In biology, I think that form is ontologically prior to function. By this I mean that a change in form implies a change in function, but that a change in function does not imply a change in form. By form I mean phenotype. So, a proper biological ontology would be:
Genotype -- Phenotype -- Function
Phenotypic expression is causally determined by environmental constrains on the ability to reproduce genotypes to the next generation. Natural selection, in this case, is the hard-stop of genotype reproduction -- those who do not pass their information on will stop passing their information on. As such, environment is actually wider than natural selection, and natural selection only plays a role at the level of genotype. Natural drift would also fall in at the level of genotype. Sexual selection, however, would be ontologically separate from natural selection because it is a selection for phenotypes which then causes a selection for some genotypes.
In most biological species function can only be changed by phenotype, and phenotype by the three preceding mechanisms.
"Consciousness" is a separate evolutionary mechanism which operates on function in the limits of phenotypic expression -- or in some extreme cases, such as cloning, acts on genotypic expression. As a mechanism of evolution, it operates in the realm of function – the brain runs on functions and this mechanism of evolution is a function of genotypic reproduction and selection. This isn’t to say that consciousness isn’t more than this – this would just be the way one could explain consciousness in a biological context. More detailed explanations of consciousness would supervene on this general sketch.
I think this ontology accounts for the biological nature of consciousness, as well as its special place in nature while staying in the bounds of an ontological naturalism. These would be the reasons for adopting it.