NOTE Johnson; Recruitment variability of Limanda ferruginea 



855 



merited, Calanus sp. and Pseudocalanus sp. are both 

 dominant representatives of the zooplankton commu- 

 nity of the North Atlantic and are important prey 

 for many spring spawned fish larvae (Gushing, 1982; 

 Runge, 1988). The widespread abundance and distribu- 

 tion of Calanus and Pseudocalanus favor them to be 

 the most likely prey choice of yellowtail flounder larvae. 

 Collectively, monthly distributions and mean densities 

 of zooplankton for May and June coincide with peak 

 occurrences of yellowtail flounder (Figs. 2 and 3). The 

 Spearman's rank-order correlation analysis was used to 

 measure the association between ichthyoplankton and 

 their zooplankton prey. 



An adaptation of the "larval food supply model" ( Mertz 

 and Myers, 1994) was used to test the match-mismatch 

 hypothesis, the relationship between spawning vari- 

 ability, and the variability of recruitment. The recruit- 

 ment estimates were obtained from the Northeast 

 Fisheries Science Center (NEFC, 1991). Two aspects 

 were explored: 1) whether or not a relationship exists 

 between spawning duration and recruitment variabil- 

 ity and 2) whether deviations from peak spawning 

 between yellowtail flounder larvae and prey are related 

 to recruitment variability. 



The mean and standard deviation of observed abun- 

 dance versus time was calculated for both yellowtail 

 larvae and their prey for each year and for both sub- 

 areas. The width and overlap of the abundance curves 

 for predator and prey were analyzed. Selected methods 

 from the Mertz and Myers ( 1994) study were applied for 

 partial use of this analysis. The following parameters — 



ip = timing between peaks of larval production and 

 food supply; 

 At^ = annual differences in (?q) from its mean value; 



8 = one-half width of the production period for larvae; 

 and 



(7 = one-half width of the production period for zoo- 

 plankton 



— were used to calculate 1) variability in peak timing 

 from the mean for individual species (a and 5), 2) vari- 

 ability in timing between larval spawning and peak 

 zooplankton production (?q), and 3) year-to-year vari- 

 ability in peak spawning and production (AIq). When ?(, = 

 0, the match between the peak larvae and production of 

 zooplankton is exact (Mertz and Myers, 1994). 



Results and conclusions 



Variability in timing for individual species 



30,000 I 



- 20,000 2 



0,000 



100 175 250 



T 30,000 N 



0,000 



10,000 



3 



400 



Day of year 



Figure 1 



(A). A match represented by the overlap of fish lai-vae and its 

 prey, of the match-mismatch hypothesis. (B). A mismatch rep- 

 resented by the lag in time and space between fish larvae and 

 its prey. {S and a represent one-half width of the production 

 period of the match-mismatch hypothesis for larvae and zoo- 

 plankton.) Diagram adapted from Gushing (1990). 



slightly under 1 day (SNE). The greatest span of time 

 when zooplankton was abundant was 24 days for GB 

 during 1981 and 35 days for SNE during 1982. The low or 

 high zooplankton occurrence was not associated with the 

 strengths or weaknesses of yellowtail recruitment. Both 

 strong (1981) and weak (1982) year classes of yellowtail 

 flounder occurred during times of high variability in peak 

 production of zooplankton. 



Zooplankton The duration of the Calanus sp. peak pro- 

 duction (a) varied from 3.4 to 26.5 days for both subareas. 

 The narrowest distribution occurred during 1984 with 3.4 

 days, and extremes occurred in 1977 with greater than 

 12 days for GB and 26.5 days for SNE. The greatest 

 span in peak production for Pseudocalanus sp. was 33 

 days. Smallest duration ranged between 7.8 days (GB) to 



Ichthyoplankton The difference in timing of peak spawn- 

 ing {Atg) from its mean value for yellowtail larvae for 

 both subareas ranged from 1.4 days to approximately 

 10 days. The largest values of At^ occurred in GB, and 

 again large differences in the timing of spawning did not 

 always reflect years with poor recruitment. Two strong 

 year classes occurred with extreme values of At^: 1977 



