From the screenshot, it looks like the pink is coming mostly from a lot of ASYNC_NETWORK_IO wait events, which usually means client applications are consuming results slower than the database is producing them, so the network buffer SQL Server is using to feed results to the application fills up and ends up waiting. The root causes are certainly going to be application specific (eg: possibly there’s a stored procedure that the optimizer is now picking a bad execution plan for, etc.), but a quick google of the wait event name led me to a MSDN blog entry that might be helpful.
If this wait event is new, then clearly something has changed either on the database or in the client applications executing on the database, or in the network between the two. I suggest the user select ASYNC_NETWORK_IO in the Events tab and identify the SQL experiencing this event: that should help lead the user to the changes/problems.
Wait Event Docs >