The reaction will have a number of bases present. The enolate of the intermediate Michael addition reaction is a base and probably is a source of regenerated ethoxide. The aldol condensation will generate another alkoxide. This can also exchange with ethanol because of the higher concentration and regenerate ethoxide. If the aldol looses water as hydroxide, it will be another base.
When I wrote mechanisms, I included all of the steps, such as regeneration of ethoxide, borohydrides, aluminum hydrides, etc. It was my objective to make every step of a mechanism understandable. I suspect that ethoxide simply appeared over an arrow and thus may appear to be an additional amount of ethoxide being added. That ethoxide may be the same ethoxide that is generated upon protonation of the intermediate enolate by ethanol.