Can we predict the number of plant species from the richness of a few common genera, families or orders?