Jay is an ecological philosopher. He wants to sketch how ecologists have used boundaries, and outline both a skepticism and an interactive approach.
He’s not talking about types of ecosystems but tokens; not biomes, for example. Second, some ecosystems are sociopolitical objects (Greater Yellowstone).
In 1935, A. G. Tansley distinguished ecosystems – the abiotic and biotic resources – and rejected communities – set of interacting species. Ecosystem ecology focused on the flow of nutrients and energy through organisms and their environments. Organisms are transducers of energy and nutrients. Gross primary production (autotrophy?), minus respiration, consumption of autotrophs.
Cycling of materials …
Ecosystem is a spatially explicit unit of the Earth along with all its components. But this presupposes boundaries. How would you figure it out? Dale Jamieson says that talking about ecosystems is like talking about constellations – the world doesn’t respond to terms.
The (n + 1)th problem – suppose a token ecosystem. Why not extend it n + 1? Could be arbitrary, but there’s another way. Interactive boundaries: sorted by causal relationships. E.g., “feeds on”. Boundary set by such a relation will exclude any species not fed upon.
There is an ecosystem when
- There is an ecosystemic causal relation between a set of biotic and abiotic objects. Won’t work…
- Try, when above, but the causal relations differ in their magntudes.
Specific case: watersheds. Drainage basin of water and sediments. There are geomorpholic boundaries (ridges) that structure the processes. The nutrients and energetic flows have differential rates inside and outside the boundaries. They are multiscalar – sub watersheds, and sub-sub-…
Been used for a long time. Constrains both observation and experimental perturbation.