If you look at any beaching, the corpses rot rather quickly, and if left to the motion of the waves, the bones get worn down by normal, pedestrian forces over a few years.
Unless they can be significantly moved away, tidal action is still problematic.
The other posters theory of a tidal wave might explain getting the corpses above tide line, true. If you assume the tidal forces were not acting on the corpses, then burying/covering the bones with a subsequent volcanic event makes sense.
As far as the 'dry and desert conditions' goes, you are projecting the current conditions backward.
In any case, if we ASSUME the present conditions also existed 2.5 million years ago, the dessicated bones would still wear to nubs with the sand/tide friction acting on them, unless of course, they were not subject to tide action, which has been my point.
A tidal wave stranding the corpses in a low lying area away from the beach, followed by a volcanic eruption or other stratifying event is plausible.
Plate tectonics are too slow, UNLESS you allow for rapid motion in the past, not a millimeter a year.
A millimeter a year is the average rate. The 2010 8.8 quake raised some beaches a couple meters. That would put anything stranded at the high tide mark well beyond the reach of waves.
It was a rapid event: http://www.threeimpacts-twoevents.com/wp-content/uploads/2012/10/SIMULTANEOUS-IMPACTS7.pdf
The plates were created by the event described in the slides at the link. Continental drift is falsified by new data (google maps).