The best option for frequent water changes is to use an electronically-controlled water dispensing system. I am sure we could have contacted an engineer to design such a system but our budget for this project is already stretched thin and paying for a professional system was not in the original budget. Thus, the task becomes do-it-yourself, which is not an uncommon phenomenon in our field. Thankfully, I have a brother who is an electrical engineer, and he gave me some free advice for designing and programming the system. We decided on a system that uses solenoid valves, which are normally closed but can be opened by passing an electric current through them. By connecting these valves to timers we can automatically program the valves to open at specific times and stay open for a sufficient amount of time to exchange the water.
In the end, setting up this single experiment required a basic understanding of plumbing, electrical, hydrology, and programming. As I said above, this is not an unusual situation for experimental ecologists. Much of the equipment we need cannot be purchased off the shelf or even out of specialty catalogs. Some of the equipment is passed down through the ages or can be built from designs by others, but most of the time a particular experiment requires a significant time investment in tinkering, building, and adapting. The difficulty in designing and building equipment is a relatively invisible facet of an experiment that is often downplayed in presentations and publications. Anyone who has tried to replicate an experimental apparatus is likely well aware of the missing details about equipment; however, most of the time the people that initially designed an experiment and the necessary equipment are more than willing to help. In the end I think all the effort is worth it because well-designed equipment leads to better experiments, which generates better data and advances our science.