Make particles easier to use

[jagoosw]

I started this intending to work on the giant kelp growth model, but realised it would be much easier if I finally cleaned up how we do particle models. This PR turns BiogeochemicalParticles into a type which stores biogeochemistry which specifies the evolution of fields in the same way as continuous biogeochemical models, e.g.:

(kelp::SugarKelp)(::Val{:A}, t, A, N, C, ...)

This separates out the numerics from the model so should make it a lot easier to implement other models.

I suppose I need to do this to sediments next...