284 
Fishery Bulletin 95(2), 1997 
Assumed N(K+ 1) 
Figure 3 
Annual fishing mortality rate (F A ) and corrected catch for 3 different cases when 
the catch occurs under the same conditions as those in Figure 2. 
lation size for that cohort at the start of the next 
year and thus the annual F will reduce N 1 to N K+V 
The annual F is then applied in the catch equation 
to generate the corrected catch (C A ) (step 4 in the 
algorithm). The resulting catch is distributed 
throughout the year according to a constant F and 
thus reflects a smooth population abundance decline 
(see Figs. 5 and 6). In both Figures 5 and 6, the sum 
of the observed monthly catches are different from 
the sum of the corrected monthly catches, whereas 
the observed and corrected population numbers fol- 
low different paths to the same endpoint. Figures 1 
and 5 have the same observed catch, but the popula- 
tion numbers (and resulting fishing mortality rates) 
under the assumption inherent in the catch equa- 
tion are different owing to the corrected catch value 
used in Figure 5. It is exactly the nonalignment of 
endpoints in Figure 1 that causes the bias in virtual 
