>I don’t think so… On looking up chatoyancy, it looks like it depends on having some sort of parallel grain in the material, while this effect depends on having linear features running in many different random directions.
Chatoyancy depends on parallel grains scattering light in the plane perpendicular to the grain and containing the source. The effect described in the OP is scattering by the sufficiently nearly parallel fraction of all the randomly oriented structures in this same configuration. That is, all the branches are random, but the reflecting subset in a particular region are all nearly parallel and tun perpendicular to the plane containing the reflecting spots, the sun, and the observer. I say in a particular region because as you travel around the line of sight to the sun, their angle changes.
So, I think the effect the OP notices is that, of all the branches, those that are at the right angle are emphasized by reflecting, which is chatoyancy at work; and, this is on a large scale a circular grain effect because you are rotating the plane in which the branches are made reflecting.