Understanding the relationship between selection, genotype-phenotype map and loss of population diversity represents an important step towards more effective genetic programming (GP) algorithms. This paper describes an approach to capture dynamic changes in this relationship. We analyze the frequency distribution of points in the diversity plane defined by structural and semantic similarity measures. We test our methodology using standard GP (SGP) on a number of test problems, as well as Offspring Selection GP (OS-GP), an algorithmic flavor where selection is explicitly focused towards adaptive change. We end with a discussion about the implications of diversity maintenance for each of the tested algorithms. We conclude that diversity needs to be considered in the context of fitness improvement, and that more diversity is not necessarily beneficial in terms of solution quality.