Kane: Zooplankton biomass and species abundance on Georges Bank 



467 



ton biomass can screen important biological events by 

 repressing anomalous values. Therefore, nonparamet- 

 ric statistical techniques were employed to test for dif- 

 ferences between years within the defined seasons. The 

 Kruskal-Wallis ANOVA was used to determine if sig- 

 nificant (P<0.05) differences existed between years 

 within each season. The test showed that all biomass 

 and species seasonal groupings had at least one pair 

 of years different from each other. The Dunns multiple 

 comparison procedure was applied to pinpoint the 

 anomalous year(s). 



Results 



Annual cycle 



Median biomass and dominant copepod abundance val- 

 ues for each of the defined seasons by year and for the 

 ten-year study period are given in Table 1, A-F. 



Biomass on average increased fourfold in early spring 

 from its winter low. It peaked in late spring and then 

 gradually declined through the summer and autumn 

 seasons (Fig. 2). Overall, zooplankton standing stock 



CO 



E 



E 

 o 



CD 



50 100 150 200 250 300 350 



Survey Mid Point (jday) 



CO 



E 



50 100 150 200 250 300 350 



Survey Mid Point (jday) 



Figure 2 



(A) Annual cycle of seasonal median biomass values in 1977 

 through 1979 and the ten-year median. (B) Annual cycle of 

 seasonal median biomass values in 1982 through 1984 and 

 the ten-year median. 



was high from 1977 through 1979. Seasonal medians 

 from early spring through late autumn were above 

 time-series median values (Fig. 2A). The opposite pat- 

 tern was evident from 1982 through 1984. Zooplank- 

 ton standing stock was below average throughout these 

 years, except for late autumn 1982 (Fig. 2B). Winter 

 biomass showed little variation throughout the study 

 period (Table 1A). 



The seasonal biomass cycle on Georges Bank from 

 1977 through 1981 was coherent; annual departures 

 from the mean were insignificant (Sherman et al., 

 1983). However, substantial departures from the aver- 

 age seasonal cycle occurred on several occasions after 



1981. Surveys in 1983, 1985, and 1986 recorded peak 

 biomass in early spring instead of late spring and in 

 two of those years, 1983 and 1986, there was an anoma- 

 lous increase in biomass during the summer months 

 (Fig. 3). However, because only below average summer 

 levels were reached during these years, the latter in- 

 creases appear to represent a recovery from the mini- 

 mal levels measured in late spring, rather than a sum- 

 mer bloom. Zooplankton standing stock increased 

 between summer and early autumn in only two years: 



1982 and 1985 (Table 1, D-E). The usual decline in 

 biomass between early and late autumn was not ob- 

 served in 1982 (Table 1, E-F). 



The average ten-year seasonal distribution of each 

 variable as a function of bottom depth is depicted in 

 Figure 4. There were no apparent long-term distribu- 

 tion shifts in biomass or species abundance during the 

 ten-year period. 



Calanus finmarchicus and P. minutus both domi- 

 nated the zooplankton community in early and late 

 spring. During the summer, their populations declined 

 and C. typicus began to increase in abundance until it 

 peaked in early autumn (Table 1). Centropages hamatus 

 and M. lucens were only prevalent in specific depth 

 strata (Fig. 4). Centropages hamatus was almost en- 

 tirely restricted to the central shoal depth region, peak- 

 ing there during summer months. Metridia lucens was 

 most numerous in deep water where its large size 

 caused it to be a major contributor to spring biomass. 

 Overall, it was the only dominant species that exhib- 

 ited a long-term abundance trend. Population estimates 

 for M. lucens were low in 1983 through 1986 (Table 1). 



All the above species showed departures from their 

 average seasonal cycles during their periods of peak 

 abundance. Calanus finmarchicus, M. lucens, and P. 

 minutus all declined between early and late spring in 



1983 and 1986 (Table 1, B-C). Centropages typicus 

 departed from its typical annual abundance pattern 

 during three years; population estimates in 1983 and 

 1986 declined from summer to early autumn and in 



1982, C. typicus did not reach peak abundance until 

 late autumn. The C. hamatus population was more vari- 

 able than the other species. From 1977 to 1981, abun- 



