Ant Colony Optimization (ACO) is a metaheuristic which has been successfully applied to tackle various combinatorial optimization problems (COPs), but its ability to cope with hard constrained COPs has not yet been explored widely. Since most real world optimization problems, e.g., job scheduling, rostering or timetabling are subject to hard constraints, the investigation of different constraint handling methods within ACO becomes an important area of study. We explore the applicability of Stochastic Ranking as a soft handling technique and propose a bounding scheme as an extension to an existing ACO implementation, that handles constraints in a hard fashion, by utilizing Constraint Programming. Based on this extension we suggest a new search guiding approach that replaces static heuristic values by dynamically calculated values. Moreover, we attempt a loose coupling of soft and hard constraint handling techniques in one algorithm. Conducted empirical studies showed that the calculated upper bound on the solution quality is not tight enough to significantly improve the performance of the existing algorithm. In consequence, we also did not observe a significant performance enhancement for the other attempted extensions.