T H E PRODUCTION O F MARINE PLANKTON
169
not catch the very smallest protozoans quantitatively, and these must
be sampled by other methods. Recently, high-speed samplers such as
the Isaacs Kidd mid-water trawl have been developed to sample large
volumes of water and to obtain samples over considerable areas. While
such techniques are of great advantage in estimating the distribution of
the larger forms, alone they cannot give an accurate quantitative
sampling of the whole zooplankton.
Another problem is the clogging of the meshes. In a rich phytoplankton bloom the filtration power of a net may be so reduced that it
prevents quantitative sampling. Equally, certain of the more gelatinous
zooplankton such as salps, medusae, ctenophores and siphonophores,
when abundant, may clog the meshes of the net and reduce filtration.
Some salps, jelly-fish and ctenophores show a marked tendency at
times to intensive swarming. Not only will the net tend to be clogged in
passing through such populations, but any quantitative sampling is
always made more difficult if the zooplankton has marked patchy
distribution. Even apart from salps and ctenophores, patchiness is an
extremely common feature of zooplankton distribution.
Various types of meter have been developed to measure the actual
volume of water passing through nets. Usually they are placed in the
mouth of the net. While such devices increase the accuracy of volume
measurements, their presence may reduce catching power, and some
animals may actively avoid the mouth of the net.
The whole problem of the proper sampling of the zooplankton is
dependent on precise knowledge of vertical distribution. Unfortunately
zooplankton varies enormously in its vert'ical distribution; often there
is a rich layer close to the surface and another maximum at a depth
approaching a thousand metres. The work of Leavitt (1935, 1938),
Jespersen (1935), Sewell (1948), Zenkevitc h and Birstein (1956),
Vinogradov (1962a) and Hardy and Gunther (1936), all deal with the
problem of the vertical distribution of animal plankton. Though there
is fairly general agreement that very deep wa,ter has reduced populations, there appears to be no clear pattern of general distribution. In
Arctic as well as Antarctic seas (cf. Hardy and Gunther, 1936; Zenkevitch, 1963; Johnson, 1963) there is a marked tendency for the greater
biomass of zooplankton to be beneath the surface layers but the distribution varies. Moreover, the zooplankton is characteristically marked
by vertical migration. This may occur diurnally, and more especially
in colder waters, there may be marked seasonal migration. The problem
of correct quantitative assessment of the richness of various strata thus
becomes more difficult.
Apart from the problem of catching plankton quantitatively there is
the question of estimating the catch. The direct numerical count is
169
not catch the very smallest protozoans quantitatively, and these must
be sampled by other methods. Recently, high-speed samplers such as
the Isaacs Kidd mid-water trawl have been developed to sample large
volumes of water and to obtain samples over considerable areas. While
such techniques are of great advantage in estimating the distribution of
the larger forms, alone they cannot give an accurate quantitative
sampling of the whole zooplankton.
Another problem is the clogging of the meshes. In a rich phytoplankton bloom the filtration power of a net may be so reduced that it
prevents quantitative sampling. Equally, certain of the more gelatinous
zooplankton such as salps, medusae, ctenophores and siphonophores,
when abundant, may clog the meshes of the net and reduce filtration.
Some salps, jelly-fish and ctenophores show a marked tendency at
times to intensive swarming. Not only will the net tend to be clogged in
passing through such populations, but any quantitative sampling is
always made more difficult if the zooplankton has marked patchy
distribution. Even apart from salps and ctenophores, patchiness is an
extremely common feature of zooplankton distribution.
Various types of meter have been developed to measure the actual
volume of water passing through nets. Usually they are placed in the
mouth of the net. While such devices increase the accuracy of volume
measurements, their presence may reduce catching power, and some
animals may actively avoid the mouth of the net.
The whole problem of the proper sampling of the zooplankton is
dependent on precise knowledge of vertical distribution. Unfortunately
zooplankton varies enormously in its vert'ical distribution; often there
is a rich layer close to the surface and another maximum at a depth
approaching a thousand metres. The work of Leavitt (1935, 1938),
Jespersen (1935), Sewell (1948), Zenkevitc h and Birstein (1956),
Vinogradov (1962a) and Hardy and Gunther (1936), all deal with the
problem of the vertical distribution of animal plankton. Though there
is fairly general agreement that very deep wa,ter has reduced populations, there appears to be no clear pattern of general distribution. In
Arctic as well as Antarctic seas (cf. Hardy and Gunther, 1936; Zenkevitch, 1963; Johnson, 1963) there is a marked tendency for the greater
biomass of zooplankton to be beneath the surface layers but the distribution varies. Moreover, the zooplankton is characteristically marked
by vertical migration. This may occur diurnally, and more especially
in colder waters, there may be marked seasonal migration. The problem
of correct quantitative assessment of the richness of various strata thus
becomes more difficult.
Apart from the problem of catching plankton quantitatively there is
the question of estimating the catch. The direct numerical count is