I think the question is not well formulated. Mostly because "language concepts" (e.g. tense-aspect-mood systems, animacy, verbal valence, case systems...) reflect the workings of the human mind far, far more than they could ever reflect the human anatomy. Anatomy/biology of course influences massive swathes of semantics, but I don't think anything other than pure phonetics or sign language is influenced by anatomy.
Facial expressions and body language probably belong there, but tend to have a huge cultural element. Position of the ears, for example, is not something humans make (or can make) use of, but maybe elves could.
Some more semantics example:
- We make distinctions useful to us at our given size. We cannot tell apart things by smell, or by UV light the way other animals (or measuring instruments) can. We do not tell apart very tiny things by basic names (e.g. ants, mites, sand). If we could sense, say, gravitic variations or psychic energies, we might have entirely different ways to describe position or time.
- Time is linear to us, cause precedes effects and so on, we cannot watch again a past event. This obviously affects how language models time and mood. Consider the Observers from Fringe or any other species with a different perception of time. Their possibly utterly alien perception of time may well be the reason their writing is (apparently) indecipherable.
- "Fish" is not a proper biological class (and whether marine mammals are excluded from it in common parlance is debatable!) because we don't feel a need to be any more precise about it, but we would have better concepts for it if we were a marine species! Similarly, plants are divided, roughly, into, flowers, lianas, grass, trees and shrubs. This system is not even very efficient for Earth ecology!
- Expanding on the UV thing, color words are constrained by our visual capacity, and typically distinguish a surprisingly limited number of basics. Forget the hierarchy part, nonderived terms usually stop around 6-8 words, why? A species with different color perception would do things differently. Mark Rosenfelder spends an entire chapter on this point in his Conlanger's Lexipedia.
- As far as number base systems, decimal and vigesimal are most common in humans because we have 10 fingers, and 20 fingers and toes. (although a huge number of bases have been documented, I don't think any languages have a binary or ternary one)
- Kinship terminology is based off our own reproductive biology (and our gregariousness). What of a species where the embryo must be incubated in a member of a different species? What about a species where reproduction requires gametes from more than two individuals?