Kane: Spatial and temporal abundance patterns of Centropages hamatus 
95 
els year round and begins to increase inshore from 
its annual low in late spring-early summer (Fig. 6) 
when Centropages hamatus is at peak abundance. 
MARMAP data indicate that it is unlikely that the 
summer decline or the abundance levels reached by 
C. hamatus are controlled substantially by C. typicus 
predation. There was no strong inverse relationship 
Figure 6 
Time series monthly log mean abundance (solid line) and 
the 95% confidence interval (dashed line) of the mean for 
the following copepod predators in the Southern New En- 
gland subarea: ctenophores, the copepod Centropages 
typicus, and chaetognaths. Single points are the log mean 
abundance of the taxon for individual surveys during cer- 
tain years. The error bars indicate the 95% confidence in- 
terval of the mean. 
between the abundance trends of the two species. 
For example, in 1987 C. hamatus reached peak abun- 
dance earlier than usual, in late April, and declined 
rapidly to below average levels (Table 1). The abun- 
dance of C. typicus was average in late April 1987 
and also declined through the summer to below av- 
erage levels (Fig. 6). High levels of C. hamatus re- 
corded in 1984 (Fig. 4) were not due to the absence 
of C. typicus predators; abundance was close to av- 
erage for the copepod during spring and summer (Fig. 
6). Monthly partial correlation coefficients between 
station abundance values of the two species during 
the time series were positive (0.07-0.24) from April 
through August, further evidence that predation by 
C. typicus is minimal. 
Peaks of Centropages hama tus abundance and the 
presence of chaetognaths do coincide in the SNE sub- 
area (Figs. 3 and 6). However, evidence that chaetog- 
nathan predation impacts C. hamatus abundance 
could not be found. All of the surveys that had ex- 
ceptional high or low C. hamatus abundance, 1979, 
1984, and 1987 (Fig. 4), had near average chaetog- 
nath density (Fig. 6). Conversely, C. hamatus abun- 
dance was close to average when chaetognath den- 
sity was high in 1977 and low in 1985 (Fig. 6). 
Monthly partial correlation coefficients between sta- 
tion abundance values of the two species were not 
significant and very low (-0.10-0.25) throughout the 
year, indicating that chaetognath predation has little 
effect on C. hamatus abundance. 
Discussion 
Temperature affects most processes in marine eco- 
systems and the life cycle of Centropages hamatus is 
no exception. Opposite extremes in temperature ap- 
pear to limit the seasonal occurrence of the popula- 
tion at the southern and northern ends of the eco- 
system. Warm summer temperatures in the MAB 
were correlated with the rapid decline of the cope- 
pod in this area as values approach or surpass the 
critical upper thermal level for the species. Similar 
relationships between temperature and C. hamatus 
were found by Deevey (1960) for the population 
present near and within Delaware Bay. She reported 
that the copepod disappears as temperatures rise in 
summer but is present year round in small numbers 
during cool summers. Grant ( 1988) also reported that 
C. hamatus abundance in the MAB declines with 
increasing temperature and is absent in some years 
during summer and fall seasons. The MAB popula- 
tion begins to reappear or increase close inshore in 
late autumn where waters cool faster than those off- 
shore. Populations farther north decline slowly as 
