Changeset 5410 for branches

DEV: tested over 200 years for a European temperate pixel with 7 forest PFTs. Ticket #302 was implemented as decsribed in the ticket. Ticket #440 was also fixed. The fix of ticket #440 should not affect the simulations because the values for when_growthinit that were used in de the code were correct but the values were not set to zero for PFTs without veget_max. This resulted in strange values for when_growthinit when the sum over age classes was plotted. The corrections resets when_growthinit to zero if veget_max = 0