Hrm... If the specific frequency response is known, an appropriately-tuned LED might be more efficient.

Many insects are attracted to UV wavelengths preferentially. Stinkbugs might prefer IR frequencies.

They like heat, which incandescent lights produce more of. I'm sure the IR would work to some extent but it seems like you're overcomplicating this!

Insects tend to have very simple sensing and processing systems and very specific signals or attractants. There's the history of the Australian beetle that was nearly driven to extinction by discarded beer bottles: the colour and shine of these precisely matched the beetle's mate-selection heuristic, and like many a man it was driven close to extinction by drink....

https://www.thoughtco.com/the-giant-jewel-beetle-1968152

At large scale (think again: ag), a low-power sensor-specific signal might be worth the overhead of a specifically-tuned IR emitter. This might also avoid collateral damage in attracting other insects based on a broader signal. In this case, the example of UV-based "bug zappers", now illegal in many areas, because they were too effective in attracting, and killing, insect life. To the extent that the insect's predator populations (including songbirds) were affected.

And, for greater irony: the zappers didn't attract the insects for which they were sold as a control device: mosquitos and other biting insects, which are instead drawn by CO2.

https://www.sciencedaily.com/releases/1997/07/970730060806.h...