Software ecosystems (e.g., npm, PyPI) are the backbone of modern software developments. Developers add new packages to ecosystems every day to solve new problems or provide alternative solutions, causing obsolete packages to decline in their importance to the community. Packages in decline are reused less overtime and may become less frequently maintained. Thus, developers usually migrate their dependencies to better alternatives. Replacing packages in decline with better alternatives requires time and effort by developers to identify packages that need to be replaced, find the alternatives, asset migration benefits, and finally, perform the migration. This paper proposes an approach that automatically identifies packages that need to be replaced and finds their alternatives supported with real-world examples of open source projects performing the suggested migrations. At its core, our approach relies on the dependency migration patterns performed in the ecosystem to suggest migrations to other developers. We evaluated our approach on the npm ecosystem and found that 96% of the suggested alternatives are accurate. Furthermore, by surveying expert JavaScript developers, 67% of them indicate that they will use our suggested alternative packages in their future projects.