Is this a perpallidus Common Kestrel?

 There's pervading view that the perpallidus subspecies of Common Kestrel is a rare winter visitor to Japan. But how rare might it be? It doesn't even occur according to the current edition (published 2012) of the OSJ Japan list! I first heard of 'pale continental birds' about 20 years ago but am still none the wiser as to their occurrence or identification; the vague “it’s paler” is as far as easily accessible information has taken me. But in what respect is it paler? 

Earlier this week I spotted a small bird of prey in flight about 800 metres distant across the fields. When I got the bins on it, it was clearly a Kestrel but a strikingly pale one, looking a rather uniform sandy colour. With the posibility of perpallidus in mind I dog-legged my way along the narrow roads through the patchwork of fields to where the bird had been. It was still present and I rattled off a few record shots from a healthy distance, not wanting to disturb it and have it disappear over the horizon.

I needn't have worried as it stuck to a very limited stretch of fields, either perching on short bamboo polls, bushes or farm outbuildings, when not hovering over the fields. In particular, one small bean field had a magnetic attraction, presumably because the cover looked/was more vole-friendly, and the bird frequently returned to hover persistently over it, as close as 10 metres from where I stood. It proved remarkably tolerant of approach when perched too. 

At one point another Kestrel overflew the bean field resulting in low aerial confrontations, leaving them sitting about 10 metres apart, first on the road then in a bare field, allowing brief but good comparison. Both were first winter birds, the pale individual a male and I suspect the other a female but that’s based purely on its larger size.

I returned two days later and found both birds in their exact same places. However temporary they may be, these areas seemed to be clearly defined territories, boundaries recognizable even to the human eye. Although they could both be on view at the same time, they remained within their respective small areas of what is actually a large expanse of fields, thus direct comparison was never possible except for that one occasion on the first visit.    

I’m not claiming this Kestrel is perpallidus, I don't know anything about their appearance, but if perpallidus is known as a pale version of the usual Kestrels here, then this has to be a bird of interest. It looked eye-catchingly pale at longer distance as no dark markings or areas of contrast could be made out. Looking at it perched in the distance I was reminded of Isabelline Shrike in terms of pale uniformity. At closer range it looked less remarkable as it shares the same basic pattern as intersinctus, however, the blackish markings aren’t heavily saturated, they look more grey than black. Further, the bars on the lower scapulars, tertials, and all wing coverts are noticeably narrower than on the other bird present. Even the flight feathers have a brownish cast compared to the blacker plumage of the larger bird. The upper scapulars are adult-type so not relevant to a direct comparison with the other bird.

One other point of interest, this bird spent as much time hovering to scan as perching to scan and never employed Merlin-like pursuit hunting methods. Rightly or wrongly, I always think this is suggestive of a continental bird where hunting for terrestrial mammals is likely to be more successful. I never saw the larger bird hover at all. 

As I haven’t been able to find any info on how to separate these taxa, or even whether they can be reliably separated; I'm at a loss. That perpallidus is believed to visit Japan in winter implies that it is possible to identify them, I've just yet to meet someone who can tell me more than 'it's paler'.

This is a pale, uniform Kestrel, even the flight feathers don't contrast greatly with the rest of the upperparts. The terminal tail band is the only feature that stands out as blackish and contrasting.  

Compare with this intersinctus present. This is an altogether darker bird and the tail band doesn't look strikingly black in contrast with the rest of the plumage. The coverts, especially towards the bend of the wing, and the whole of the upperparts look blackish with pale markings. Whereas the possible perpallidus is looks pale with dark markings.  

The intersinctus again. The black bars are broader than the paler ground; the wings coverts are particularly blackish. Compare with the possible perpallidus below.

The possible perpallidus. Ignoring the adult-type upper scapulars, the others have dark bars distinctly narrower than the pale ground, additionally the saturation is such that they appear grey not black. This is also true of the wing coverts, most markedly the lesser coverts. The primaries have a brownish cast.  

Quite unconcerned by my presence, sheltering from the slanting, sleety rain in the lee of a farm outbuilding.