As thick as thieves - a bee and a moth

The sky islands of Arizona are treasure islands for naturalists. When the desert wildflowers wilt in the per-summer heat of April and May, in these mountain ranges spring is only beginning. One of my favorite mountains is Mount Graham in eastern Arizona. On its highest elevations, it has beautiful mixed conifer areas and mountain meadows.

Mertensia macdougalii.

 in early summer, a herbaceousl plant in the family Borraginaceae covers acres of these meadows: Mertensia macdougalii. The flowers are set in swirls typical for this family. They are blue when fresh but change color to purple and pink when they age - probably due to ph changes in the aging flowers. They remind me very much of  Pulmonaria, an early spring herald of European forests.
Studying those flowers (looking for bugs on them) I soon noticed little injuries on the upper part most of the drooping flowers. The culprits were obvious too: bees were chewing through the side of the flower's throat to get to the nectar, instead of laboriously crawling into the flower. Easier on the bees, but of course this way, the flowers were probably left unpollinated. The bees that I observed were all of one species in the genus Colletes.

Colletes sp. bee chewing into the side of the flower to get to the nectar

Many Colletes are specialists, foraging for pollen on only one group of plants (Wilson and Messinger Carril). Even though my Mnt. Graham species is not identified and Mertensia is not mentioned in the bee-plant pairings of the book by Wilson and Merringer Carril, I may have stumbled on another such pair, or I just happened upon a group of bees (they are non-social but often nest in aggregations) that had found a good easy nectar source and stuck to it for the time being for reasons of efficiency (constancy principle).

middle: Colletes bee. right: damaged flower; left: moth

But I may have observed a clue that points to a long established relationship, which in this case is based on thievery: After trying to photograph the thieving bees for a while and thus establishing a search image for them, I realized that I was repeatedly looking not at a bee, but at a moth that was using the bee-created access to the nectar source.  Interstingly, the moth (Caloreas apocynoglossa) was always sitting head down while the bees would sit head-up. However, the moths' wing pattern 'took that into account': a dark hind-end  gave the impression of the dark head of the bee, even including antennae, and  a light patch imitated the  reflection on the bee's wings.

Moth Caloreas apocynoglossa

Two  interpretations are possible:
The bees may be armed with a painful sting - at least in comparison with a helpless moth. Predators would avoid bees and moths. So this would be direct Batesian mimicry.
But solitary bees are not usually heavily defended and many birds feed on them.

Another possibility is that the moth gains some protection from its head-down orientation: Flycatchers tend to grab their prey by head and thorax, in this case the pretend-bee-thorax - so the moth may be able to escape from the misdirected attack with minor hind-wing injuries (same idea as in hairstreaks).

But: If the relationship between bee and moth is old and established enough to have resulted in adaptive changes in behavior and pattern of the moth, it may be save to assume that the relationship of this Colletes species and Mertensia macdougalii is even older and that the bees are specialists. Too bad for the plant then, that these bees are specialized thieves!

Quoted: Thee Bees in your Backyard by JS Wilson, OM Carril, Priceton University Press 2016