One of the best techniques is the "panel". For example, with the system of rotation adopted the type 2-2-2; every moth enters the sample for two consecutive months; following a break for two months, after which it is included again for another two months, and then leave permanently.
The rotation system adopted keeps unchanged half the sample from month to month and half from one month to the same month of the following year. This lets you use the sample data to assess:
the cyclical variation, defined as the change compared to the previous survey;
the trend change, meaning the change compared to the survey conducted in the same period last year.
You could mark the moths with ink when they are first captured and released. Then if they are caught again, you would not use data from the previously-captured moths. Or if the general activity of the moths is of interest, you might want to know how many times moths are re-captured; you could mark them with a different code at each re-capture.