Demography

Emergent Population Ecology in a Virtual Ecosystem

Virtual Ecology provides all the kinds of information obtained from classical population ecology. The demographic properties of a population of plankton in the mesocosm are emergent properties of the virtual ecosystem. They depend on the functions of individual plankters. It is no longer necessary to depend on population-based equations.

The demography of a species describes the number of plankton in its population and the rate of change due to reproduction, natural mortality, predation and migration.

Species

You can design a virtual ecosystem to have a simple or complex community with as many species of plankton as you like. They can be phytoplankton, zooplankton, fi sh larvae, bacteria and viruses. But they cannot be fi sh, which can usefully change their ambient environment by migrating horizontally. Fish migration cannot be tracked in a one-dimensional virtual ecosystem. But, more importantly, fi sh can learn new behaviour and arbitrarily adopt a behaviour pattern according to new circumstance. We do not yet know how to establish primitive equations for learning and choice. Some fi sheries modellers seek to overcome that diffi culty by using the method of induction based on past migration patterns. Induction can never be used in virtual ecology because it denies emergent behaviour.

Particles

Every species is represented by a cloud of particles. Each particle behaves like a single organism following a unique trajectory. The plankton in each particle can be alive or dead. Other particles represent fæcal pellets. Bacteria remineralize the chemicals in dead plankton and fæcal pellets.

Lagrangian Ensemble Integration

Lagrangian Ensemble integration (Woods & Onken 1982) is the precursor of the "super particle" concept in terrestrial population ecology (Reference 1996). It is a method for scaling-up from the number of agents in the computation to the number of planktors in the ecosystem. The uncertainty (demographic noise) is smaller when more particles are used. In practice the number is limited to about one million per processor in your computer. So a community of, say, ten species can be represented by 100,000 particles per population. That limits the demographic noise to a few percent (see "The Global Stability of a Virtual Ecosystem").

Sub-populations

A particle that represents a living plankter carries a packet of information about a dynamic sub-population of identical plankters. The number of plankton in a sub-population is continually changing due to reproduction, starvation, disease and predation.

Populations

A population in the virtual mesocosm comprises all the individual plankters of that species. It equals the sum of the plankters in every sub-population of that species in the mesocosm.

Computing demography

The population of one species is the number of living organisms in the virtual mesocosm.The population is computed by summing over all its sub-populations, each of which is associated with one of the thousands of particles used to represent that species. Similarly, the biomass in that population equals the sum of the biomasses in all the sub-populations of the species.

Demographic processes

The population is continually changing due to a combination of several processes: reproduction, natural mortality, predation and migration. The contribution of each process is computed by summing the corresponding process in all the sub-populations of that species.

Profiles

The sums can performed separately in each (one-metre thick) layer of the virtual mesocosm to produce a vertical profi le of each demographic property.