I read an article that discussed a Japanese physicists observations about earthworms exiting the ground before one of their big earthquakes.
http://animalsandearthquakes.com/ikeya.htm
Electric currents have been used for earthworm extraction going back many decades (worm getters), but I could not find any studies where their sensitivity to specific currents levels was measured.
I have raised garden beds which are prolific with earthworms so I decided to see if I could get them to respond with fairly low electric currents. I used #12 solid copper wire as electrodes, starting out with 6" length and then later making some 12". I started out with a 38.6 VDC source, an old HP printer transformer, the 6" electrodes vertically in the ground and 10" spacing, this giving me approximately 60 ma no matter where I inserted the electrodes in the garden bed.
I had multiple worms exiting the ground within 60 seconds (worms move slow). I changed the spacing to 16" (to a fresh area) and the current was just a few milliamps lower, and the same worm extraction experience was noted.
The next day I used two different transformers, a 12 VDC and a 9 VAC and experimented with the spacing. The current at 12 VDC was about 20 ma and that for the 9 VAC, approx. 15 ma current. There was a slight difference between worm exit counts and size AC vs DC, but the current was slightly different.
The following day I prepared the longer electrodes, 12", and experimented with various spacings, AC and DC (both 12 and 38 V). I also prepared a 13" insulated electrode where only 1" was exposed and inserted that into the ground so that the current had a somewhat vertical component. The number of worms exiting varied somewhat with the applied current, which depended upon electrode size and insertion depth. In the maximum current area I suspect worms were immobilized, but this is conjecture only because of where I observed worms exiting relative to the electrodes.
I did not determine the minimum current at which worms would exit the ground but I could get exiting with a little as 3 ma between electrodes, which means the actual current differential experienced by the worms has to be in the sub-milliamp range.
I can conclude the earthworms will exit the ground if there is a sustained current. Worms don't move fast, thus they probably started toward the surface as soon as they experienced a current level that was uncomfortable to them. I also had small garden centipedes (beneficials) exiting from the ground in the same area as the worms, thus they were being irritated by the same in-ground currents as the worms.
There were a number of peculiarities noted in where the worms exited relative to the electrodes but more studies are needed with better controls to see if these peculiarities are significant. I also placed some of the worms that had exited and moved away from the current influence directly in between the electrodes and watched their reaction. Size seems to matter, and if they couldn't exit immediately from the area of maximum current influence they had problems. The next day I used multiple electrode configurations and switching the polarity of inner and outer electrodes, this to examine if worms have a polarity sensitivity.
From what is known, the earth currents generated before and during earthquakes are more a DC type with variations in magnitude, it is not sinusoidal with plus and minus swings.