Remove eaten biomass

Once the animals module passes back in "eaten" fruits, we can remote this from the appropriate biomass pools. Would be easy to fake this data to integrate piping while waiting for the actual functionality from Animals

Alternatively, if the animals module uses the same biomass pools in the data object as we use, do we even need to do this? If we're essentially sharing biomass pools, maybe there's nothing to do (once the protocol is set & agreed on)

No files or tests are named. First resolve the biomass-pool ownership and protocol between the animals module and this module, then identify the integration entry points; the work is done when the agreed handling of eaten fruits is implemented and verified without duplicating shared-pool updates.

Written by the indexing model from the issue text.