32 



P.C. RASMUSSEN AND N.J. COLLAR 



exclusively found in the Eastern Himalaya as well as others occur- 

 ring only in the north-western portions of India.' A statement by Ali 

 and Ripley (1969) concerning intermedia - 'never consciously seen 

 alive in the wild state by any ornithologist' - remains true today. 

 Recent reports within India (R. Bhargava pers. comm. 1996, in litt. 

 1997, 1998; Ahmed et al. 1996, Anon. 1997, 1998a, b; Mookerjee 

 1997; Bhargava 1998; Them 1998), and three birds identified as 

 intermedia (Sane 1975, 1977, Sane et al. 1987, S. R. Sane pers. 

 comm. 1 997), do not clear up the mystery, as all refer to captive birds 

 of uncertain provenance, and some of these are of problematic 

 identification as well (Rasmussen and Collar 1998, Bhargava 1998). 

 Appeals for information (Rothschild 1907, Sane 1977, Wirth 1990, 

 Inskipp and Inskipp 1995) have not led to the discovery of a wild 

 population. 



If intermedia were a typical diurnal, noisy Psittacula, it would be 

 a most unusual bird, not only for having escaped the attentions of 

 field ornithologists for over a century in one of the best-known parts 

 of tropical Asia, but also for showing complete intermediacy in 

 numerous characters between two clearly differentiated congeners. 

 There are two possible explanations for this double circumstance: 

 the first is that it is an extremely rare species and therefore requires 

 concerted conservation attention; the second is that it is not a species 

 at all, but a hybrid. Only the second explanation accounts for both of 

 its unusual traits. In this paper we reexamine the evidence for 

 specific status vs. hybrid origin of intermedia, based on plumage and 

 mensural analyses both of museum specimens and of newly located 

 captive birds of known parentage. 



METHODS 



Specimens examined 



Eight museum specimens have been published in the primary litera- 

 ture as Psittacula intermedia, although one (AMNH 621545) has 

 been considered an immature himalayana (Biswas 1959, 1990, 

 Forshaw 1973). Each was thoroughly examined, photographed, and 

 measured for this study: AMNH 62 1 539 (holotype), 62 1 540-62 1542, 

 621544-621545; BMNH 1980.3.1; and BNHS (Bombay Natural 

 History Society) 26758. In addition, we examined another 

 uncatalogued specimen belonging to Mr. Sane, as well as a photo- 

 graph of three unaccessioned specimens in the possession of R. 

 Bhargava. 



In the early 1930s, Rothschild's entire series of intermedia went 

 to AMNH along with most of the rest of his collection. Subse- 

 quently, one (BMNH 1980.3.1, formerly AMNH 621543) was 

 exchanged to the then British Museum (Natural History) (M.R 

 Walters, pers. comm. 1997), where it had already resided since 1959 

 on long-term loan (Knox and Walters 1994). The BMNH specimen 

 was lent to AMNH so that we could compare it there with the 

 remainder of the series. 



A colour transparency of AMNH 621540 (placed with the speci- 

 mens; date and photographer unknown) taken sometime after 1973 



- based on an accompanying note: 'Psittacula 'intermedia' believed 

 to be a hybrid himalayana x cyanocephala see Forshaw (1973: 336)' 



- shows that it had long central rectrices when the photo was taken, 

 but these were lacking when the specimen was first photographed by 

 PCR in 1993, and it cannot now be determined if the rectrices were 

 fully grown. Estimates of lengths of the tail and of the yellow tip of 

 the central rectrix of AMNH 621540 were made from the transpar- 

 ency, in which the subject is 1/3 natural size and photographed from 

 the side. 



BNHS 26758 is essentially dataless (label data: male, aviary bird. 



S. R. Sane, Bombay, 12/90), as is Sane's second uncatalogued 

 specimen; these are two of the three birds examined from his 

 collection. The first may or may not be the specimen described in 

 Sane (1975, 1977), referred to as having died in 1978, and as being 

 in the BNHS collection (Sane et al. 1 987), but if ' 1 2/90 ' refers either 

 to date of death or to date of accession it can hardly be the same 

 individual. 



The only other specimens reputed to be intermedia of which we 

 are aware are those preserved by R. Bhargava, and we have seen 

 photos of three of those. However, adult female and especially 

 immature intermedia (see below) would readily escape notice among 

 series of similar congeners, and may well exist undetected in mu- 

 seum collections. 



We assessed variability among the eight published putative 

 intermedia specimens in the plumage and mensural characters listed 

 in Tables 1^4 and the Appendix. 



Photographic evidence 



Most of the published information on intermedia was recently 

 summarized by Inskipp and Inskipp (1995). Through perusal of the 

 avicultural literature we located an additional, previously unrecog- 

 nized, published photograph of an intermedia-tike bird, and 

 correspondence with aviculturists and researchers (after the main 

 statistical analyses for this paper were complete) resulted in addi- 

 tional unpublished information, including the location of several 

 more captive birds, some of documented parentage. 



Hybrid diagnoses 



For hybrid diagnoses (Graves 1990), plumage, other external char- 

 acteristics, and measurements of adult males were compared among 

 the species of Psittacula that either had been suggested previously as 

 possible parental taxa (Husain 1959, Forshaw 1973, Wolters 1975) 

 or for which the phenotype of the presumptive hybrids indicated the 

 likelihood of those species being involved. Sane's birds were com- 

 pared indirectly with the other intermedia specimens (Table 1 ) and 

 with series of adult male cyanocephala and Rose-ringed Parakeets P. 

 krameri (Table 2), while other intermedia were compared with 

 series assembled at the National Museum of Natural History (USNM) 

 of each of the potential parental species: cyanocephala (n = 21), 

 himalayana (n = 26), Grey-headed Parakeet P. finschii (n = 17), 

 krameri (n = 10), and Blossom-headed Parakeet P. roseata (n=l 1). 

 Additionally, all adult males of these species in the collections of 

 AMNH, the Academy of Natural Sciences of Philadelphia (ANSP), 

 the Museum of Comparative Zoology (MCZ), and the University of 

 Michigan Museum of Zoology (UMMZ), as well as several smaller 

 collections, were examined and measured; their plumage characters 

 (which did not differ materially from those of the series assembled 

 at USNM) were not included in the analyses, but their measurements 

 are included in the Appendix and in the statistical analyses. A few 

 unsexed specimens with plumage characteristics diagnostic of males 

 were included. All other species of Psittacula were ruled out as 

 potential parental species as they have plumage characters strongly 

 incompatible with the phenotype of specimens reputed to be 

 intermedia. 



Mensural characters of adult males (listed in Appendix) were 

 used to evaluate which (if any) of the species listed above could 

 potentially be parental species of the intermedia specimens. Measure- 

 ments taken as far as possible for each specimen were: culmen 

 length (from distal edge of cere); height and width of maxilla (upper 

 mandible, at distal edge of cere); minimum distance between nares; 

 width of (lower) mandible; wing length (straightened and flattened); 

 shortfalls of each primary (P1-P10, with PI outermost) from 

 wingpoint; for PI, distance from notch on inner web to feather tip, 

 maximum width, and width at notch; widths of P2-5, each taken at 



