Okay, having looked at photos of all the possible species, I’m pretty well satisfied that Arizona white oak (Quercus arizonica) was the host of all of the galls I collected in Madera Canyon that actually produced adult cynipid wasps. Armed with this information, I took a look at Lewis H. Weld’s Cynipid Galls of the Southwest (which I’ve just discovered is available online here, along with Cynipid Galls of the Pacific Slope. Cynipid Galls of the Eastern United States isn’t there, but maybe it will be eventually.) While perusing the black and white photographs in the back of the book to see if anything matched galls #1 and 2, I happened to spot one with a clear resemblance to gall #3, which is a fuzzy red gall attached to the midrib on the lower leaf surface.
I checked Weld’s description, which clinched it: “Wool crimson.” I’m pretty sure that this gall, which I encountered both times I visited Madera Canyon, is caused by Cynips quercusnubila.
The first wasps to emerge from these galls were eulophid parasitoids, about 2 mm long:
Christer Hansson identified them as belonging to the genus Aprostocetus, and that’s as far as the identification will go. He tells me that not only are there no useful keys to the North American species of any genus in this subfamily (Tetrastichinae), but Aprostocetus in particular has “probably hundreds” of species in North America, of which most are undescribed.
A week after the eulophids started emerging, the first of the adult cynipids appeared. Like those from gall #1, and unlike the ones I’m used to seeing from eastern oak galls, they had distinct wing markings.
After another week, a third type of wasp emerged.
This one is torymid parasitoid, and if I ever get anyone to examine the specimen, I suspect I will be told that it is in the genus Torymus and can’t be identified to species because it’s a male.