In fact, there are a large number of feedbacks in ecosystems of wide variety in strength and timescale. Of course, the model is just a replica of the concerned system, and it is impossible to include all feedbacks into a model. Therefore, to formulate the flows in an ecosystem as described before, we should understand what kinds of feedbacks are dominant in the targeted ecosystem, and discuss how to introduce these feedbacks into the model. It would not be an exaggeration to say that the modeling of an ecosystem structure is the finding out of the dominant feedback mechanisms in the ecosystem, and mathematical description of these dominant feedbacks in the model.
Actually, as J0rgensen pointed out, there are various kinds of feedbacks in the ecosystem, with widely different action mechanisms. Especially, the response timescale, which is one of the important points to describe the structural dynamics, varies greatly from case to case.
For example, in a short-term timescale, the growth of phytoplankton is controlled by the available nutrient concentrations dissolved in water and/or the intracellular nutrient concentration. When the concentration of a nutrient is far less relative to requirement (e.g., halfsaturation constant of the phytoplankton for the concerned nutrient), the growth rate of the phytoplankton is far below the maximum growth rate. On the other hand, if the concentration is enough (and, of course, if no other limitation, such as toxic substances, occurs), the phytoplankton grows according to the maximum growth rate. This is a foundational property of the phytoplankton growth, and most of the lake ecological models include these feedbacks. However, more long-term feedbacks -such as the biochemical adaptation, biological adaptation, species selection, selection of predator-prey relationship, etc. - are sometimes required to introduce the structural dynamic models. Furthermore, the emergence of new species may be required in a structural dynamic model which is used to assess the very long-term ecological succession.
The modeling strategy should be changed depending on the feedback timescales which are introduced into a model. Particularly, the long-term feedbacks tend to be important in the structural dynamic models.
Was this article helpful?