How does the representation rate of features in a landscape affect visual preferences? A case study from a post-mining landscape