Well I think that getting a basline like you are getting now from your findings, and changing one variable such as types of food fed to your T and recording observations over time will give you data to prove or disprove a hypothesis. Is that the six sigma approach? I guess the aquarium is the place to start because it is definately easier to control than your pond. Just my $.02, I am very interested in your findings. I hypothesize that the T in the Aquarium might prefer the crickets after regular exposure to both types of forage.
