Kane: Spatial and temporal abundance patterns of Centropages hamatus 
93 
area there were no strong correlations for extended 
periods between variables. 
Chlorophyll Estimates of the abundance of Centro- 
pages hamatus were highest at locations where chloro- 
phyll biomass was also high (Fig. 5B). Total chlorophyll 
and abundance measures at stations were significantly 
(PcO.Ol) correlated during certain times of the year in 
all subareas (Table 2). In the MAB, variables were posi- 
<7 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ;>21 
Surface temperature (°C) 
< 0.2 0.4 0.6 0.8 1.0 1.4 1.8 2.2 3.5 ;> 3.5 
Chlorophyll (mg/m 3 ) 
<31 31 31.5 32 32.5 33 33.5 34 34.5 *35 
Salinity (psu) 
Figure 5 
Mean abundance of Centropages hamatus by (A) surface 
temperature, (B) chlorophyll, and (C) salinity interval. All 
time series data from the entire survey area was used. 
tively correlated from May through July and, in SNE 
waters, during October and February. Variables on GBK 
were positively correlated from May through January, 
except for October. GOM correlations were significantly 
positive in August, November, and December. 
Partitioning of total chlorophyll values into net- 
plankton and nanoplankton size fractions did not 
typically change the correlation coefficients between 
Centropages hamatus and phytoplankton abundance 
listed in Table 2. There were a few scattered months 
in the subareas where coefficients with netplankton 
were 0. 1-0.2 units higher. The most substantial change 
occurred during October on GBK. The correlation coef- 
ficient with netplankton was 0.27 units above the value 
in Table 2 and was positively correlated (P=0.02). 
Salinity Centropages hamatus was present at sta- 
tions where integrated water-column salinity ranged 
from 27.09 to 36.00 psu. Maximum abundance oc- 
curred in the lower region of this range (Fig. 5C). 
Monthly correlation coefficients between station 
abundance and salinity were usually negative and 
oftentimes significant during the year (Table 2). No- 
table were the comparatively high negative correla- 
tions found during January in both the MAB and 
SNE subareas. Values in the MAB were also nega- 
tively correlated in February and again in August 
and September. SNE correlations were also signifi- 
cantly negatively correlated during April, July, and 
from September through December. GBK correla- 
tions, though not always significant, were positive 
from February through July and negative in the re- 
maining six months. GOM coefficients were gener- 
ally weak throughout the year. 
Predation Pressure On average, Centropages 
hamatus and ctenophores both reach peak abundance 
during June in the SNE subarea (Figs. 3 and 6). 
During June and July of 1981 a large patch (9-12 
stations) of ctenophores occupied inshore waters in 
the southern region of the subarea offshore of Long 
Island, New York. This concentration pushed over- 
all mean abundance in the subarea to an 11-year high 
(Fig. 6). Predation on C. hamatus was apparently 
minimal; its mean abundance in late spring 1981 was 
slightly above the 11-year average (Table 1; Fig. 3). 
However, the abundance of C. hamatus in June 
within a ctenophore patch was much lower (611/ 
100m 3 ) than outside (2,712/lOOm 3 ) it. Evidence for 
predation pressure was also found in the July survey; 
C. hamatus density was 8,138/lQQm 3 where it co- 
occured with ctenophores, 22,87 l/100m 3 where cteno- 
phores were absent. 
In the SNE subarea, the omnivorous copepod 
Centropages typicus is present at relatively high lev- 
