There's quite a few process that transfer carbon, nitrogen and phosphorus between the animal and soil+litter models. The number of variables could be reduced by expressing these as CNP triplets instead of individual variables.
I've already done this for excrement and carcass decay (see #1224), but it would be good to do it for all the exchange processes