This commit mostly concern changes and bug fix for land_cover_change and species_change subroutine. (1) it is now possible to use portfolio scenario like BESTLCA or BESTT2M FM and species change maps. (2) A check for NAN value in the mass_balance_check subroutine was added in order to stop the run when pool_start, pool_end or check_intern are NAN. I solved one bug related to NAN pool into stomate_phenology.f90. (3) Fix a bug in land_cover_change wich initialize PLANT_STATUS= inone(instead of iprescribe), LAST_CUT=1(instead of 0), AGE_STAND=1(instead of 0), when the Veget_max is equal to zero at the end of the subroutine. This bug produced a model divergence when PFT proportions were forced by PFT map with ou without land cover change.
