82
GORDON A. RILEY
should vary so precisely with changes in viscosity, is questionable.
Riley (1965) suggested alternatively that this apparent sinking rate
might actually be due to convective overturn, a concept which has
been developed more fully in the present discussion.
If this concept is correct, regional variations in apparent sinking
rate would be related to thermal structure of the water rather than
to differences in viscosity. I n general, the seasonal thermocline tends
to be somewhat deeper in tropical and subtropical waters than in
higher latitudes. Many local exceptions can be found, such as the shallow thermoclines associated with upwelling. However, when the thermal
structure develops as it simple function of wind stress, the depth of the
mixed layer for any given wind speed theoretically should be inversely
proportional to the sine of the latitude (Rossby and Montgomery,
1935). This kind of variation was implicit, though not explicitly stated,
in the model of regional variations of phytoplankton (Riley, 1965).
The assumption of a regional variation in sinking speed could have
been stated alternatively as an advective removal of a certain fixed
proportion of the population in the mixed layer each day, and the
results would have been essentially the same. I n order t o get numerically realistic results one would have to assume that about 5-7% of
the mixed layer loses its population each day.
Convective motion obviously deserves more thorough study, both
as a physical process and in respect t o its biological implications. The
hypotheses proposed here remain tentative, but they provide possible
explanations for problems that eluded solution for a long time. There
are further implications with regard to non-living particulate matter.
Particles which have the same spectrum of sinking rates as phytoplankton should have approximately the same area of retention and the same
rate of loss from the mixed layer, and this rate of loss could be several
times larger than observed sinking rates. Of course this must be
qualified by saying that there are times when negative gradients in
the surface layer are too strong to permit convective overturn, and at
such times both phytoplankton and non-living matter will be removed
at their natural sinking rate.
A correlation has been noted between the quantity of phytoplankton
and non-living matter. Moreover, flakes and aggregates are of a size
that should make them susceptible t o removal by indiscriminate filter
feeders to about the same degree as phytoplankton. These similarities
in processes of removal suggest that rates of production also are
probably more or less similar. I n general this means that the daily
increase in particles of 5 p or larger could be about 5-15% of the amount
present.
GORDON A. RILEY
should vary so precisely with changes in viscosity, is questionable.
Riley (1965) suggested alternatively that this apparent sinking rate
might actually be due to convective overturn, a concept which has
been developed more fully in the present discussion.
If this concept is correct, regional variations in apparent sinking
rate would be related to thermal structure of the water rather than
to differences in viscosity. I n general, the seasonal thermocline tends
to be somewhat deeper in tropical and subtropical waters than in
higher latitudes. Many local exceptions can be found, such as the shallow thermoclines associated with upwelling. However, when the thermal
structure develops as it simple function of wind stress, the depth of the
mixed layer for any given wind speed theoretically should be inversely
proportional to the sine of the latitude (Rossby and Montgomery,
1935). This kind of variation was implicit, though not explicitly stated,
in the model of regional variations of phytoplankton (Riley, 1965).
The assumption of a regional variation in sinking speed could have
been stated alternatively as an advective removal of a certain fixed
proportion of the population in the mixed layer each day, and the
results would have been essentially the same. I n order t o get numerically realistic results one would have to assume that about 5-7% of
the mixed layer loses its population each day.
Convective motion obviously deserves more thorough study, both
as a physical process and in respect t o its biological implications. The
hypotheses proposed here remain tentative, but they provide possible
explanations for problems that eluded solution for a long time. There
are further implications with regard to non-living particulate matter.
Particles which have the same spectrum of sinking rates as phytoplankton should have approximately the same area of retention and the same
rate of loss from the mixed layer, and this rate of loss could be several
times larger than observed sinking rates. Of course this must be
qualified by saying that there are times when negative gradients in
the surface layer are too strong to permit convective overturn, and at
such times both phytoplankton and non-living matter will be removed
at their natural sinking rate.
A correlation has been noted between the quantity of phytoplankton
and non-living matter. Moreover, flakes and aggregates are of a size
that should make them susceptible t o removal by indiscriminate filter
feeders to about the same degree as phytoplankton. These similarities
in processes of removal suggest that rates of production also are
probably more or less similar. I n general this means that the daily
increase in particles of 5 p or larger could be about 5-15% of the amount
present.