in a variety of situations where the data are 

 limited. However, the statistical foundation underly- 

 ing the technique is weak thus precluding the making 



901 



