around our heads and calling, and occasionally This is the first breeding record of European perching on rocks allowing scrutiny through a roller for Arabia. More conclusive breeding telescope. There were two or more females, one evidence will be sought in 1991 when their imF.ature male and one adult male, showing the presence at Digdaga will be monitored more diagnostic bluish hood and unspotted bright closely. chestnut mantle. Some of the birds also had projecting central tail feathers (or perhaps just Colin Richardson, P.O. Box 2825, Dubai, UAE. lack of white tips to these feathers). The juvenile was disturbed from a rock ledge, but was NB The European roller breeds throughout much of not yet able to fly. Iran, in Iraq at the head of the Arabian Gulf and probably in Jordan. Ed. A month later, on 3 July, another adult male was seen 50 km or so to the south, but apart from WHEATEARS OF THE OENANTHE LUGENS COMPLEX these records and another sighting of a probable (MOURNING WHEATEAR) IN ARABIA female in spring 1989, records of lesser kestrels in the region seem to be very few and far Mike Jennings recently (1989, Phoenix 6) between. Green (1984, Sandgrouse 6:48-50) makes expressed the view that South Arabian mourning no mention of the species. However, on the same wheatears Oenanthe "Iugentoides" should be mountain in March 1988, a group of kestrels seen considered a separate species from O.Iugens. from a distance by Paul Goriup and Peter Symens Hollom et al. (Birds of the Middle East and were possibly this species. The likelihood is North Africa, 1988, Poyser: Calton) also followed that the colony has in fact been present for at this course. In this note I aim to present least three years. evidence which suggests that it would be preferable to continue to regard South Arabian The lesser kestrel is a common summer breeding wheatears as a group of subspecies of the visitor to parts of the Mediterranean, the Middle mourning wheatear o. Iugens. East and Asia, although it has undergone declines in many countries in recent years. It is a Wheatear taxonomy has been confused by the high social falcon, inhabiting open, arid country, but lability of plumage characters in the genus and has adapted to using nest-sites in old buildings wheatears include an unusually high proportion in urban areas. (for birds) of polymorphic species, species with a range of plumage types within the same sex of a John Norton, c.o National Conservation Bureau, single subspecies. Wheatear species and 136 Kingfisher Court, Newbury, Berkshire, RG14 subspecies are differentiated by variations on 5SJ. two themes, of black-and-white or sandy plumage, but many species are polymorphic in one or more of crown, throat or belly colour. Also, the 0841 CORACIAS GARRULUS European roller degree of sexual _or age dimorphism may vary between races of single species. Characters The status of the European roller in the United involved in polymorphism or sexual dimorphism in Arab Emirates is much the same as that recorded some species are responsible for subspecific throughout he rest of the Arabian Peninsula. It differentiation in others and for distinction occurs regularly on spring and autumn passage, in between yet other full species. I recently transit during its rather extravagant seasonal reviewed the relationships of wheatears in an migration between the warm temperate zones of attempt to sort out some of this confusion (Tye, Europe and Central Asia and its African wintering A. 1989, Bonn. zool. Beitr. 40: 165-182 and come grounds. It is most likely to be found in the to the following conclusions regarding mourning Emirates from April to mid May and September to wheatears. October. Some occur in March and August, and it is exceptional to find any from late May to mid This species has more geographical races (eight) July. than any other wheatear. The races are eight isolated populations, in North and East Africa Its 1990 spring passage was healthy, with a and the Middle East, including Arabia. These number of reports of ones and twos seen races are often divided into three groups. The throughout April and early May, whilst parties of "Lugubris" group comprises three dark plumage, up to six together were reported from 10-25 April sexually dimorphic subspecies: O. Iugens at Dubai, Asab in the western desert and in the Iugubris, 0.1. vauriei and O. 1. schalowi. These expanding fodder fields of Digdaga, in the far are isolated, montane subspecies, found in north of the country. Ethiopia, Somalia and East Africa respectively. The "Iugentoides" group is what Hollom et al. Digdaga (VB28) near Ras al Khaimah has been a (1988) and Jennings (1989) refer to as "South source of ornithological discoveries in recent Arabian wheatears"; it consist of two sexually years, with European bee-eaters recorded breeding dimorphic subspecies, O. 1. Iugentoides of SW for the first time in 1989. This year the fields Arabia and 0.1. boscaweni of the extreme south. have been expanded further over the gravel The "I u gen s" group has three subspec ies : plains, and indigenous Ghaf Prosopis cineraris O. 1. Iugens breeds in the Middle East from trees have been carefully retained wherever they approximately lower Egypt to Syria, including NW stand in and around the new fields. This has Arabia, 0.1. persica breeds in Iran and created a marvellous landscape of mature trees possibly NE Arabia; these two are sexually and grassland: ideal for Indian rollers, which monomorphic. The third, O. 1. halophila of North breed in the area. Africa from Morocco to Libya, is sexually dimorphic. The "Iugens" group is partially European rollers were recorded on all visits from migratory, overlapping with other races in early April 1990, and nine birds wee found on 25 winter. May, when passage should have been well over. They seemed at home perching on irrigation spray The three subspecies groups, and even some of the heads, many of which were already occupied by individual subspecies, eg o. 1. schalowi, are Indian rollers, and surprisingly the two species appeared to be co-existing quite happily. They + + + + + + + + + + + + + + + + + + + + + + + + + were still present in June, and on 4 July, I found five birds, including at least one * Scientific names of all species included in the juvenile. This young bird had buff edges to its project and their reference numbers are given on blue wing coverts and was very pale around the Form 2: List of Breeding Birds of Arabia ­ face and throat (mainly due to creamy pale issued free to all contributors. The names of feather tips). The brown to the back was pale other species and additions to the list are given and indistinct. Nearby, a pair of adults were in full. To save space, localities mentioned in chasing in what appeared to be a half-hearted the text (except major towns) are suffixed by the display. Juvenile birds were later recorded on atlas square reference and these can be seen on several subsequent visits to the fields, which the maps in this issue. Bibliographic references are spread widely over several hundred hectares. are kept to a minimum and are given in This habitat appears to have supported several abbreviated form. All articles are attributed to pairs of European rollers this year. the Editor unless otherwise shown. vV undoubtedly at least incipient species, being As regards the South Arabian birds, they are geographically isolated and with distinctive clearly intermediates (between 1ugens and plumages. vauriei, and it is hard to justify separating them as a full species. If one did so, and However, problems arise when one tries to decide wished to remain consistent, half a dozen other how to divide mourning wheatears up into a nunilier wheatear species would have to be similarly of smaller species. The problem is caused by the split, not to mention other bird species. fact that, although O. 1. scha10wi at one extreme appears strikingly different from O. 1. ha10phi1a So, at the risk of denying Arabia an endemic at the other, these extremes are linked by a species, I sug~est retention of all eight chain of the other subspecies and along the chain populations in O. 1ugens. This does not remove there is no obvious break in characteristics; the value of recording for ABBA purposes the rather, characters change gradually or one by South Arabian 1ugentoides-boscaweni separately one. The difficulty is exacerbated by the fact from O. 1. 1ugens of NW Arabia and from O. 1. that each subspecies is an iSQlated population; persica of (perhaps) the southern shores of the one cannot apply biological species definition of Arabian Gulf. One may perhaps hope to discover lack of interbreeding. This means that, if one thereby an area of overlap where interbreeding wants to divide mourning wheatears up, it is does not occur, and which would thus prove that impossible to say where the dividing lines should they are different species. However, recorders be. For example, should the 1ugubris group be do need to beware of temporary population treated as one species, or as two (scha10wi and overlaps cased by the migrations of the 1ugens Lugubris-vauriei) or as three (0.scha10wi, O. group which may, in spring, be found in areas 1ugubris, O.vauriei)? On the other hand, the where the South Arabian birds are breeding at the Somalian vauriei is quite similar to the time. 1ugentoides South Arabian birds and should perhaps be grouped with the latter? Or perhaps Alan Tye, 2 School Lane, King's Ripton, the 1ugentoides group should be merged with 0.1. Huntingdon, Cambridgeshire, PE17 2NL, UK.