A%*%N1*(1-h)
to
build_Leslie(s = s*(1-h), f = f)

Exact same results, but easier to understand

(to avoid artificial dilution due to high expected nb of newborns from the SAD)

Also modified "pop_project" functions accordingly

added small piece to avoid red-error on plots while wainting for a value

models : change sensitivity of lambda approximation (claibration) to speed up the simulation (it really improved the speed !)