Nudibranch diversity along a radioactive gradient caused by the Fukushima Daiichi disaster.
One could devise a whole bunch of projects dealing with gradients: salinity gradients, fertilizer run-off gradients, animal husbandry waste, industrial pollutants (coal ash, hot water discharge from nuke power plant), etc... So something like "Freshwater benthic algal community diversity change in a headwater stream caused by sheep ranching in alpine meadows." How about genotypic changes in cattail associated with adaptation to salt discharge from deicing roadways.
It would be possible to do some projects using microcosms, or mesocosms. There are some artificial enhanced CO2 experiments that could be used as models for such a project. This could work if you are setting up the start of a long term project.
The problem is usually not in finding interesting questions. The problem is usually figuring out how to pay for the people, equipment, and supplies needed to conduct the project.