Appendix 1.

The 'type description' of tristis

The original description of “Phylloscopus tristis” (Blyth 1843, Journ. Asiatic Soc. Bengal XII, p. 966-7) was based upon individuals on the wintering grounds in India and is as follows:

'Ph. tristis, Nobis.  Closely allied to Ph. rufus, but devoid of any greenish or yellowish tinge on the plumage, except on the fore-part of the wing underneath, and faintly margining the quills and tail externally; legs and claws black, or rather dull black (much darker than in Ph. rufus), except the under-surface of the toes which is yellow; bill also blackish, tinged with yellow at base of lower mandible, and the gape also yellow.  Length four inches and a half, by six inches and a half across; of wing two inches and one-eighth; tail an inch and three-quarters; bill to gape half an inch; and tarsus seven-sixteenths of an inch.  General colour greyish-brown, beneath paler and albescent, a faint rufous tinge on the breast, and not trace of yellowish on the lower tail-coverts, nor elsewhere than as described.*

[Footnote] * There is the faintest possible greenish tinge on the upper-parts of some that I have since procured, which colour is most developed on the margins of the secondaries, towards their base.'

 [Ed note: Ph. rufus was the scientific name for collybita and abietinus combined, before their separation.]

Individuals wintering in India may not all originate from eastern Siberia but the 'type description' conforms with modern interpretations of 'classic' tristis, viz. plumage lacking yellow away from the underwing and lacking olive except faintly on the margins of remiges and rectrices; upperparts grey-brown, underparts off-white with a faint 'rufous' tinge on the breast. The 'footnote' perhaps hints at individuals with 'fulvescens' traits sensu Sushkin 2015 et al.

The form 'fulvescens' was first described as a new taxon by Severtzov (1873), based upon a series of 100 migrants collected in Turkestan.

Severtzov’s description (kindly provided and translated by Dr Vladimir Loskot, Curator of the Ornithological Department, Zoological Institute, Russian Academy of Sciences, St. Petersburg) was as follows:

‘Ficedula (Phyllopneuste) fulvescens, nob. - (Ph. tristis? Gould). Upper parts, from forehead to upper tail coverts and small wing coverts rusty-grey with olive tint, olive-brown in autumn; supercilium and under parts rather pale rufous-yellowish, brighter in autumn; cheeks not pure rusty; wings and tail feathers blackish, with olive fringes which, on coverts, cover blackish middle of feathers; small underwing coverts sulphur yellow; first primary twice as long as its coverts; 3=4=5>6>7>2>8. Male and female do not differ from each other; in juvenile birds, differing in general only in looser texture of feathers, sometimes  unclear longitudinal stripes of pale sulphur yellow colour are present on the breast; in others, the breast stripes are similar but greyish (var. naevia), and then all the plumage is more greyish than usually. Bill and legs black; bill is relatively small even for a Leaf Warbler, and claws are large, especially on the hind and mid toes.’

Appendix 2.

Taxonomic history of 'fulvescens'

The form 'fulvescens' was described from migrants collected in Turkestan (Severtzov 1873) but the name and breeding range were equated by Sushkin and others with western Siberia. Later, Vaurie (1954) included in 'fulvescens' chiffchaffs from much further south, in SW Transcaspia and NE Iran. This was repeated by Williamson (1962), though he noted that this was a 'curiously disjunct distribution' and he also subsumed 'fulvescens' with tristis. Locations noted by Ticehurst (1938) of individuals with intermediate 'riphaeus' traits (deemed to be 'on passage', presumably because they were assumed to originate from the tristis / abietinus overlap zone) included the south coast of the Caspian and the Elborz Mountains in Iran. Chiffchaffs breeding in woodlands in NE Iran and Kopet Dagh in Turkmenistan are indeed greyer above, whiter below and deficient in yellow compared with typical abietinus but lack other features of tristis / 'fulvescens' (see main text). Their song is much like collybita / abietinus but their call is similar to tristis. They are now designated as menzbieri. At the time of Ticehurst's monograph (1938), menzbieri had only recently been described (Shestoperov 1937) and was not included (while caucasicus was not described (by Loskot) until 1991). As late as 1992, this confusion over menzbieri persisted, with BWP6 opting to include it within tristis but noting that 'it may belong to abietinus'. The precise appearance of menzbieri, the exact boundaries of its range (see Raković et al. 2019) and its appropriate taxonomy remain equivocal. See  <here> for examples of Chiffchaffs in the Elborz mountains of Iran.

A key issue in defining acceptable boundaries for diagnosing tristis (including in validation of the tristis panel's approach) is : ‘does any level of introgression exclude diagnosis as tristis?’. The following comments have been separated from the main text as they involve genetic issues which seem relevant but for which adequate evaluation requires greater expertise in genetics than I possess. Thus, they are subject to correction.

As noted in the main text, at the population level there are significant differences in the levels of genetic admixture between the sympatric zone and the allopatric zone (hence the adoption here of the colloquial terms 'riphaeus' and  'fulvescens' sensu stricto). In the sympatric region ('riphaeus' zone), the gene pool embraces significant levels of both tristis and abietinus alleles. The genetic studies of Marova et al. and Shiplilina et al. have shown that there are substantial numbers of first generation (F1) hybrids (tristis x abietinus as popularly understood) and early backcrosses. Such 'direct' hybrids cannot be designated as 'Siberian Chiffchaffs'. In contrast, the gene pool in the allopatric region in the West Siberian Plain (which embraces around a third of the total range of tristis) remains predominantly that of tristis. Individuals with trace levels of 'misplaced yellow' most probably derive from a 'fraction' of abietinus alleles (to quote Shipilina et al.) scattered in a tristis gene pool and arising from isolated instances of hybridization radically diluted by many generations of recurrent backcrossing with thoroughbred tristis. Such low-level introgression does not necessarily disbar diagnosis as tristis. There are certain genes which are critical in defining the taxon identity and divergence and which may be relatively 'robust'. See Talla et al. (2017) for on-going studies of the influence of differing regions of the genome on speciation within the chiffchaff complex.

While genetic admixture in the West Siberian Plain (beyond the direct hybridization zone) may be low at the at the population level, this does not mean that the genetic composition of a given individual from the West Siberian Plain can be guaranteed to be close to thoroughbred tristis. Although there will be a high probability that this is so, the 'genetic hybrid index' of an individual will be unknown. Unfortunately, as appearance too is determined by a relatively small region of the genome, neither does appearance provide a reliable guide to genetic composition. The findings of Shipilina et al. indicate that an individual Chiffchaff with mixed alleles may yet resemble and sound like a 'good' tristis.

Discussing Chiffchaffs from the sympatric region, Talla et al. (2017) also wrote :

'We foresee that the next step will be more extensive sampling within the sympatric region, including birds with intermediate phenotypes and using detailed morphological and behavioural studies, combined with resequencing efforts of candidate [genetic] regions, to find significant associations between specific alleles and phenotypic traits of interest.'

Such studies could be extended usefully beyond the sympatric region, to Chiffchaffs from the West Siberian Plain.

The 'pragmatic approach to identification' advocated in the discussions on this website and elsewhere includes recognition that there will be a margin of error. Among other factors, the level of error will be depend upon the origins of tristis candidates reaching western Europe : what percentage arrives from the hybrid zone, what percentage from the introgression zone and what percentage from beyond the Yenisey (the last likely to be small?). Only extensive whole-genome studies of tristis candidates reaching Western Europe seem capable of answering this question.