Unitary associations 2 and 6 are found in only one section each (section 2 and 9, respectively), and are therefore weakly reproducible. They do not constitute useful biozones on their own.

Maximal cliques 2, 5, 10 and 11 occur in cycles (strong components). This implies serious inconsistencies in the stratigraphical ordering of these cliques.

Removing horizon 4 of section 1 significantly improves the quality of the biostratigraphical correlation. The number of contradictions is reduced from 22 to 15, and the cycles in the maximal cliques graph disappear.