When the smoke alarm started beeping at 4 in the morning (even the stepladder wouldn't have been enough for me to reach the cabin ceiling), we embraced necessity and got up. Outside Mercury and Venus were rising above the hill and dawn was threatening, so we sat and watched while a small squadron of bats flashed back and forth around us. My better half was glad to see them, saying that they'd eat the mosquitoes.
Since the mosquitoes were eating me instead, I wondered what the bats were up to--clearly not deterrence. And they didn't fly all that close to us, where you'd think there'd be great pickings.
Light dawned. On me, anyway. Full mosquitoes are slower fliers, and are simultaneously easier targets and flavored with a little something extra. Perhaps the bats were waiting to get the full ones leaving us instead of interdicting the hungry ones headed our way.
How would you test this? Release known numbers of empty and full mosquitoes into a bat pen, and count the number that escape? Sounds kind of tedious, but it might be inexpensively do-able, with a large room, a pen, and big fans forcing the insects into a small screened room where you BlackFlag the beasts and count the corpses. You might even be able to get some preliminary separation between full and empty by shaking the tray of dead mosquitoes. You'd need lots of repetitions to get good statistics, since it would be easy to make counting errors, and sometimes the input group would be all empties and sometimes all fulls to let you try to measure the counting biases.
Now where does one find volunteers for the feeding chamber?