SHERMAN ET AL.: COHERENCE IN ZOOPLANKTON 



Species Shifts in Dominance 



Although the three species co- occur on the shelf, 

 their temporal and spatial patterns of dominance are 

 different. These patterns are coherent among the 5 

 yr. The proportion of the total zooplankton account- 

 ing for these three dominant species is shown for 

 each subarea as a function of time in Figure 2b. In the 

 Gulf of Maine and over Georges Bank, C. fin- 

 marchicus, a species that overwinters in the cooler, 

 deep waters of the Gulf of Maine (Bigelow 1926), is 

 dominant in spring and early summer. During early 

 autumn, when temperatures in the upper layer are 

 warmest, dominance shifts to C. typicus, a species 

 which undergoes greatest egg production in water 

 warmer than 13°C (Dagg 1978). The shift from C. fin- 

 marc hicus to C. typicus dominance occurs earlier (in 

 late summer) on Georges Bank, where the change in 

 abundance is of greater magnitude and persists to 

 early winter. In the southern portion of the shelf, the 

 dominance of C. finmarchicus in late spring is 

 replaced by P. minutus, C. typicus, and other less- 

 abundant zooplankters, including other copepods, 

 cladocerans, larval echinoderms, salps, and barnacle 

 larvae in Southern New England and principally 

 cladocerans in the Mid-Atlantic Bight. Annual 

 deviations in the dominance patterns of C. fin- 

 marchicus, P. minutus, and C. typicus from the 5-yr 

 mean were insignificant at the 0.05 level in 95 V< of the 

 comparisons made within the subareas (Table 1). 



The numerical abundance of the three copepods 

 were coherent within the envelope of one standard 

 error of the mean and within the mean range in each 

 of the subareas during the 5 yr (Fig. 2c). The 

 zooplankton standing stocks, dominance patterns, 

 and abundance levels of the principal species in each 

 of the four subareas are different. The spring peak in 

 zooplankton standing stock in the Gulf of Maine and 

 on Georges Bank (Fig. 2a) is represented by C. fin- 

 marchicus in the Gulf of Maine and a combination of 

 C. finmarchicus and P. minutus on Georges Bank, 

 (Fig. 2b, c); the shift to C. typicus dominance in 

 autumn is not of sufficient magnitude to register a 

 secondary pulse in standing stock in the Gulf of 

 Maine or Georges Bank.In Southern New England 

 waters the bimodal peaks in zooplankton standing 

 stock are represented by C. finmarchicus and P. 

 minutus dominance in spring and early summer 

 followed by large-scale C. typicus swarming in late 

 summer and autumn (Fig. 2b, c). Further south in the 

 Mid-Atlantic Bight, C. finmarchicus abundance is 

 diminished, and is replaced by P. minutus and C. 

 typicus in late winter and early spring, followed by an 

 increase in the standing stock of zooplankton from 



summer through autumn (Fig. 2a) related to the 

 growing abundance of cladocerans and other 

 zooplankters in summer and large-scale swarming of 

 C. typicus in autumn (Fig. 2b, c). Deviations from the 

 5-yr mean temporal patterns of abundance of the 

 three dominant copepods were not significant at the 

 0.05 level in 95 ( /r of the comparisons (Table 1). 



DISCUSSION 



Observations on the zooplankton of the northeastern 

 continental shelf made during the past half century 

 (Bigelow 1926; Bigelow and Sears 1939; Grice and 

 Hart 1962; Judkins et aL 1980) can be divided into 

 four periods: 1) The first measurement of volumes 

 and species abundance made by Bigelow between 

 1912 and 1920, 2) the volume measurements by 

 Bigelow and Sears from 1929 to 1932, 3) the volume 

 and species measurements of Grice and Hart in 1 960, 

 and 4) the more contemporary measurements of 

 species abundance made by Judkins et al. in 1975. 

 Data from these studies were converted where pos- 

 sible from volumes per standard haul and volumes 

 per square meter to volumes per 100 m 3 ; data from 

 stations showing evidence of net clogging due to large 

 amounts of gelatinous zooplankton, large number of 

 organisms >2.5 cm length, or sampling gear and 

 methods differing significantly from MARMAP meth- 

 ods were excluded. Throughout the sampling periods 

 the mean seasonal zooplankton values of the earlier 

 investigators were not significantly different from the 

 mean values of the contemporary MARMAP data 

 base (Table 2). The greatest range in biomass from 

 year to year is on Georges Bank and is likely related to 

 variability in retention of zooplankton resulting from 

 the seasonal formation and decay of the Georges 

 Bank gyre (Butman et al. 1982). In the earlier studies 

 (Bigelow 1926; Bigelow and Sears 1939; Grice and 

 Hart 1962; Judkins et al. 1980) copepods were the 

 predominant zooplankters: Calanus finmarchicus 

 and Pseudocalanus minutus were the most abundant 

 species in the spring, with a shift to Centropages 

 typicus in late summer and autumn. These three 

 species are important links in the energetics of the 

 shelf ecosystem since they provide food for larval, 

 juvenile, and adult fish (Sherman and Honey 1971; 

 Sherman and Perkins 1971; Marakl974; Shermanet 

 al. 1981b; Cohen and Lough 1982). 



Our results provide evidence that the biomass and 

 species compositon of zooplankton have not changed 

 substantially over the past 70 yr. The persistent pat- 

 terns of abundance and species dominance reflect 

 coherence within the range of interannual variability 

 observed since the early part of the century. These 



857 



