A report from the British Ornithologists’ Union Taxonomic Sub-committee The taxonomic status of Carrion and Hooded Crows David T. Parkin, Martin Collinson, Andreas J. Helbig,Alan G. Knox and George Sangster

Steve McQueen ABSTRACT The taxonomic status of Carrion Corvus corone and Hooded Crows C. cornix is reviewed. As well as the obvious differences in plumage between the two, there is good evidence for non-random mating and reduced fitness of hybrids between Carrion and Hooded Crows, which together provide sufficient evidence for them to be regarded as separate species under most species concepts. Differences in vocalisations and ecology support this distinction. It is therefore recommended that Carrion Crow and Hooded Crow be treated as separate species.

here are two distinct groups of crows west and orientalis in the east. The grey-and- Corvus corone in the Palearctic. The all- black cornix group (the ‘Hooded Crows’) com- Tblack corone group (the ‘Carrion Crows’) prises a number of races in northern and has two disjunct races: nominate corone in the eastern Europe, from Corsica and Italy east-

274 © British Birds 96 • June 2003 • 274-290 Taxonomic status of Carrion and Hooded Crows

Fig. 1. The distribution of Carrion Corvus corone (red) and Hooded Crows C. cornix (green), showing the approximate extent of the various races of each species. For more details see text. Distribution based on the maps in BWP, with kind permission of Oxford University Press. wards, through the Middle East to northern, The race sharpii is found in Sardinia, Sicily and western and central Asia. These include cornix, south mainland Italy; from coastal Croatia east capellanus, sharpii and pallescens. to Romania, and Bulgaria south to Greece (including Crete); through Moldova and Turkey Distribution (except in the south), east through southern ‘Carrion Crow’ C. c. corone (Goodwin 1986; Ukraine and northern Kazakhstan to the Cramp & Perrins 1994; Madge & Burn 1994) is western Altai; and through the Caucasus and found from England and Wales south to Iberia northern Iran. It grades into cornix in Ukraine, and the southern slopes of the Alps, and east- European Russia, and Kazakhstan, overlapping wards to Austria (fig. 1). It hybridises with C. c. and partly hybridising with orientalis in Turk- cornix in several areas, including northern Scot- menistan and Kazakhstan. The race pallescens land, southern Denmark, parts of eastern occurs from Cyprus and southeast Turkey to Germany and the Czech Republic, along the northern Iraq and Egypt. The race capellanus borders of Austria with Hungary and Slovakia, occurs in Iraq and southwest Iran. and in the southern foothills of the Alps. The eastern race, orientalis, occurs from Iran north Geographical variation to the Aral Sea, across to the western Altai. Else- Within the Carrion Crow group, there is only where in Asia, it is found through south and slight variation (Cramp & Perrins 1994): orien- east Kazakhstan to Kashmir, Sinkiang, Mon- talis differs from nominate corone by having golia, central and northern China, Korea and longer wing, tail and tarsus. In Spain, corone is Japan. Farther north, it occurs through the smaller than elsewhere in western Europe, espe- Yenisey valley eastwards to Kamchatka. cially in terms of bill depth. By contrast, those ‘Hooded Crow’ C. c. cornix (Goodwin 1986; orientalis from the mountains and north Cramp & Perrins 1994; Madge & Burn 1994) average larger than elsewhere. The bill is slightly occurs in the Faeroes, Ireland and the Isle of longer and more slender in the west, compara- Man, northern Scotland, Denmark and tively thicker and shorter in the east, but these Fennoscandia, and east to the Yenisey valley differences are slight and clinal: no races other (where it interbreeds with orientalis). It also than orientalis are generally recognised in the occurs in eastern Germany, Hungary, northwest east. Romania, northern Croatia, northern Italy and Within the Hooded Crow group, variation is Corsica, and northeast to about 49°N in Russia. also slight and clinal (Cramp & Perrins 1994),

British Birds 96 • June 2003 • 274-290 275 Taxonomic status of Carrion and Hooded Crows

involving the dimensions of wing, bill and cially in terms of bill, tarsus and foot size. The tarsus, and tone of grey feathering. Birds are throat feathers are rather long, and, as in orien- generally smaller in the south, and paler in the talis, the tail is long and graduated. It is, in fact, south and east. In northern Europe (including closely similar to Pied Crow C. albus, but the the Faeroes, Scotland and Ireland), cornix is greyish-white is more extensive and the centre typically large, close to nominate corone from of the breast is black. The race capellanus is western Europe in size, with a light ash-grey sometimes considered a separate species, but body. Those from southern and central Europe the plumage pattern is similar to that of other are slightly darker, and are sometimes separated races of the cornix group, its proportions are as ‘subcornix’. In Siberia, cornix averages paler, rather similar to orientalis, and it is said to but the differences are very slight. The race intergrade with sharpii in Iran (Meinertzhagen sharpii is equal in size to cornix, or slightly 1926; Meise 1928; Vaurie 1959). smaller, but the grey is distinctly paler. Euro- pean sharpii are larger than those from Asia, Hybrid zones especially in terms of bill size: those from the The boundary between the corone and cornix Don, Ural and lower Volga rivers are sometimes groups is usually sharp, with a narrow zone of separated as ‘khozaricus’. Farther south and hybridisation where birds of both phenotypes west, the situation becomes more complex. and various hybrid forms occur. Outside this Birds from Turkey, Caucasus (‘kaucasicus’) and hybrid zone, a wider border or introgression northern Iran are slightly smaller than sharpii, zone occurs, in which most birds are typical of and the grey is slightly darker, intermediate one form but where some individuals show between that of cornix and sharpii. Variation in influences of the other. Detailed maps of the the size and colour of birds from Italy through hybridisation zones between nominate corone Turkey is slight and inconsistent, so all are and cornix can be found in Meise (1928), usually included in sharpii. The race pallescens Richter (1958), Cook (1975), Dybbro (1976), from southern Turkey through the Middle East Picozzi (1976) and Bährmann (1978). In Asia, to Egypt is the smallest race, and the grey corn orientalis and cornix/sharpii were perhaps origi- be even paler than on sharpii. Finally, capellanus nally well separated and contemporary overlap forms the pale end of a cline in colour, being may be a result of habitat change. The zone of cream or pale silver rather than grey in fresh hybridisation here is apparently not stable (see plumage, and almost white when worn. This Meise 1928; Vaurie 1954; Korelov et al. 1974). taxon is larger than adjacent populations, espe- The introgression zone in Scotland has been Robin Chittenden 193. Carrion Crows Corvus corone, London, November 1995.

276 British Birds 96 • June 2003 • 274-290 Taxonomic status of Carrion and Hooded Crows mapped twice in the last 40 years, during field- extensively studied in the 1980s and 1990s. work for the two Breeding Bird Atlases (Shar- Italian researchers (e.g. Rolando 1993) recog- rock 1976; Gibbons et al. 1993). Additional nised a fifth category of hybrid, between observations were incorporated for the Euro- Picozzi’s 2 and 3, for birds which are interme- pean Atlas (Hagemeijer & Blair 1997), but these diate between pale and dark hybrids. The Italian are less detailed. Sharrock and Gibbons et al. studies investigated habitat, pair composition, both show that the hybrid zone is relatively vocalisations and reproductive success, and the narrow. It moved north and west during the results from these are discussed below. These twentieth century, with corone displacing cornix and other studies have shown evidence of from the lower (agricultural) land around the partial ecological segregation and slight behav- North Sea coast of eastern Scotland and ioural differentiation between cornix and restricting it to more upland areas. Relatively corone. There is strong evidence of non-random little ecological or genetic research on crows has mating in the zones of hybridisation and, in been carried out in this region, however Italy, hybrid progeny reproduce less successfully. (Picozzi 1975, 1976). Hybrids show intermediate plumages, with Habitat preference every intergradation between typical all-black There have been two studies of habitat prefer- and ‘hooded’ birds, often with black dappling ence in the Italian hybrid zone. The first (Saino on the back and chest (Richter 1958; Melde 1992) was undertaken in two areas of Piedmont 1984). Picozzi (1976) separated these into four during the winter of 1989/90. The main habitats categories, although the underlying variation is were meadows, grazed and ungrazed pastures, continuous: cereals and maize. Saino separated the birds into Carrion, Hooded and hybrids, and 1. Black crows. All-black, either pure Carrion obtained very similar results in both study Crows or hybrids indistinguishable from these. areas. Table 1 summarises the results of his 2. Dark hybrids. Hybrids which have a dark grey highly detailed analyses in terms of the relative back and/or belly, including those with a grey frequency of phenotypes in each habitat com- collar. pared with that expected if the forms were dis- tributed randomly, in an attempt to eliminate 3. Pale hybrids. Hybrids with a pale back and the problem of the relative abundance of the belly. These are distinguished from Hooded three forms. Saino found significant variation Crow by uppertail- and undertail-coverts, among the three types in their habitat use. which are concolorous with the back and These results suggest that Hooded and belly in Hooded and darker in hybrids. Carrion Crows differ slightly, yet significantly, in their habitat use. In a separate part of the 4. Grey crows. Hooded Crows or hybrids indis- same hybrid zone, however, Rolando & Laiolo tinguishable from these. (1994) obtained rather different results. They There is also a zone of hybridisation between found less difference between Hooded and corone and cornix in northern Italy, which was Carrion Crows in winter, but marked differ-

Table 1. Habitat preference of Carrion Corvus corone and Hooded Crows C. cornix, and hybrids between the two studied at two sites in northern Italy in winter. Plus (or minus) signs indicate that the form was present at a frequency above (or below) that expected if the birds had been distributed randomly across habitats. * = p<0.05 ** = p<0.01. Data from Saino (1992).

Carrion Hybrid Hooded Site A Site B Site A Site B Site A Site B Meadows + / * + / ** Ungrazed pasture Maize stubble <15 cm + / ** + / ** Maize stubble >15 cm - / * Recently seeded grass fields - / ** - / * - / ** - / ** Cereal - / ** - / ** - / ** Ploughed fields Others - / **

British Birds 96 • June 2003 • 274-290 277 Taxonomic status of Carrion and Hooded Crows

Table 2. The proportion of Carrion Corvus corone and Hooded Crows C. cornix observed feeding in a variety of habitats in northern Italy in spring/summer. * indicates that the frequencies are significantly different. Data from Rolando & Laiolo (1994).

Carrion significance Hooded Meadows 54.1% * 69.3% ‘Dunged’ meadows 20.6% * 3.8% Maize fields 0.0% * 4.1% Maize stubble 8.7% * 3.1% Plough 9.5% - 12.8% Cereal shoots 7.1% - 6.9%

ences during spring and summer. Table 2 sum- Pair composition marises the spring/summer distribution of the Mated pairs were identified during the breeding two forms in their study. Carrion Crows were season and the phenotype of both partners found significantly more frequently in maize recorded by Rolando (1993), although he could stubble and meadows treated with manure. not distinguish the sexes. In the zone of Hooded Crows, on the other hand, were signifi- hybridisation in Italy, there was clear evidence cantly more abundant in non-manured that pair composition was not random (table meadows and maize fields. Manure treatment 3). Omitting the hybrid birds, of which there comprised scattering dung across the field so are small numbers and whose mating prefer- that pieces lay on the surface. Birds selectively ences are not a priori predictable, there is clear examined these for the presence of inverte- evidence of non-random pair composition brates. Rolando & Laiolo (1994) noted that the between Hooded and Carrion Crows (or at least differences between habitats disappeared when ‘Grey’ and ‘Black’ crows) (table 4). Random pair meadows were pooled irrespective of fertiliser formation would result in far more mixed pairs treatment, and when maize crops were pooled than were observed. independent of crop state. Since the differences Interestingly, fewer of the intermediate forms are significant, it is questionable whether backcrossed to the black (9) than to the grey pooling these data is valid, and evidence of dif- (24) crows. This would be expected to be equal, ferential utilisation of habitat between the two and the difference is significant (p<0.01). This taxa still remains. In summary, the results of suggests that the visual appearance of the birds both of these studies suggest that Hooded Crow is important in mate selection. Hybrid individ- is a bird of poorer quality land, a finding which uals preferentially choose ‘grey’ rather than is supported by its displacement by Carrion ‘black’ crows as mates. Since black is, geneti- Crows in northern Scotland following changes cally, at least partially dominant (see below), in land use during the twentieth century. These grey hybrids are likely to be closer to ‘pure’ ecological differences merit further study. Hoodeds, than black hybrids are to ‘pure’ Car- /Windrush David Tipling 194. Hooded Crows Corvus cornix cornix, Finland.

278 British Birds 96 • June 2003 • 274-290 Taxonomic status of Carrion and Hooded Crows

Table 3. Composition of mated pairs of Carrion Corvus corone (‘black’) and Hooded Crows C. cornix (‘grey’), and the hybrids between them, observed during the breeding season in northern Italy. The sexes could not be differentiated so that ‘51’ indicates that there were 51 pairs comprising two black crows. Data from Rolando (1993).

Black B/H Hybrid G/H Grey Black 51 B/H 5 0 Hybrid 2 0 0 G/H 2 0 0 1 Grey 16 9 3 12 41

Table 4. This table shows the data from table 3, omitting birds of mixed plumage, rearranged to show a deficiency of pairs comprising one black (Carrion) and one grey (Hooded) crow. Data from Rolando (1993).

Pair composition Black × Black Black × Grey Grey × Grey Total Observed 51 16 41 108 χ2 = 53.2, 1 df Expected (random) 32.2 53.5 22.2 107.9 p<0.001 rions. Since the progeny of mixed genetic com- tatively, again with fewer mixed pairs than position are selectively disadvantaged (see would be expected if mating was simply random. below), hybrids which choose a Hooded Crow Saino (1992) also recorded the composition as a mate will maximise the proportion of of groups of crows in the Italian hybrid zone, ‘Hooded’ genes (~75%), and consequently the away from the nest. He examined the composi- fitness, of their progeny. Those that choose a tion of groups of two and of more than two black crow may have a partner with a heteroge- birds separately, and in both he found highly neous genetic constitution, and consequently significant evidence for a lack of mixing com- may produce progeny of lower fitness. pared with what would be expected at random. Similar evidence for non-random pairing was He extended this analysis to allow for differ- reported in the zone of hybridisation between ences in the distribution of phenotypes across cornix and orientalis in central Siberia (Kryukov habitats. Even when looking only at meadow- & Blinov 1989; Blinov & Kryukov 1992), and land, there was still evidence that the associa- between cornix and corone in Schleswig-Hol- tion of phenotypes was not random, with a stein, Germany (Risch & Andersen 1998). In significant deficiency of mixed flocks of grey Siberia, Kryukov and Blinov found an excess of and black crows (table 5). homotypic (like-with-like) pairs, and a defi- ciency of mixed pairs. Risch & Andersen (1998) Composition of populations within the hybrid recorded the phenotypes of pairs of birds on zone the island of Amrum, Germany. The island is in When the individual colour patterns of the a region on the corone side of the hybrid zone, breeding birds reported by Rolando (1993) so that grey phenotypes were rare. They found within the Italian hybrid zone are summed, that hybrids and black crows were paired assor- there is evidence of a deficiency of hybrid

Table 5. The composition of flocks of Carrion Corvus corone and Hooded Crows C. cornix outside the breeding season in meadows in northern Italy, compared with that expected if the birds associated at random. Observations are separated into parties of two and more than two individuals. Data from Saino (1992).

Groups of two Groups of more than two Observed Expected Observed Expected Carrion-Carrion 38 22.5 - - Hooded-Hooded 61 45.5 - - ‘Homotypic’ (all the same) - - 46 29.7 Mixed 33 64.0 135 151.3 Total 132 132.0 181 181.0 χ2 = 31.0, 2 df, p<0.001 χ2 = 10.7, 1 df, p<0.005

British Birds 96 • June 2003 • 274-290 279 Taxonomic status of Carrion and Hooded Crows phenotypes (table 6). genetics of plumage pattern in crows, it is diffi- Little is known of the genetics of pigmenta- cult to predict the likely array of phenotypes in tion in crows, but Picozzi (1976) recorded the a hybrid zone where random mating occurs, but phenotypes of some parents and their broods in there can be little doubt that, under any genetic the hybrid zone in northern Scotland (table 7). model, intermediates should be more abun- These nests came from the southern (i.e. dant. The ‘deficiency’ of hybrids in Rolando’s Carrion Crow) end of the hybrid zone, so grey (1993) data clearly suggests non-random birds were sufficiently rare that pure Hooded mating and/or some selective disadvantage to Crows were not present in his study area. He the hybrid forms. recorded 36 nests over three years, and cate- gorised the birds by phenotype as black Aspects of reproductive biology (‘Carrion’), dark hybrid, pale hybrid and grey Saino & Villa (1992) recorded the breeding per- (‘Hooded’; see above). The majority of pale formance of pairs of crows across the hybrid hybrid nestlings occurred in broods where at zone in northern Italy over a number of years. least one adult was also pale. They recorded the phenotype of each bird as Although based on relatively little data, these ‘Carrion Crow’ (all black), ‘Hooded Crow’ results support the idea that there is genetic (black head, throat, breast, wings and tail) or control of plumage colour, with black being ‘hybrid’ (intermediate between the two in dominant over grey. This conforms with other amount or extent of black feathers in the body studies of plumage polymorphism where the plumage). They separated the area into three darker phenotype usually seems to be domi- parts: the hybrid zone, and the areas on either nant, for example in Snow Geese Anser side where only Hooded or Carrion Crows were caerulescens (Cooke et al. 1995), Arctic Skuas found (i.e. ‘in allopatry’). They recorded the Stercorarius parasiticus (O’Donald 1983) and composition of each pair, and (where possible) Rock Dove Columba livia (Murton et al. 1973). the clutch size and the number of chicks which Recently, Theron et al. (2001) have shown that fledged successfully (table 8). melanic plumage in Bananaquits Coereba flaveola (which is also dominant over non- a. Clutch size of Carrion and Hooded Crows melanic plumage; Wunderle 1981) is perfectly in northern Italy associated with a mutation in the Saino & Villa (1992) found no evidence of vari- melanocortin-1-receptor gene. It is not possible ation in clutch size across years, so they pooled to be more precise about the finer details of their data relating to clutch size of individual genetic control in crows without information pairs (table 8, a). It is possible that, in some from controlled crosses, but the broad spread of instances, the same pairs were recorded in suc- intermediate patterns among hybrids suggests a cessive years, so the data may not be entirely degree of multi-gene control, rather than the independent, but the effects of this are likely to simple pattern of inheritance in skuas, be slight. Although there is evidence that hybrid Bananaquits, etc. females laid fewer eggs than either Carrion or With this rather limited knowledge of the Hooded Crows in the hybrid zone, the differ-

Table 6. The number of crows Corvus corone of five different colour classes observed in a hybrid zone in northern Italy.The classes are described in the text. Data from Rolando (1993).

Black Black/hybrid Hybrid Grey/hybrid Grey Total 127 14 5 16 122 284

Table 7. The relationship between the plumage colour of parents and their progeny at a series of nests of crows Corvus corone in the hybrid zone in northern Scotland. Note the deficiency of pale progeny in the nests of dark parents. Data from Picozzi (1976).

Dark/black progeny Pale/grey progeny Total Both adults dark/black 92 5 97 One pale/grey (hybrid) adult 16 18 34 χ2 = 39.7 1 df Total 108 23 131 p<0.005

280 British Birds 96 • June 2003 • 274-290 Taxonomic status of Carrion and Hooded Crows

Table 8. Five reproductive traits related to the adult plumage type across a hybrid zone of Carrion Corvus corone and Hooded Crows C. cornix in northern Italy comprising (a) clutch size; (b) egg volume (in ml); number of chicks reared (c) per female and (d) per male; and (e) probability of chick survival relative to female parent. Data show: mean (standard error) sample size.

Trait CC allopatry CC sympatry Hybrids HC sympatry HC allopatry Ref a Clutch size 4.4 (0.09) 118 4.7 (0.13) 41 4.2 (0.20) 29 4.5 (0.09) 88 4.5 (0.09) 114 1 b Egg volume 18.7 (0.08) 447 18.2 (0.12) 163 17.3 (0.21) 104 17.9 (0.10) 334 17.7 (0.09) 459 2 c Chicks/female 3.1 (0.11) 91 2.5 (0.19) 37 1.6 (0.28) 24 2.6 (0.13) 89 2.6 (0.12) 118 1 d Chicks/male 3.1 (0.11) 91 2.3 (0.20) 41 2.7 (0.33) 20 2.5 (0.13) 84 2.6 (0.12) 118 1 e Survival/female 0.71 (0.027) 75 0.53 (0.045) 35 0.38 (0.067) 24 0.57 (0.030) 81 0.56 (0.027) 101 2 References: 1 Saino & Villa (1992); 2 Saino & Bolzern (1992) ences are not statistically significant. The cantly fewer young than Hooded Crows inside ‘hybrid’ category will be heterogeneous, or outside the hybrid zone (table 8, c). including birds recorded by Rolando (1993) as Male Carrion Crows outside the hybrid zone pale or dark hybrids. It is interesting to note reared significantly more young than those that the variance of clutch size for hybrids is inside the zone, and significantly more than higher than that for Carrion or Hooded Crows, Hooded Crows in either area, although other either in allopatry or in the hybrid zone. ‘Vari- comparisons were not significant (table 8, d). In ance’ is an estimate of variability which, in a both sexes, hybrids showed a higher variance in trait such as this, can be genetic, environmental brood size than did either parental form. or (usually) a combination of the two. In a similar study, Saino & Bolzern (1992) reported chick survival in another part of the b. Egg volume of Carrion and Hooded Crows hybrid zone in Italy (table 8, e). They estimated in northern Italy the proportion of the clutch that survived to Using essentially the same criteria, Saino & fledge, which is effectively the probability of Bolzern (1992) showed that, unlike clutch size, each egg in the nest becoming a fledged bird. egg volume varies significantly among pheno- They related this back to the phenotype of the types (table 8, b). They found evidence of female parent, and again examined the areas bimodality in hybrids and (again) the variance separately across the hybrid zone. They found is higher in this category. that the fledging success was significantly higher Comparing egg volume by phenotype shows for Carrion Crows in allopatry; no other signifi- that Carrion Crows outside the hybrid zone (in cant differences were found, although the value allopatry) laid significantly larger eggs than for hybrids was lower than for the other cate- those in the hybrid zone. Furthermore, inside gories. Yet again, the variance was higher for the hybrid zone, Carrion Crows laid signifi- hybrids. Interestingly, Saino & Villa (1992) cantly larger eggs than Hooded Crows, and described, albeit anecdotally, the presence of both Hooded and Carrion Crows laid signifi- abnormal eggs in five nests of hybrid females. cantly larger eggs than the hybrids. As with These included unusual, or even a complete clutch size, the variance of the egg volume of lack of pigment in the shells. They also reported hybrids is greater than for either parental form, that shells were brittle, such that most of them whether inside or outside the hybrid zone. broke during incubation and thus failed to hatch. Nothing akin to this was found among c. Reproductive success of Carrion and Hooded or Carrion Crow pairs nesting in the Hooded Crows in northern Italy same hybrid zone. The sample sizes concerned Saino & Villa (1992) reported the number of were too small for these differences to be statis- chicks which were reared to ‘about the age of tically significant. fledging’ in the three areas. They found that female Carrion Crows outside the hybrid zone Vocalisations reared significantly more young than either The acoustics of birds may give an indication of hybrids or Hooded Crows, whether inside or their phylogenetic relationships. In particular, outside the hybrid zone. Hybrids reared signifi- playback experiments can assist in the clarifica-

British Birds 96 • June 2003 • 274-290 281 Taxonomic status of Carrion and Hooded Crows

tion of taxonomic status (e.g. Irwin et al. 2001). work has involved birds from Siberia in the Playback studies have not been undertaken with hybrid zone between Hooded Crows of the race Hooded and Carrion Crows, but sonogram cornix and Carrion Crows of the race orientalis. analyses have been reported by Palestrini & Chelomina et al. (1991) examined restriction Rolando (1996) examined the characteristics of fragment variation in nuclear DNA from a call comprising a short series of ‘caws’ sepa- Hooded and Carrion Crows and their hybrids. rated by short intervals. These were analysed by A subsequent investigation (Kryukov et al. dissection of the sonograms, recording caw 1992) extended this to include an elec- duration, lowest and highest frequency, har- trophoretic investigation of serum proteins. monic duration and total frequency range. Neither they nor Saino et al. (1992) discovered These were then examined both individually any major differentiation between the parental and by discriminant function analysis (DFA). forms, although Kryukov et al. (1992) reported Data were collected from an area where Hooded that ‘new variants’ were identified among the and Carrion Crows occurred in sympatry, but hybrids in albumin, post-transferrin and insufficient recordings were available from esterase. It is, however, possible that these also hybrids for these to be included in the analysis. occur among the parental forms but that the Palestrini & Rolando found significant differ- samples were simply too small to detect them. A ences in five of the six parameters, and DFA further study, based upon DNA fingerprinting permitted Hooded and Carrion Crows to be (Ufyrkina et al. 1995), found considerable vari- differentiated on their acoustic characters. They ation among 15 Hooded and Carrion Crows suggested that these differences might be eco- and hybrids from the area of sympatry in logical or geographic rather than taxonomic, Siberia. There were, however, no diagnostic but presented no evidence in support of this. In DNA fragments, and the genetic structure of some species, acoustics have been shown to the hybrid population was consonant with pan- covary with body size (e.g. larger birds having mixia, despite field evidence of assortative deeper voices); there was no evidence of this for mating. Once again, however, the sample sizes the crows, either within or between taxa. There were too small for detailed analysis to be pos- is also no evidence as to whether the two taxa sible. Chelomina et al. (1995) claimed to find themselves can differentiate between their calls. further evidence of difference in patterns gener- ated using a slightly different technology. They Molecular studies suggested that Hooded Crows (cornix) from The molecular research undertaken into this Novosibirsk and Large-billed Crows C. group has been limited and piecemeal. Most macrorhynchos from Sakhalin were more similar /Windrush George McCarthy 195. Carrion Crow Corvus corone.

282 British Birds 96 • June 2003 • 274-290 Taxonomic status of Carrion and Hooded Crows /Windrush Tom Ennis Tom 196. Hybrid Carrion Corvus corone x Hooded Crow C. cornix, Bangor, Co. Down, February 1991. to each other than either was to Carrion Crows loop’,is required. (orientalis) from Sakhalin. Although this sug- To some extent, this has been provided by gests that Hooded and Carrion Crows might Kryukov & Odati (2000) in a phylogenetic have diverged genetically, the sample sizes again analysis of various species of corvid, using are small and the data hard to interpret. about 1000 bp of cytochrome b, although again A more extensive study by Kryukov & Suzuki the number of individuals was limited. In a (2000) examined the DNA sequence of a short study which included 12 other species, they fragment (336 bp) of the cytochrome b gene in analysed one each of corone (Paris), orientalis a series of individuals from Paris, France, to (southern Primorskiy, Russia), orientalis Hokkaido, Japan. They found uniformity in (Sakhalin) and cornix (Moscow). Their analysis sequence from France (corone) across eastern indicates that corone and cornix are sister-taxa, Europe and western Siberia (cornix) and but that the two orientalis samples are not: the through to northern Sakhalin (orientalis). Inter- Sakhalin sequence clusters with corone/cornix, population differentiation was between 0 and but that from Primorye is more different. 2.5%, although this is difficult to assess since ‘Carrion Crows’ from Japan and adjacent parts they do not give detailed results. Birds from of eastern Asia may be genetically divergent southern Sakhalin, from the far southeast of from the rest of the taxon ‘orientalis’, b u t m u c h Siberia (Primorye) and from northern Japan more research is needed to resolve this. showed similarly limited differentiation, although (again) the extent of this is not Diagnosability entirely clear from their data. Nonetheless, The black and grey forms of adult C. Kryukov & Suzuki reported that birds from the corone/cornix are documented thoroughly in two regions (southern Sakhalin/Japan and the Cramp & Perrins (1994). In summary, corone is rest of Eurasia) are strikingly different in their completely black, and the grey forms of the DNA sequences. They found 4-8 substitutions cornix group have the head, tail and wings (in c. 330 bases) between the birds on either black, while the nape, sides of neck, back and side of a divide in central Sakhalin, although shorter uppertail-coverts are ash-grey. In juv- this is based upon fewer than ten individuals. It enile corone, the entire body is dull, sooty black, would be preferable to see more individuals in apart from the head, which has a slight oily the analysis (especially since they also suggested gloss. The mantle, scapulars, tail-coverts and that at least one sequence was found on both median coverts have glossy purplish-black tips, sides of the divide). A more detailed analysis, and light grey feather-bases may be visible, involving a longer sequence of the cytochrome especially on the mantle and rump. The under- b gene or a more variable region such as the ‘D- parts are matt black, although again grey

British Birds 96 • June 2003 • 274-290 283 Taxonomic status of Carrion and Hooded Crows Nick Picozzi 197. Hybrid Carrion Corvus corone x Hooded Crow C. cornix, Northeast Scotland, 1967.This photo shows a distinctive hybrid which somewhat resembles Collared Crow C. torquatus.This type is not common in the Northeast Scotland hybrid zone.

feather-bases may be visible. The remiges, rec- zones and clinal variation. In the latter, when trices, greater wing-coverts and alula are black, two morphologically differentiated taxa come less intensively glossed than in adults. The head, into contact, there may be a smooth and clear tail and wing of juvenile cornix are similar to transition from one through to the other. Such those of corone, but the lesser and median a situation is typified by a gradual change in the upper- and underwing-coverts are grey, with phenotypic composition of the populations dull black towards the tips. The body is grey, from one extreme to the other, as exemplified tinged brown on the upperparts, and appears by many species conforming to Bergmann’s or less uniform than in adults. The remainder of Allen’s biogeographic rules. A hybrid zone is the underparts are light brownish-grey. clearly distinct when populations across the zone of transition include both parental forms Discussion as well as hybrids. Helbig et al. (2002) discussed operational cri- teria for the assignment of specific rank to avian Diagnosability taxa. They recognised that populations will ‘Pure’ Hooded and Carrion Crows are diagnos- remain distinct only if they are reproductively able in both adult and juvenile plumages, but isolated to the extent that their gene pools do hybrids show varying degrees of intermediacy. not merge. Such populations can be regarded as It could be argued that hybridising taxa can following separate evolutionary trajectories, in never be 100% diagnosable since there will line with the Evolutionary Species Concept always be intermediates which do not meet the (Mayden 1997) or General Lineage Concept (de diagnostic criteria of either parental taxon. The Queiroz 1998). Thus, Carrion and Hooded hybrid zones of crows are somewhat mosaic, Crows should be regarded as separate species if but it seems that many local populations within barriers to gene flow can be demonstrated. these zones contain not only hybrids, but also These barriers might be prezygotic, such as dif- more or less of the ‘pure’ phenotypes, or at least ferences in courtship behaviour or species birds which are indistinguishable from these, recognition through vocalisations. Alternatively, found outside the hybrid zone. Consequently, they might be postzygotic, for example if they differ from clines where the populations hybrids show reduced fitness. Sympatric or change progressively from one phenotype to the parapatric taxa are easier to deal with since the other, and are more akin to the ‘bimodal hybrid frequency and consequences of hybridisation zone’ of Jiggins & Mallet (2000). In the light of can sometimes be recorded directly. Care is this pattern of variation within a hybrid zone, needed, however, to distinguish between hybrid the Biological Species Concept would regard

284 British Birds 96 • June 2003 • 274-290 Taxonomic status of Carrion and Hooded Crows

Hooded and Carrion Crows as semispecies, ferentiating markers, such as microsatellites or whereas the Evolutionary and Phylogenetic the mitochondrial control region, might be Species Concepts would treat them as specifi- more appropriate to such relatively recent cally distinct. divergences. The molecular studies are insufficient to support this, however. Nonetheless, the markers Ecological/behavioural differentiation which have been used (enzymes, DNA finger- Taxa can be regarded as specifically distinct if printing and the mitochondrial cytochrome b there are intrinsic barriers to gene flow which gene) are less sensitive to slight genetic differen- ensure that the gene pools do not merge. There tiation, and a lack of difference across the is strong evidence for prezygotic barriers hybrid zone is unsurprising. More rapidly dif- between Hooded and Carrion Crows, at least in Nick Picozzi Nick Picozzi 198 & 199. A range of hybrid crows Carrion Corvus corone x Hooded Crow C. cornix, Northeast Scotland, 1967. Although rather grisley, these two photos illustrate the variability of hybrid types in the Northeast Scotland hybrid zone, grading from black to grey.The dark hybrids in particular are subtle, and only close views reveal the evidence of mixed parentage.

British Birds 96 • June 2003 • 274-290 285 Taxonomic status of Carrion and Hooded Crows northern Italy. Although sympatric, they show tical differences in clutch size, hybrid females partial ecological separation at some times of laid fewer eggs than did either ‘parental’ form in the year. In two separate valley systems in the the hybrid zone, and the variance of the hybrids region, there were consistent and statistically was much higher. Egg volume differed signifi- significant differences in distribution between cantly among phenotypes. Hybrid crows laid meadows and maize fields. There were differ- smaller eggs than did either parental form in ences in flock composition, with a deficiency of the hybrid zone. There was no difference mixed flocks even within the same habitat between Hooded Crows inside and outside the (meadow). Finally, pair formation was distinctly zone, but Carrion Crow eggs were significantly non-random; in three different studies, fewer smaller inside the hybrid zone, compared with mixed pairs were recorded than would be those outside it. If egg volume reflects the pro- expected given random mating patterns. In one vision of nutrients for the developing embryo, study (Rolando 1993), birds of intermediate these differences might have considerable bio- plumage mated preferentially with the grey logical significance. These results are sum- phenotype rather than with the black. marised in table 9, where each character is We have argued that plumage colour is under ranked from 1 to 5 across the five categories of genetic control. Consequently, the positive crow. For example, clutch size is highest for assortative mating which Rolando demon- Carrion Crows inside the hybrid zone (rank = strated would in itself produce a deficiency of 1) and lowest for hybrids (rank = 5). intermediate phenotypes in the population, There is a degree of consistency in these even without differences in fitness. Theoretical rankings across fitness parameters which can be population genetics (e.g. Crow & Kimura 1970) analysed statistically using Kendall’s coefficient tells us that positive assortative mating leads to of concordance (W). Carrion Crows outside the a steady decrease in heterozygosity until an hybrid zone tend to have the highest fitness and equilibrium is reached which depends upon the hybrids the lowest, although overall the result is strength of the assortment. If the hybrids have not significant (W = 0.40, p>0.05). The data reduced fitness, then any genetically determined are, however, slightly heterogeneous: four of the behavioural attribute which reduces the likeli- parameters are essentially ‘female’ – clutch size hood of mixed pairings will be selectively and egg volume are both maternal characters, advantageous. Such ‘isolating mechanisms’ will and two of the measures of chick survivals be favoured since progeny resulting from (from different studies) are related to the phe- homotypic pairs will survive better and carry notype of the mother. In addition, of course, any genes which caused the positive assortative there is non-random mating between the sexes mating in the first place (e.g. Butlin 1989). so the phenotypes of the parents are correlated. If we omit the data relating to paternal pheno- Fitness of the hybrids type, we have four aspects of female fitness, and The Italian studies examined several aspects of these are significantly concordant (W = 0.59, reproductive success, including clutch size, egg p<0.05). Female Carrion Crows outside the volume, number of chicks raised to fledging, hybrid zone have the highest fitness; hybrids and the probability of survival to fledging of (which live inside the hybrid zone) have the individual eggs. Although there were no statis- lowest.

Table 9. This table shows the data from each row of table 8 converted to ranks: 1 = largest value, 5 = smallest.The mean rank for all five variables is given, and also the mean rank based upon the female parent only (i.e. rows a, b, c and e). See text for details.

Trait CC allopatry CC sympatry Hybrids HC sympatry HC allopatry a Clutch size 4 1 5 2.5 2.5 b Egg volume 1 2 5 4 3 c Chicks/female 1 4 5 2.5 2.5 d Chicks/male 1 5 2 4 3 e Survival/female 1 4 5 2 3 Mean Rank 1.6 3.2 4.4 2.8 3.0 Females only 1.75 2.75 5.0 2.5 3.0

286 British Birds 96 • June 2003 • 274-290 Taxonomic status of Carrion and Hooded Crows Hugh Harrop 200. Hooded Crow Corvus cornix cornix, Maywick, Shetland, September 2002.

Ranking of variances Vetukhiv 1956), and has been explained as We can also examine the variation in reproduc- follows. Populations become adapted to the tive characters within the five groups of birds area which they inhabit, owing to natural selec- (table 10). The best estimate of this is the vari- tion. One form (allele) of a gene (A) is favoured ance, which reflects the variability of individ- in locality LA and spreads through that popula- uals within a sample, and can be determined tion, and another allele (B) in locality LB from the data in the original papers. These also spreads there. This may happen for several (or show a striking consistency (W = 0.77, p<0.01), many) genes so that the two populations a result which is not affected by omitting the diverge in genetic structure. But genes do not ‘male’ character. In general, hybrids have the act in isolation, and the effect of replacing A highest variance (i.e. are most variable), fol- with B may have secondary effects upon other lowed by Hooded Crows outside the hybrid genes functioning elsewhere in the animal’s zone, Hooded Crows inside the hybrid zone are body. These genes will also be selected, so that next, and the least variable (most uniform) are the entire genetic structure of the population Carrion Crows, both inside and outside the evolves to produce a harmonious array of genes hybrid zone. and alleles (a ‘coadapted gene complex’) which It has long been known (e.g. Dobzhansky interact to produce a functioning organism, 1951) that crosses of individuals from different well adapted to exist in that place. Crossing populations often show differences in ‘fitness’ individuals from different populations will (fecundity, fertility, longevity, etc.) compared result in ‘hybrid’ progeny and cause the coad- with crosses within the same population. This apted complexes to begin to break up. In suc- has been demonstrated experimentally (e.g. cessive generations, progeny are produced

Table 10. This table shows the variances of the measures given in table 8 and the ranked values (in bold) for each row. Again, the mean rank, and the mean based upon female parents only are given. See text for details.

Trait CC allopatry CC sympatry Hybrids HC sympatry HC allopatry Clutch size 0.956 2 0.693 5 1.160 1 0.713 4 0.923 3 Egg volume 2.861 4 2.347 5 4.586 1 3.340 3 3.718 2 Chicks/female 1.101 5 1.336 4 1.882 1 1.504 3 1.699 2 Chicks/male 1.101 5 1.640 3 2.178 1 1.420 4 1.699 2 Survival/female 0.0547 5 0.0709 4 0.1077 1 0.0729 3 0.0736 2 Mean Rank 4.2 4.2 1.0 3.6 2.2 Female only 4.0 4.2 1.0 3.25 2.25

British Birds 96 • June 2003 • 274-290 287 Taxonomic status of Carrion and Hooded Crows

which contain varying proportions of alleles a genetic contribution from a Carrion Crow from each parental population, to give ancestor. It is, therefore, perhaps not surprising increased variability. This disruption of the that black crows inside the hybrid zone will genetic structure in subsequent generations has reveal a pattern of fitness intermediate between been shown experimentally to have profoundly that of ‘pure’ Carrion Crows and undoubted debilitating effects upon fitness. The situation hybrids. These patterns also lend support to the which we observe in the crows is akin to this. fitness differences observed between hybrids Hybrids are much more variable, and often and parental forms. These again presumably inferior – for example in clutch size, shell thick- reflect genetic differentiation between Hooded ness and chick survival. It is strong evidence and Carrion Crows on either side of the hybrid that the genetic structure of Hooded and zone. Where these birds come into contact, Carrion Crows differs by more than just the there is some hybridisation. The crows which colour of their feathers, and the increased vari- are produced are likely to be intermediate in ance within the area of overlap indicates a phenotype. The genetic dominance of the black hybrid zone rather than a cline. alleles will mean that hybrid birds will tend In the case of the crows, there are few data towards the Carrion (‘black’) Crow phenotype, relating to the post-fledging survival of yet they contain alleles from both populations. nestlings, but there is certainly evidence of dif- Whether the reduced viability of hybrid and ferences prior to fledging. In separate studies, ‘backcross’ crows extends beyond the nestling hybrid females have been shown to lay fewer stage is not known. There was, however, a defi- and smaller eggs, and to rear fewer progeny to ciency of intermediate phenotypes among the fledging than the parental forms. Although adults recorded in the hybrid zone by Rolando there was little difference in chick survival (1993). Perhaps it is not surprising that non- between Hooded (i.e. ‘grey’) Crows inside and random mating has evolved in these regions. outside the hybrid zone, it is striking that, in Hybrid progeny are selectively disadvantaged, these studies, Carrion (i.e. ‘black’) Crows inside so any behavioural characters which result in the hybrid zone performed less well than assortative mating will spread in the hybrid outside. If the genetic model suggested above is zone (Butlin 1989). broadly correct, and black is at least partly dominant over grey in the hybrid zone, the Conclusions and recommendations black Carrion Crow phenotype might mask a The differences in plumage, the non-random variable fraction of recessive ‘Hooded’ genes. mating and the reduced fitness of hybrids are On the other hand, the recessive nature of the sufficient to regard Hooded and Carrion Crows Hooded phenotype will be less likely to conceal as specifically distinct under most species con- Robin Chittenden 201. Hooded Crow Corvus cornix pallescens, Egypt, December 1992.

288 British Birds 96 • June 2003 • 274-290 Taxonomic status of Carrion and Hooded Crows /Windrush David Tipling 202. Carrion Crow Corvus corone. cepts (Helbig et al. 2002). The differences in data at present to support these observations. vocalisations and ecology support the differen- Until further data emerge, it is perhaps wise to tiation but are, of themselves, not sufficient. err on the side of conservatism and to retain Under the Biological Species Concept, they orientalis as a subspecies of C. corone. would be treated as ‘semispecies’, because gene flow occurs constantly between them, albeit References counteracted by assortative mating and natural Bährmann, U. 1978. Eine biometrische analyse zur selection (the reduced fitness of hybrids). The morphologie der Nebelkrähe (Corvus corone cornix L.) und ihre intraspecific variation. Zool. Abh. Mus.Tierkd. deficiency of mixed pairs indicates a strong Dres. 35: 223-252. prezygotic barrier to gene flow which is rein- Blinov,V. N., & Kryukov, A. P.1992.The evolutionary forced by the fitness differences between the stability of hybrid zones: the assortment instead of the parental and hybrid phenotypes. These differ- elimination of hybrids of Hooded and Carrion Crows. Dokl. Ross. Akad. Nauk 325: 1085-1087. ences would also support taxonomic separation Butlin, R. K. 1989. Reinforcement of pre-mating isolation. of the two types under most species concepts. In: Otte, D., & Endler, J. A. (eds.), Speciation and its It is recommended that Hooded and Carrion consequences: 158-179. Sinauer, Sunderland, Mass. Chelomina, G. N., Ivanov, S.V., & Kryukov, A. P.1995. Crow be treated as separate species. As they Peculiarities of RFLP of highly repetitive DNA in crow overlap locally with only limited hybridisation, genomes. Russian J. Genetics 31: 147-151. Russian authors (e.g. Korelov et al. 1974, —, Pashkova, L. N., & Krykov, A. P.1991.The restriction analysis of the nDNA from the Corvids (Passeriformes, Stepanyan 1990) consider the two groups to be Aves). Comp. Biochem. Physiol. 100B: 601-604. full species; orientalis is then usually combined Cook, A. 1975. Changes in the Carrion/Hooded Crow with the corone group into a single species C. hybrid zone and the possible importance of climate. corone, which comprises two races. Nominate Bird Study 22: 165-168. Cooke, F., Rockwell, R. F., & Lank, D. B. 1995. The Snow corone and orientalis may, however, be less Geese of La Perouse Bay. OUP,Oxford. closely related to each other than nominate Cramp, S., & Perrins, C. M. (eds.) 1994. The Birds of the corone is to cornix (Eck 1984); and the limited Western Palearctic. Vol. 8. OUP,Oxford. Crow, J. F., & Kimura, M. 1970. An Introduction to Population molecular results from eastern Asia (Kryukov & Genetics Theory. Harper & Row, New York. Odati 2000; Kryukov & Suzuki 2000) lend some de Queiroz, K. 1998.The general lineage concept of support to this. The form capellanus, which in species, species criteria, and the process of speciation: a some structural characters is closer to orientalis conceptual unification and terminological recommendations. In: Howard, D. J., & Berlocher, S. H. than to cornix, may form a fourth group in the (eds.), Endless forms: species and speciation: 57-75. OUP, corone complex (Cramp & Perrins 1994), while Oxford. Collared Crow C. torquatus (synonym C. pec- Dobzhansky,T. 1951. Genetics and the origin of species. Third edition. Columbia University Press, New York. toralis) from China, which is similar to orien- Dybbro,T. 1976. De danske ynglefugles udbredelse. Dansk talis, may form a fifth. There are, however, few Ornitologisk Forening, Copenhagen.

British Birds 96 • June 2003 • 274-290 289 Taxonomic status of Carrion and Hooded Crows

Eck, S. 1984. Katalog der ornithologischen Sammlung Dr Udo Picozzi, N. 1975. A study of the Carrion/Hooded Crow in Bährmanns. Zoo.Abh. Staatl. Mus.Tierkde. Dresden 39: 71-98. northeast Scotland. Brit. Birds 68: 409-419. Gibbons, D.W., Reid, J. B., & Chapman, R. A. 1993. The New — 1976. Hybridisation of Carrion and Hooded Crows Atlas of Breeding Birds in Britain and Ireland: 1988-1991. Corvus c. corone and C. c. cornix in northeastern Poyser, London. Scotland. Ibis 118: 254-257. Goodwin, D. 1986. Crows of the World. Brit. Mus. Nat. Hist., Richter, H. 1958. Untersuchungen an einer London. Aaskrähenpopulation in Mittelmecklenburg. Abh. Berl. Hagemeijer,W. J. M., & Blair, M. J. 1997. The EBCC Atlas of Staatl.Mus.Tierkde. Dresden 23: 219-240. European Breeding Birds. Poyser, London. Risch, M., & Andersen, L. 1998. Selektive partnerwahl der Helbig, A. J., Knox, A. G., Parkin, D.T., Sangster, G., & Aaskrähe (Corvus corone) in der hybridisierungszone Collinson, M. 2002. Guidelines for assigning species von Rabenkrähe (C. c. corone) and Nebelkrähe (C. c. rank. Ibis 144: 518-525. cornix). J. Orn. 139: 173-177. Irwin, D. E., Alström, P., Olsson, U., & Benowitz-Fredericks, Rolando, A. 1993. A study on the hybridisation between Z. M. 2001. Cryptic species in the genus Phylloscopus Carrion and Hooded Crow in northwest Italy. Ornis (Old World leaf warblers). Ibis 143: 233-247. Scand. 24: 80-83. Jiggins, C. D., & Mallet, J. 2000. Bimodal hybrid zones and — & Laiolo, P.1994. Habitat selection of Hooded and speciation. Trends in Ecol. & Evol. 15: 250-255. Carrion Crows in the Alpine hybrid zone. Ardea 82: Korelov, M. N., Kuz’mina, M. A., Gavrilov, E. I., Kovshar’, A. F., 193-200. Gavrin,V. F., & Borodikhin, I. F. 1974. Birds of Kazakhstan. Saino, N. 1992. Selection of foraging habitat and flocking Vol. 5. Alma Ata. by Crow Corvus corone phenotypes in a hybrid zone. Kryukov, A. P.,& Blinov,V. N. 1989. Interaction of Hooded Ornis Scand. 23: 111-120. and Carrion Crows (Corvus cornix L. and C. corone L.) — & Bolzern, A. M. 1992. Egg volume, chick growth and in the zone of sympatry and hybridisation: is there survival across a Carrion/Hooded Crow hybrid zone. selection against hybrids? Zh. Obshch. Biol. 50: 128-135. Boll. Zool. 59: 407-415. — & Odati, S. 2000. Phylogenetic relationships within the — & de Bernardi, F. 1994. Geographic variation in size and Corvine assemblage (Aves, Corvidae) based on partial sexual dimorphism across a hybrid zone between sequencing of the mitochondrial DNA cytochrome b Carrion Crows (Corvus corone corone) and Hooded gene. Russian J. Genetics 36: 1054-1060. Crows (C. c. cornix). Can. J. Zool. 72: 1543-1550. — & Suzuki, H. 2000. Phylogeography of Carrion, Hooded — & Villa, S. 1992. Pair composition and reproductive and Jungle Crows (Aves, Corvidae) inferred from success across a hybrid zone of Carrion Crows and partial sequencing of the mitochondrial DNA Hooded Crows. Auk 109: 543-555. cytochrome b gene. Russian J. Genetics 36: 922-929. —, Lorenzini, R., Fusco, G., & Randi. E. 1992. Genetic —, Ufyrkina, O.V., & Chelomina, G. N. 1992. An analysis of variability in a hybrid zone between Carrion and the genomes of crows (Corvidae, Passeriformes) from Hooded Crows (Corvus corone corone and C. c. cornix) a zone of overlapping areas and hybridisation. Genetika in north-western Italy. Biochem. Syst. & Ecol. 20: 605-613. 28: 136-140. Sharrock, J.T. R. (ed.) 1976. The Atlas of Breeding Birds in Madge, S., & Burn, H. 1994. Crows and Jays. Christopher Britain and Ireland. Poyser, Calton. Helm, London. Stepanyan, L. S. 1990. Conspectus of the ornithological fauna Mayden, R. L. 1997. A hierarchy of species concepts: the of the USSR. Nauka, Moscow. denouement in the saga of the species problem. In: Theron, E., Hawkins, K., Bermingham, E., Ricklefs, R. E., & Claridge, M. F., Dawah, H. A., & Wilson, M. R. (eds.), Mundy, N. I. 2001.The molecular basis of an avian Species:The Units of Biodiversity: 381-424. Chapman & plumage polymorphism in the wild: a melanocortin-l- Hall, London. receptor point mutation is perfectly associated with the Meinertzhagen, R. 1926. Introduction to a review of the melanic plumage morph of the Bananaquit, Coereba genus Corvus. Nov. Zool. 33: 89-111. flaveola. Current Biol. 11: 550-557. Meise,W. 1928. Die Verbietung der Aaskrähe Ufyrkina, O.V.,Vasil’ev,V. A., Kryukov, A. P.,& Ryskov, A. P. (Formenkreis Corvus corone L.). J. Orn. 76: 1-203. 1995. Genomic fingerprints in crows: a study of the Melde, M. 1984. Raben- und Nebelkrähe. Die neue Brehm- genetic structure of populations in the hybrid zone. Bucherei,Wittenberg, Lutherstadt. Russian. J. Genetics 31: 753-757. Murton, R. K.,Westwood, N. J., & Thearle, R. J. P.1973. Vaurie, C. 1954. Systematic notes on Palearctic birds. No. Polymorphism and the evolution of a continuous 5. Corvidae. Amer. Museum Novitates No. 1668. breeding season in the pigeon Columba livia. J. Reprod. — 1959. The Birds of the Palearctic Fauna.Witherby, London. Fert. Suppl. 19: 563-577. Vetukhiv, M. 1956. Fecundity of the genotype in local O’Donald, P.1983. The Arctic Skua: a study of the ecology populations of Drosophila pseudoobscura. Evolution 10: and evolution of a seabird. CUP,Cambridge. 139-146. Palestrini, C., & Rolando, A. 1996. Differential calls by Carrion Wunderle, J. M. 1981. An analysis of a morph ratio cline in and Hooded Crows (Corvus corone corone and C. c. the Bananaquit (Coereba flaveola) on Grenada,West cornix) in the Alpine hybrid zone. Bird Study 43: 364-370. Indies. Evolution 35: 333-344.

Prof. David T. Parkin, Institute of Genetics, University of Nottingham, Queen’s Medical Centre, Nottingham NG7 2UH; e-mail: [email protected] (corresponding author) Dr Martin Collinson, Biomedical Sciences, Institute of Medical Sciences, Foresterhill, University of Aberdeen, Aberdeen AB25 2ZD Dr Andreas J. Helbig, Universität Greifswald, Vogelwarte Hiddensee, D-18565 Kloster/Hiddensee, Germany Dr Alan G. Knox, Historic Collections, King’s College, University of Aberdeen, Aberdeen AB24 3SW George Sangster, Stevenshof 17, 2312 GM Leiden, The Netherlands

290 British Birds 96 • June 2003 • 274-290