Can't simulate proportions with known correlation and contrains in R

Question

The problem:

I have been sampling 5 categories over 6 months and their effects on certain enviromental activity over the months their proportion has varied like this: