MORSE: CATCHABILITY, GROWTH. AND MORTALITY OF LARVAL FISHES 



ential regression equations fitted to the mean 

 catch per 100 m~ and to the length for each taxa 

 are shown in Table 4. The regi-ession fitted to 

 twilight catches is often not as accurate as for 

 day or night. This is not sui-prising because twi- 

 light catches represent the transitional time be- 

 tween light and dark regimes when visual avoid- 

 ance responses to the net are expected to be 

 most variable. This variability is evident in the 

 often high twilight catches at the larger sizes. 



The ratios of expected catches calculated from 

 the exponential regi-essions for night:day and 

 twilight:day catches were fit to polynomial equa- 

 tions and yield corrections for catchability for 

 day, night, and twilight catches (Table 4). 



Mortality 



Catches of each taxa were corrected for day, 

 night, and twilight catchabiHty, and the mean 

 catch per tow (number/100 m^) was again calcu- 

 lated for each mm length and the exponential 

 decay model fit to the length frequencies. The 

 slope of the fitted line was used as an estimate of 

 length-dependent mortality (Ebert 1973). Mor- 

 talities ranged from 0.114 for rock gunnel to 

 0.701 for Atlantic mackerel (Table 5). A review 

 of the spawning times of the 26 taxa (Colton et 

 al. 1979; Morse et al. 1987) reveals that 4 of the 6 

 taxa with the lowest mortalities are winter 

 spawners, and 5 of the 6 taxa with the highest 



(C) versus larval length (mm) as C = a * exp(length « b) for day, night, and twilight by taxon for larvae caught off the northeast 

 twilight:day mean catches verses larval length (/.) where R = a * b, L + bz L^. Numbers in parenthesis are standard error of 



437 



