Based on the theory of robust coexistence by Meszéna et al. (TPB, 69:68, 2006), we establish a clear meaning of ecological niche assuming that the intended use of the word is the precise formulation of Gause's principle. In line with Gause’s principle the expected content of a ``niche theory'' can be vaguely summarized as:
-Species partition an abstract “niche space” of the ecological possibilities among themselves.
-They are able to coexist, i.e. to avoid competitive exclusion, exactly because they partition that space.
Our central notion is the regulating variable, a generalization of the resource concentration. An environmental variable is regulating, if it is involved in the population regulation feedback loop. The niche space is identified with the set of the regulating variables. This is the set, which should be partitioned between the coexisting populations to reduce competition. This “space” is not necessarily a continuous entity. In case of two distinct resources it is a two-element set. Hutchinson’s niche axes/variables represent continua of regulating variables. The concentrations of the infinitely many different seed-sizes are infinitely many different regulating variables, distinguished by the niche variable, i.e. the seed-size.
Regulating variables at different locations, distinguished by different environmental conditions, are to be considered as separate ones. Therefore, conditions, like temperature, can serve as niche variables. Successional niche-segregation in a metapopulation is a separation along the patch-age, as niche axis. Here the regulating variables are the frequencies of the patches of different ages. Mathematical implementation of these cases of spatio-temporal heterogeneity employs the theory of structured (meta)populations.
The niche of a species is defined by its way of being regulated. Two species may coexist because their growth is regulated differently enough. Quantitatively, a species’ niche is given by the strength of its impact on, and sensitivity towards, the regulating variables. Specifying the pair of impact and sensitivity is the measurable alternative to the intuitive notion of resource utilization. The strength of competition between two species can be calculated from the overlap between the impact of one species and the sensitivity of the other. Robust coexistence requires sufficient differentiation in the impact as well as in the requirement niches; increasing similarity decreases robustness.