62
Chapter 4: Physical Control of Ecological Processes
Rule-Based Models of Ecological Response
to External Forcing
Those who originally coined the biosphere concept during the 19th century believed
that pattern within it was created principally by reaction between its component animals
and plants (Westbrook, 2002). But central to the argument of this book is the observation that the structure and functioning of both the marine and terrestrial parts of
the biosphere result principally from the effects of their external physical environment:
both terrestrial and marine ecologists have used very similar constructs to represent
how the external forcing of ecosystems selects dominance among plant assemblages.
Margalef (see later discussion) suggests that phytoplankton assemblages respond principally to turbulence and nutrient concentration, and he is remarkably close to the logic
of Holdridge (1947), or Atjay et al. (1979), who showed how the occurrence of vegetation types or biomes can be predicted from just a few indices: Holdridge requires only
annual precipitation, annual mean temperature, and the potential evapo-transpiration
ratio to map tropical plant biomes. To predict the distribution of terrestrial biomes, Atjay
et al. require information on only latitude, altitude, exposure, rainfall, and geological
substrate.
Rule-based modeling of this kind can be carried sufficiently far to reproduce the
observed pattern of natural terrestrial vegetation types, except (of course) where these
have been replaced by intensive agriculture. By coupling these routines to atmospheric
General Circulation Models the dynamic response of terrestrial ecosystems to changes in
the global climate may be predicted.
The same could be done at sea if we could identify a comparable set of factors
in marine ecology that control the growth of plant cells in the open ocean. To the
extent that this approach is successful, and yielded a partition defined by characteristic modes of phytoplankton ecology, it would also suggest which organisms of higher
trophic levels are likely to occur in each biome because phytoplankton-herbivorepredator relationships in characteristic pelagic habitats are well known at least in outline, differ significantly between habitats, and should be generally predictable for each
partition or biome defined by a unique pattern of environmental forcing of algal
growth. Here, biome is used in the sense of Odum (1971): “The largest community
unit that it is convenient to recognize. In a given biome the life form of the climatic climax vegetation is uniform. Thus, the climax vegetation of the grassland biome
is grass.” Here, it is assumed that knowledge of the characteristic seasonal phytoplankton cycle for any region is equivalent to a description of its “climax vegetation”
and that much can be inferred about characteristic regional ecology from just this
knowledge.
Although the principles involved should be common to both fields, such models have
been taken much farther in terrestrial than marine ecology, perhaps because of the relative
ease of observing and mapping terrestrial plant communities for verification. Production
of a sufficiently detailed field of the forcing factors required by the simulation process
is surely now possible, using the synoptic global fields of satellite data discussed in the
previous chapter. Marine ecologists can surely now hope to match the achievements of
their dry-land colleagues.
To do so, a useful starting point might be Margalef’s diagram, which has been
the basis of many fruitful discussions of the relations between physical processes and
pelagic ecosystems. Partly because it was drafted before the existence of autotrophic
microbiota was known, and emphasized the reaction of larger phytoplankton to their
environment, Cullen et al. (2002) revisited Margalef’s diagram that suggests that four types
of phytoplankton community are determined by differential interaction between nutrients
Chapter 4: Physical Control of Ecological Processes
Rule-Based Models of Ecological Response
to External Forcing
Those who originally coined the biosphere concept during the 19th century believed
that pattern within it was created principally by reaction between its component animals
and plants (Westbrook, 2002). But central to the argument of this book is the observation that the structure and functioning of both the marine and terrestrial parts of
the biosphere result principally from the effects of their external physical environment:
both terrestrial and marine ecologists have used very similar constructs to represent
how the external forcing of ecosystems selects dominance among plant assemblages.
Margalef (see later discussion) suggests that phytoplankton assemblages respond principally to turbulence and nutrient concentration, and he is remarkably close to the logic
of Holdridge (1947), or Atjay et al. (1979), who showed how the occurrence of vegetation types or biomes can be predicted from just a few indices: Holdridge requires only
annual precipitation, annual mean temperature, and the potential evapo-transpiration
ratio to map tropical plant biomes. To predict the distribution of terrestrial biomes, Atjay
et al. require information on only latitude, altitude, exposure, rainfall, and geological
substrate.
Rule-based modeling of this kind can be carried sufficiently far to reproduce the
observed pattern of natural terrestrial vegetation types, except (of course) where these
have been replaced by intensive agriculture. By coupling these routines to atmospheric
General Circulation Models the dynamic response of terrestrial ecosystems to changes in
the global climate may be predicted.
The same could be done at sea if we could identify a comparable set of factors
in marine ecology that control the growth of plant cells in the open ocean. To the
extent that this approach is successful, and yielded a partition defined by characteristic modes of phytoplankton ecology, it would also suggest which organisms of higher
trophic levels are likely to occur in each biome because phytoplankton-herbivorepredator relationships in characteristic pelagic habitats are well known at least in outline, differ significantly between habitats, and should be generally predictable for each
partition or biome defined by a unique pattern of environmental forcing of algal
growth. Here, biome is used in the sense of Odum (1971): “The largest community
unit that it is convenient to recognize. In a given biome the life form of the climatic climax vegetation is uniform. Thus, the climax vegetation of the grassland biome
is grass.” Here, it is assumed that knowledge of the characteristic seasonal phytoplankton cycle for any region is equivalent to a description of its “climax vegetation”
and that much can be inferred about characteristic regional ecology from just this
knowledge.
Although the principles involved should be common to both fields, such models have
been taken much farther in terrestrial than marine ecology, perhaps because of the relative
ease of observing and mapping terrestrial plant communities for verification. Production
of a sufficiently detailed field of the forcing factors required by the simulation process
is surely now possible, using the synoptic global fields of satellite data discussed in the
previous chapter. Marine ecologists can surely now hope to match the achievements of
their dry-land colleagues.
To do so, a useful starting point might be Margalef’s diagram, which has been
the basis of many fruitful discussions of the relations between physical processes and
pelagic ecosystems. Partly because it was drafted before the existence of autotrophic
microbiota was known, and emphasized the reaction of larger phytoplankton to their
environment, Cullen et al. (2002) revisited Margalef’s diagram that suggests that four types
of phytoplankton community are determined by differential interaction between nutrients