500
D. J. TRANTER:
v. Trophic Relationships
1. Evidence of Trophic Succession
If the seasonal variation ob$erved in nearly every biological property is the result of a
seasonal cycle of plankton production, then the variations in these properties ought to be
inter-related.
In its simplest form, the plankton cycle is conceived to be a "food chain" with algae
as the first link, algal herbivores (zooplankton) as the second, primary carnivores
(zooplankton) as the third, secondary carnivores (micronekton) as the fourth, pelagic
squid as the fifth, and predatory nekton as the sixth. Each trophic level is considered,
ideally, to draw on the energy resources of the level immediately below. So, matter is
conveyed along the food chain leaving depletion in its wake. Each standing stock increases
in succession, reaches a maximum, and then declines. The problem is to distinguish
changes such as these from those of horizontal translation.
Where peaks at adjacent trophic levels are close in phase, their values will be wellcorrelated (STEEMANN NIELSEN, 1957; CUSHING, 1958). Where the time lag is considerable, there will be high negative correlations - for example when grazing or predation
pressure is increasing or declining. With such a model of plankton production in mind,
I have sought to determine whether or not there is evidence of trophic succession in our
study area during the year of study.
It has often been proposed that time lags are short in warm waters (e.g. CUSHING,
1958). Therefore, in the analysis that follows of data from meridian 110 0 E, correlations
based on concurrent pairs of values are considered first of all.
Table 5 shows correlation values between various components of the zooplankton
catch, representing holoplankton, meroplankton, and each major trophic type. The table
shows that these diverse components are well correlated; they vary in a similar way.
Of a total of 112 correlations, no less than 102 were significant at the 95% level, of
which not more than 5 - 6 could be due to chance. Most of these (86) were significant
also at the 99% level. Of the 26 correlations that were not significant at this level, more
than half were for the tropical zone by night; of the 10 that were not significant even at
the 95% level, 5 involved the rather unique component fish eggs and larvae, 2 involved
decapod larvae, 2 Chaetognatha, and 1 Copepoda. The various components of the
midwater trawl catch are similarly well-correlated, particularly in the subtropical zone
(Table 6 b).
It is clear, then, that the time lag between the seasonal cycles of the various zooplankton components is very short. Consequently it is not possible to locate evidence
of trophic succession. It is necessary to extend the search more widely.
Table 6 shows the correlations between zooplankton (as a whole) and other biological
properties, viz. phytoplankton (chlorophyll a), productivity (light-saturated), phosphate,
and nitrate. The values on which the correlations are based are column means, 200 m
in the case of zooplankton and nutrients, and 150 m in the case of chlorophyll a and
productivity. The nutrients column was taken to 200 m rather than to the bottom of
the mixed layer, because the NIP profile showed a progressive increase from the surface
to round about this depth (Fig. 3).
In calculating correlation coefficients, values for zooplankton (species, taxa, and
biomass) and for the various midwater trawl components were transformed by logs.
D. J. TRANTER:
v. Trophic Relationships
1. Evidence of Trophic Succession
If the seasonal variation ob$erved in nearly every biological property is the result of a
seasonal cycle of plankton production, then the variations in these properties ought to be
inter-related.
In its simplest form, the plankton cycle is conceived to be a "food chain" with algae
as the first link, algal herbivores (zooplankton) as the second, primary carnivores
(zooplankton) as the third, secondary carnivores (micronekton) as the fourth, pelagic
squid as the fifth, and predatory nekton as the sixth. Each trophic level is considered,
ideally, to draw on the energy resources of the level immediately below. So, matter is
conveyed along the food chain leaving depletion in its wake. Each standing stock increases
in succession, reaches a maximum, and then declines. The problem is to distinguish
changes such as these from those of horizontal translation.
Where peaks at adjacent trophic levels are close in phase, their values will be wellcorrelated (STEEMANN NIELSEN, 1957; CUSHING, 1958). Where the time lag is considerable, there will be high negative correlations - for example when grazing or predation
pressure is increasing or declining. With such a model of plankton production in mind,
I have sought to determine whether or not there is evidence of trophic succession in our
study area during the year of study.
It has often been proposed that time lags are short in warm waters (e.g. CUSHING,
1958). Therefore, in the analysis that follows of data from meridian 110 0 E, correlations
based on concurrent pairs of values are considered first of all.
Table 5 shows correlation values between various components of the zooplankton
catch, representing holoplankton, meroplankton, and each major trophic type. The table
shows that these diverse components are well correlated; they vary in a similar way.
Of a total of 112 correlations, no less than 102 were significant at the 95% level, of
which not more than 5 - 6 could be due to chance. Most of these (86) were significant
also at the 99% level. Of the 26 correlations that were not significant at this level, more
than half were for the tropical zone by night; of the 10 that were not significant even at
the 95% level, 5 involved the rather unique component fish eggs and larvae, 2 involved
decapod larvae, 2 Chaetognatha, and 1 Copepoda. The various components of the
midwater trawl catch are similarly well-correlated, particularly in the subtropical zone
(Table 6 b).
It is clear, then, that the time lag between the seasonal cycles of the various zooplankton components is very short. Consequently it is not possible to locate evidence
of trophic succession. It is necessary to extend the search more widely.
Table 6 shows the correlations between zooplankton (as a whole) and other biological
properties, viz. phytoplankton (chlorophyll a), productivity (light-saturated), phosphate,
and nitrate. The values on which the correlations are based are column means, 200 m
in the case of zooplankton and nutrients, and 150 m in the case of chlorophyll a and
productivity. The nutrients column was taken to 200 m rather than to the bottom of
the mixed layer, because the NIP profile showed a progressive increase from the surface
to round about this depth (Fig. 3).
In calculating correlation coefficients, values for zooplankton (species, taxa, and
biomass) and for the various midwater trawl components were transformed by logs.