260
N. A. HOLME
different st'ations to show up a difference of 2% between the two
stations ( P = 0.01). Even this method, however, suggests much
larger numbers of samples than can normally be taken at sea.
Application of statistics
When making any survey a decision must be made as to the area to
be sampled at each station, and some of the features of distribution
outlined above may be of some assistance. Other factors to be considered will be dependent on local conditions and the purpose for
which the samples are required. There remains a considerable element
of personal choice, for one worker will take samples at many stations,
with one haul at each station, while another will prefer to take samples
totalling perhaps a square metre at a lesser number of stations. Either
method is probably equally valid, but it would be advisable initially
to undertake some test series of samples at a few representative
stations to determine what degree of reliance could be placed on
individual samples.
One further cause of variation has not so far been considered-the
effect of seasonal changes in the fauna. It is well-known that year-toyear changes in the stock of invertebrates occur (Thorson, 1957b),
but the effects of seasonal changes may be overlooked. These are
caused by migration, mortality, and the settlement and subsequent
growth of spat up to a size at which they may be separated by sieving.
Seasonal migrations of crustacea and certain other members of
the epifauna are known to occur and such changes will be most marked
in shallow water where temperatures are higher in summer and lower
in winter than in deeper water. Similarly, seasonal mortalities will be
greater in shallow waters where near-zero temperatures may be
approached in winter.
There is not much evidence in the literature that seasonal changes
greatly affect grab sampling offshore. Birkett (1953, 1959) has traced
the growth and mortality of Mactra spat on the Dogger Bank, and
such changes must be very significant in areas of high spat-fall. Most
workers on offshore grounds have not noted any significant seasonal
changes (Steven, 1930; Jones, 1962, 1956; Holme, 1063), although
Steven did notice the absence of the amphipod Ampetisca in samples
taken during the winter.
Information on the length of life of different benthic animals is
given in Thorson (1967b). Those species which are annuals will
naturally show greater seasonal variation than the majority of benthic
animals which live 3, 4, or more years. In general it may be said that
faunistic surveys offshore may be made without too much concern
N. A. HOLME
different st'ations to show up a difference of 2% between the two
stations ( P = 0.01). Even this method, however, suggests much
larger numbers of samples than can normally be taken at sea.
Application of statistics
When making any survey a decision must be made as to the area to
be sampled at each station, and some of the features of distribution
outlined above may be of some assistance. Other factors to be considered will be dependent on local conditions and the purpose for
which the samples are required. There remains a considerable element
of personal choice, for one worker will take samples at many stations,
with one haul at each station, while another will prefer to take samples
totalling perhaps a square metre at a lesser number of stations. Either
method is probably equally valid, but it would be advisable initially
to undertake some test series of samples at a few representative
stations to determine what degree of reliance could be placed on
individual samples.
One further cause of variation has not so far been considered-the
effect of seasonal changes in the fauna. It is well-known that year-toyear changes in the stock of invertebrates occur (Thorson, 1957b),
but the effects of seasonal changes may be overlooked. These are
caused by migration, mortality, and the settlement and subsequent
growth of spat up to a size at which they may be separated by sieving.
Seasonal migrations of crustacea and certain other members of
the epifauna are known to occur and such changes will be most marked
in shallow water where temperatures are higher in summer and lower
in winter than in deeper water. Similarly, seasonal mortalities will be
greater in shallow waters where near-zero temperatures may be
approached in winter.
There is not much evidence in the literature that seasonal changes
greatly affect grab sampling offshore. Birkett (1953, 1959) has traced
the growth and mortality of Mactra spat on the Dogger Bank, and
such changes must be very significant in areas of high spat-fall. Most
workers on offshore grounds have not noted any significant seasonal
changes (Steven, 1930; Jones, 1962, 1956; Holme, 1063), although
Steven did notice the absence of the amphipod Ampetisca in samples
taken during the winter.
Information on the length of life of different benthic animals is
given in Thorson (1967b). Those species which are annuals will
naturally show greater seasonal variation than the majority of benthic
animals which live 3, 4, or more years. In general it may be said that
faunistic surveys offshore may be made without too much concern