28
J. A. GULLAND AND J. E. CARROZ
for trawls, selection takes place over a range of sizes. If it is desired to
take fish of a certain size and bigger, the mesh size can be chosen t o
retain half of the fish of the critical size, and release half of them, but
some fish considerably smaller will be retained, and others considerably
larger than the critical size will escape. If the fisherman sees some of
these larger fish escape while he is hauling his net, he will become less
willing to use larger meshes, especially as a fish in the water looks
larger than one on deck. Because of the range of sizes over which
selection takes place, and also because the selection apparently varies
with the condition of the fish, the exact material and treatment of the
twine in the net (especially the cod-end of the trawl), the magnitude of
the catch, duration of haul, etc., the measurement of selection raises
many practical problems (Gulland, 1964b). A considerable volume of
literature has grown up concerned with the subject (see especially
the report of the Joint I .C.N. A.F. /I. C .E . S . /F. A. 0. Special Scientific
meeting (I.C.N.A.F., 1963) which contains papers not solely concerned
with selection in trawls, and the reports of the various I.C.E.S. working
groups (I.C.E.S., 1964, 1965)). For selectivity experiments outside the
North Atlantic, see Aoyama (1961), Cassie (1955) and Longhurst
(1959, 1960).
A more serious limitation to mesh size and similar regulations is that
many of the fisheries to which they are applied exploit several species,
which may be caught in the same haul. Especially when different
groups of fishermen prefer different species, i t is impracticable to
introduce a mesh size sufficiently large to provide effective regulation
for the stocks of the larger species without causing unacceptable
reductions in the catches of the smaller species. For example, in the
North Sea the biggest catches of bottom-living species are of cod,
haddock and plaice, all of which, and the trawl catch as a whole, would
be increased by the use of meshes of 100 mm and over-probably for
cod and plaice up to a t least 150 mm. However, it has proved impossible to introduce meshes larger than 80 mm because they would cause
losses of the smaller species (especially sole (Xolea solea L.) and whiting
(MerZangus rnerlangus (L.)), and these are of major importance to some
groups of fishermen. I n any case, even the mesh size (ca 110 mm)
which would give the greatest total yield of all species would give
considerably less than the sum of the yields from individual species,
if the optimum mesh was used for each: ca 80 mni for soles and ca
150 mm for cod, etc. (I.C.E.S., 1957).
The most fundamental disadvantage of mesh regulation as the only
method of management is that its very success tends to cause changes
which lose much of the benefits. The improved catches attract new
J. A. GULLAND AND J. E. CARROZ
for trawls, selection takes place over a range of sizes. If it is desired to
take fish of a certain size and bigger, the mesh size can be chosen t o
retain half of the fish of the critical size, and release half of them, but
some fish considerably smaller will be retained, and others considerably
larger than the critical size will escape. If the fisherman sees some of
these larger fish escape while he is hauling his net, he will become less
willing to use larger meshes, especially as a fish in the water looks
larger than one on deck. Because of the range of sizes over which
selection takes place, and also because the selection apparently varies
with the condition of the fish, the exact material and treatment of the
twine in the net (especially the cod-end of the trawl), the magnitude of
the catch, duration of haul, etc., the measurement of selection raises
many practical problems (Gulland, 1964b). A considerable volume of
literature has grown up concerned with the subject (see especially
the report of the Joint I .C.N. A.F. /I. C .E . S . /F. A. 0. Special Scientific
meeting (I.C.N.A.F., 1963) which contains papers not solely concerned
with selection in trawls, and the reports of the various I.C.E.S. working
groups (I.C.E.S., 1964, 1965)). For selectivity experiments outside the
North Atlantic, see Aoyama (1961), Cassie (1955) and Longhurst
(1959, 1960).
A more serious limitation to mesh size and similar regulations is that
many of the fisheries to which they are applied exploit several species,
which may be caught in the same haul. Especially when different
groups of fishermen prefer different species, i t is impracticable to
introduce a mesh size sufficiently large to provide effective regulation
for the stocks of the larger species without causing unacceptable
reductions in the catches of the smaller species. For example, in the
North Sea the biggest catches of bottom-living species are of cod,
haddock and plaice, all of which, and the trawl catch as a whole, would
be increased by the use of meshes of 100 mm and over-probably for
cod and plaice up to a t least 150 mm. However, it has proved impossible to introduce meshes larger than 80 mm because they would cause
losses of the smaller species (especially sole (Xolea solea L.) and whiting
(MerZangus rnerlangus (L.)), and these are of major importance to some
groups of fishermen. I n any case, even the mesh size (ca 110 mm)
which would give the greatest total yield of all species would give
considerably less than the sum of the yields from individual species,
if the optimum mesh was used for each: ca 80 mni for soles and ca
150 mm for cod, etc. (I.C.E.S., 1957).
The most fundamental disadvantage of mesh regulation as the only
method of management is that its very success tends to cause changes
which lose much of the benefits. The improved catches attract new