70
Martin V. ANGEL
Fig. 3.12. The distribution of the biogeochemical subdomains based on the productivity cycles illustrated in Fig. 3.11 and the approximate
climatological positions of major oceanographic discontinuities identified from a combination of remotely-sensed and hydrographic data.
These correlate closely with classical biogeographical provinces identified in each ocean, suggesting that these large-scale distribution patterns
are determined by bottom-up processes. There are indications that these domains are also mirrored in the deep benthic communities. Redrawn
from Longhurst (1995).
to be capable of diel vertical migration are still limited
to as little as 10–15% of the total ocean volume.
Herbivores are absent from most oceanic volume, and
so detritivory is the basis for food chains in most deepocean scenarios. Thus, unlike the Eltonian concept,
Platt et al. (1981) found that oceanic food webs can
be modelled more closely on the basis of size spectra
than on functional relationships.
These characteristics of ocean food webs also mean
that there are few opportunities for specialization, and
this may account for the low global species richness
of open ocean fishes; certainly relative to freshwater
species. A single river system in Africa, the Zaire, is
inhabited by >690 species, 84% of which are endemic.
In Lake Malawi there are >600 fish species of which
96% are endemic, and 92.5% belong to a single
family, the cichlids (Ribbink, 1994). Together these two
freshwater systems host nearly as many fish species as
are found throughout the whole of the North Atlantic.
PATTERNS OF PRODUCTIVITY AND
BIOGEOGRAPHY
In the open ocean the annual quantity and seasonal
cycling of primary production is determined by vertical
stratification, the light cycle and the persistence of
nutrient supplies. Longhurst (1995) has used remotely
sensed climatological data for chlorophyll (derived
from the Coastal Zone Color Scanner) and sea surface
temperature data (from the satellite Nimbus 7) in
combination with climatological data for mixed-layer
depths and nutrient concentrations (Levitus et al., 1993)
to identify three basic production domains in the
open ocean – polar, temperate and tropical. Each of
these domains has a fundamentally different seasonal
cycle of water column stability, nutrient supply and
illumination. Coastal (shelf) waters provide a fourth
domain, which is subdivided into a mosaic of very
much smaller-scale regions. Longhurst also identified
eight basic types of production cycle (Fig. 3.11),
which sub-divides the three basic domains into 56 biogeographical provinces. The provinces are bounded
by recurrent physical features – fronts bounding
ocean currents, topographic features and sea-surface
chlorophyll distributions (Fig. 3.12). Many of the
boundaries to Longhurst’s provinces coincide with the
biogeographical patterns described in classical studies
of biodiversity and species distributions (e.g., Backus,
1986); the inference is that the zoogeographical patterns are probably determined by the production cycles
and hence the prevailing biogeochemical environmental
conditions (e.g., Angel, 1993).
Within each province the structure of pelagic food
webs and communities appears to be relatively consistent. The quantities and dynamics of export production
(i.e., the amounts of organic carbon that are exported
from the euphotic zone into the deep ocean) are
also likely to be directly influenced by the production
Martin V. ANGEL
Fig. 3.12. The distribution of the biogeochemical subdomains based on the productivity cycles illustrated in Fig. 3.11 and the approximate
climatological positions of major oceanographic discontinuities identified from a combination of remotely-sensed and hydrographic data.
These correlate closely with classical biogeographical provinces identified in each ocean, suggesting that these large-scale distribution patterns
are determined by bottom-up processes. There are indications that these domains are also mirrored in the deep benthic communities. Redrawn
from Longhurst (1995).
to be capable of diel vertical migration are still limited
to as little as 10–15% of the total ocean volume.
Herbivores are absent from most oceanic volume, and
so detritivory is the basis for food chains in most deepocean scenarios. Thus, unlike the Eltonian concept,
Platt et al. (1981) found that oceanic food webs can
be modelled more closely on the basis of size spectra
than on functional relationships.
These characteristics of ocean food webs also mean
that there are few opportunities for specialization, and
this may account for the low global species richness
of open ocean fishes; certainly relative to freshwater
species. A single river system in Africa, the Zaire, is
inhabited by >690 species, 84% of which are endemic.
In Lake Malawi there are >600 fish species of which
96% are endemic, and 92.5% belong to a single
family, the cichlids (Ribbink, 1994). Together these two
freshwater systems host nearly as many fish species as
are found throughout the whole of the North Atlantic.
PATTERNS OF PRODUCTIVITY AND
BIOGEOGRAPHY
In the open ocean the annual quantity and seasonal
cycling of primary production is determined by vertical
stratification, the light cycle and the persistence of
nutrient supplies. Longhurst (1995) has used remotely
sensed climatological data for chlorophyll (derived
from the Coastal Zone Color Scanner) and sea surface
temperature data (from the satellite Nimbus 7) in
combination with climatological data for mixed-layer
depths and nutrient concentrations (Levitus et al., 1993)
to identify three basic production domains in the
open ocean – polar, temperate and tropical. Each of
these domains has a fundamentally different seasonal
cycle of water column stability, nutrient supply and
illumination. Coastal (shelf) waters provide a fourth
domain, which is subdivided into a mosaic of very
much smaller-scale regions. Longhurst also identified
eight basic types of production cycle (Fig. 3.11),
which sub-divides the three basic domains into 56 biogeographical provinces. The provinces are bounded
by recurrent physical features – fronts bounding
ocean currents, topographic features and sea-surface
chlorophyll distributions (Fig. 3.12). Many of the
boundaries to Longhurst’s provinces coincide with the
biogeographical patterns described in classical studies
of biodiversity and species distributions (e.g., Backus,
1986); the inference is that the zoogeographical patterns are probably determined by the production cycles
and hence the prevailing biogeochemical environmental
conditions (e.g., Angel, 1993).
Within each province the structure of pelagic food
webs and communities appears to be relatively consistent. The quantities and dynamics of export production
(i.e., the amounts of organic carbon that are exported
from the euphotic zone into the deep ocean) are
also likely to be directly influenced by the production