Chapter 7
Provinces: The Secondary
Compartments
A
partition into four biomes clearly does not fully satisfy the requirement for a
regional geography of the ocean because it fails to capture the fine detail observed
in satellite images of the surface chlorophyll field, and so fails to reflect the regional
structure of the pelagic ecosystem. Equally obviously, the two polar regions function
differently, though following similar general rules, and the midlatitude and tropical
regions of the Pacific, Atlantic, and Indian Oceans each have individual characteristics
that we wish to understand. And each semienclosed sea and shelf region differs from
every other in its characteristic ecology.
The primary biomes of the pelagic realm were characterized rather simply in Chapter 6,
but if we are interested in a smaller-scale partition of the ocean, then these four biomes
must each be rationally subdivided. To do this, we must have recourse to a wider set of
factors, especially those apt to define interfaces between physically, and therefore ecologically, distinct regions. We should examine the factors that determine the characteristics
of regional circulation and stratification at all scales, of which the resistance of the
mixed layer to deepening (the primary determinant of biomes) is only a single example.
Bathymetry, river discharges, characteristic coastal wind systems, location of islands, and
the distribution of land masses must all be reviewed. The procedures adopted to perform
this finer partition are discussed in this chapter.
It is important to accept, as I have reiterated at several points in this book, that
establishing a partition of the ocean’s surface is a fractal problem, because physical
phenomena occur across such a wide range of dimension. Though some phenomena
have a characteristic scale, many others that are critical to this task do not. An example
that is frequently used to illustrate the fractal concept is vorticity in the atmosphere,
which occurs across a wide range of horizontal dimension: dust devils (10 m), tornadoes
(1000 m), cyclonic storms (100 km), and weather systems (1000 km). Similarly, vorticity
in ocean circulation exists from small swirls, to mesoscale eddies, and up to entire ocean
gyres. Perturbation of flow, and fronts between opposing flow, in water bodies may be
observed alongside a jetty or across an ocean basin.
We cannot expect, therefore, that an objective review of the processes that are likely
to be relevant to pelagic ecology shall constrain us to converge on an objective number
of compartments. Rather, it will be better to pursue our investigations with a subjective
ideal in mind. It has already been suggested that about 50 compartments globally would
be convenient for the computation of global primary production (Sathyendranath et al.,
1995), a suggestion that has been allowed to guide these proposals for a global partition
of the pelagic ecosystem. This is a matter not only of practical convenience but also
of necessity because we are constrained by the available number of observations of any
ecological variable or parameter. Where too few observations are binned among too many
103
Provinces: The Secondary
Compartments
A
partition into four biomes clearly does not fully satisfy the requirement for a
regional geography of the ocean because it fails to capture the fine detail observed
in satellite images of the surface chlorophyll field, and so fails to reflect the regional
structure of the pelagic ecosystem. Equally obviously, the two polar regions function
differently, though following similar general rules, and the midlatitude and tropical
regions of the Pacific, Atlantic, and Indian Oceans each have individual characteristics
that we wish to understand. And each semienclosed sea and shelf region differs from
every other in its characteristic ecology.
The primary biomes of the pelagic realm were characterized rather simply in Chapter 6,
but if we are interested in a smaller-scale partition of the ocean, then these four biomes
must each be rationally subdivided. To do this, we must have recourse to a wider set of
factors, especially those apt to define interfaces between physically, and therefore ecologically, distinct regions. We should examine the factors that determine the characteristics
of regional circulation and stratification at all scales, of which the resistance of the
mixed layer to deepening (the primary determinant of biomes) is only a single example.
Bathymetry, river discharges, characteristic coastal wind systems, location of islands, and
the distribution of land masses must all be reviewed. The procedures adopted to perform
this finer partition are discussed in this chapter.
It is important to accept, as I have reiterated at several points in this book, that
establishing a partition of the ocean’s surface is a fractal problem, because physical
phenomena occur across such a wide range of dimension. Though some phenomena
have a characteristic scale, many others that are critical to this task do not. An example
that is frequently used to illustrate the fractal concept is vorticity in the atmosphere,
which occurs across a wide range of horizontal dimension: dust devils (10 m), tornadoes
(1000 m), cyclonic storms (100 km), and weather systems (1000 km). Similarly, vorticity
in ocean circulation exists from small swirls, to mesoscale eddies, and up to entire ocean
gyres. Perturbation of flow, and fronts between opposing flow, in water bodies may be
observed alongside a jetty or across an ocean basin.
We cannot expect, therefore, that an objective review of the processes that are likely
to be relevant to pelagic ecology shall constrain us to converge on an objective number
of compartments. Rather, it will be better to pursue our investigations with a subjective
ideal in mind. It has already been suggested that about 50 compartments globally would
be convenient for the computation of global primary production (Sathyendranath et al.,
1995), a suggestion that has been allowed to guide these proposals for a global partition
of the pelagic ecosystem. This is a matter not only of practical convenience but also
of necessity because we are constrained by the available number of observations of any
ecological variable or parameter. Where too few observations are binned among too many
103