What about a similar gradual rule for varying sentience levels of animal?

A quantitative measure of sentience seems much more reasonable than a binary measure. I'm not a biologist, though, and so don't have a good sense of how sharp the gradations of sentience in animals are; I would naively expect basically every level of sentience from 'doesn't have a central nervous system' to 'beyond humans' to be possible, but don't know if there are bands that aren't occupied for various practical reasons.