The automated provisioning and management of Cloud applications is supported by various general-purpose technologies that provide generic management functionalities such as scaling components or automatically redeploying parts of a Cloud application. However, if complex applications have to be managed, these technologies reach their limits and individual, application-specific processes must be created to automate the execution of holistic management tasks that cannot be implemented in a generic manner. Unfortunately, creating such processes from scratch is time-consuming, error-prone, and knowledge-intensive, thus, leading to inefficient developments of new applications. In this paper, we present an approach that tackles these issues by enabling the usage of choreographies to systematically combine available management workflows of existing application building blocks. Moreover, we show how these choreographies can be merged into single, executable workflows in order to enable their automated execution. To validate the approach, we apply the concept to the choreography language BPEL4Chor and the Cloud standard TOSCA. In addition, we extend the Cloud application management ecosystem OpenTOSCA to support executing management choreographies.