I have a particular non-standard view of ecology based on my studies of General Schemas Theory. In that theory it is noted that the System schema has an inverse dual called the meta-system (openscape). ‘Meta’ here means what is beyond or outside the system. The meta-system tries to give a term to concepts like ecology, environment, situation, context, media and other like terms that have no overarching concept to unify them in a single category in English. It turns out that the best term in English is the word Scape, but that always takes a prefix so we also call it an Open-Scape.
The point is that in the West the Openscape schema has been obscured, and this I think is the root cause of our insensitivity to environmental issues. We always think about Systems, but we do not think about the inverse dual of the System which is the Meta-system, which corresponds to the environment and its ecology. If we understood Schemas Theory better then we would have a better grasp of the nature of the environment and ecology because we would see them not as systems but as openscapes. The openscape is what is seen from a given point to the horizon in a landscape that is outside the boundary of the system. Thus systems have determinate boundaries and openscapes have indeterminate horizons. Both Systems and Meta-systems can be recursively applied to each other, so you can have nested sub-systems, systems, and supersystems (system of systems), and the same goes for the meta-system which also can be applied to itself. But what is not generally recognized is that Systems and Meta-systems are interleaved and mutually nested so that they are like Russian dolls, in that the doll is the system and the space between the dolls into which the smaller doll fits is analogous to the meta-system.
Another analogy is the turing machine and the universal turing machine. A universal turing machine is one that runs other turing machines and we call these “operating systems”. Thus the system is like the application running on the “operating system” which is like the meta-system. The halting problem does not apply to the universal turing machine because it is never meant to stop.
The System/Meta-system relation is like that proposed by G. Bataille in Accursed Share between the Restricted and General Economies. A meta-system is a general economy. Arkady points this out in his book Complementarities.
The point is that ecology and the biology of the natural environment is a discipline that needs the theory of the meta-system, and I think that there are some ecologists that have such a view, yet most are still thinking in terms of systems rather than the next higher scoped schema, the openscape.
So my answer is that Ecology is the application of the openscape or meta-system schema to the environment, such that it understands that it is an environment for systems which is differently organized than the system. I have given details of this in my dissertation (see http://emergentdesign.net). In other words Ecology seeks to understand the nature of the environment as a network of interacting organisms and the geology that serves as their cohabitatational foundation. Species fit into niches in the environment which is broken up into patches where different colonies of species cluster together. These species and their niches are covalent, i.e. they are sculpted and fine tuned to fit each other with the overall ecological meta-system. In an ecology, it is the network of different species interactions that may form the system, rather than the organism itself. An organism may be seen as a form within a resilient system of mutual relations between cohabiting species.