1972 to 1994 for sectors of the northwest Atlantic shelf from eastern Nova Scotia to
the Labrador Sea (Fig. 9.11a) showed strong negative correlations (the opposing time-
series trends shown in Fig. 9.11b) between benthic fish stocks (again, cod and
company) and small pelagics (herring, etc.). The implied food-chain effect apparently
extended down the food web to phytoplankton estimated from CPR records. It was
low when fish stocks were high. Application of the trophic-cascade hypothesis seems
straightforward, but ocean ecology consistently provides alternative hypotheses.
Greene and Pershing (2007) have pointed out that, simultaneous with the cod
collapse, Arctic Ocean and North Atlantic circulation shifted, reducing temperature
and salinity downstream all the way to the mid-Atlantic Bight. They suggest that
enhanced stratification due to salinity reduction allowed stronger phytoplankton
blooms and increased abundance of small copepods (like Pseudocalanus and
Centropages), for which some data show strong shifts from about 1991 in the Gulf of
Maine and over Georges Bank. Thus, lower trophic-level impacts reasonably
attributable to top-down effects may instead (or also!) have had bottom-up causes.
Ocean ecology is not an experimental science with excellent control of all conditions
except those under test.
Fig. 9.11 (a) Subdivision into fisheries statistical areas (1 through 9) of the northwest
Atlantic continental shelf. (b) Comparison from 1975 to 1994 from fishery-agency
census trawling of groundfish (solid lines) and small pelagic fish abundances (dotted
lines with some points as •). The plots are values of annual mean numbers/trawl haul
plotted as standardized anomalies from the long-term means. The data included a total
of 55,043 trawls that collected 26,286,369 fish of 412 species.
(^) (After Frank et al. 2005.)
Shrimp-for-cod substitutions similar to those adjacent to maritime Canada and New
England have been observed in coastal Alaskan fisheries. See Baum and Worm (2009)