You may want to separate your crawled data into different indexes. If that's the case, you can use actions to accomplish this by specifying a different indexName for each action. You can also use the parameter pathsToMatch in an action to specify the webpages you want to limit. This will add the records you add in your action to separate indices.
You can either use the common domain as the StartUrls, or use a set of entry points. We recommend reviewing this guide for more insight on how to configure your StartUrls.