You have a Kubernetes Deployment managing an application workload that has changing demand throughout the day. What feature of Kubernetes can be used to automatically scale the number of pods in the Deployment as load increases and delete pods when demand decreases?