Wind River earthquake of 2013: unusually deep and in the mantle

Scenary of the Wind River area illustrating that the
surface is covered with sedimentary rocks and
supporting the observation that there is no
active volcanism within about 200 km. From the BBC
article by Matt Walker cited in the text.

Today the BBC has a nice article  by Matt Walker pointing out an article by T.J. Craig and R. Heyburn in Earth and Planetary Science Letters (425, pp. 12-23, 2015) on a very deep earthquake in the Wind River range of Wyoming. The authors point out that while earthquakes in the mantle of the oceanic lithosphere are common, well-documented and well-constrained earthquakes in continental mantle are rare, partly because not only must the depth of the earthquake be constrained, but the depth of the Moho also has to be known. In 2013 there was a M4.8 earthquake in the Wind River Range of central Wyoming, a region that is normally relatively quiet seismically.  Only once in the past ~60 years has a M5 earthquake been recorded, and most quakes do not even exceed M4. This earthquake, and one single aftershock, were initially recorded to be between 70-80 km depth.
     The (very mathematical) analysis in this paper constrains the depth of the earthquake to 75 km (plus or minus 8 km), and makes it the second deepest earthquake now identified under a stable continental region. The depth of the Moho in this area is well constrained to be between 42-50 km, so the earthquake occurred well within the mantle, probably >20 km below the base of the crust.  The only two other comparable earthquakes that the authors know of are the 1979 Randolphe, Utah, quake at 90 km, and the 2000 Arafura Sea earthquake at 61 km.
     What caused this earthquake? The authors mention the possibility that the quake may result from the migration of fluids within the mantle.  Such activity is known to cause microseismic activity at great depths in volcanic regions. However, the Wind River range is more than 200 km from the nearest volcanic region, the hot spot of Yellowstone. They also argue that since the Wind River earthquake ruptured an area of about 1,000,000 square meters, this area is much larger than would be expected from fluid-related origin.  They cannot rule out this possibility, but prefer an explanation that the earthquake resulted from brittle fracture due to tectonically-derived stresses.