Studies of West Palearctic birds 191. Azores Bullfinch Colin J. Bibby, Trevor D. Charlton and Jaime Ramos he Azores comprises nine islands lying in mid Atlantic. Few landbirds Tbreed and all are of Palearctic origin, with the avifauna bearing a strong similarity to die species lists of Madeira and the Canary Islands. High in the mountains on a remote corner of one island lives a rather odd Bullfinch: Pyrrhula (pyrrhula) murina. Azores Bullfinches differ considerably from the other forms in western Eu­ rope. The sexes are virtually identical in colour, though males sometimes show the slightest suffusion of buffish-orange on the upper breast and ear-coverts. The general coloration is dark greyish-brown above and buffish below, similar to that of European females. The rump is buffish, rather than white, and not very distinct. Azores Bullfinches are quite large, weighing around 30 g (British Bullfinches average about 22 g in summer). Such a distinct and isolated population could merit the rank of a separate species (Knox 1989) and no doubt would if the phylogenetic taxonomists had their way. Current taxonomy offers no hard-and-fast way of recognising when isolation and differentiation have gone far enough to justify treating island forms as separate species. Larger size and duller plumage are quite frequent trends among island birds compared with their continental counterparts. Separate species or not, this is a very poorly known bird. In the nineteenth century it was a locally abundant pest of fruit orchards, with a bounty on its head. The major fruits grown were oranges, but this crop was largely aban­ doned as a result of disease in the late nineteenth century. The early collectors had no difficulty in obtaining specimens, but the Azores Bullfinch became rare and, after the 1920s, slipped into oblivion. The great explorers of die Adantic islands, Bannerman & Bannerman (1966), failed to see it on visits in 1963 and 1964. It seemed that it could be close to extinction, although it was still known and remembered by local people at that time. In the late 1970s, a small population, estimated at 30-40 pairs, was rediscovered (Le Grand 1983, 1985). Even when it was abundant, it was confined in range to a small part of the largest island, Sao Miguel, and it was there that it was rediscovered. In 1989, CJB and TDC visited the area at the east of Sao Miguel to make [Brit. Bird, 85: 677-680, December 1992] 677 678 Studies of Azores Bullfinch the first formal estimate of population size and seek to assess any conservation needs. Locally, the birds were not difficult to find in and near laurel forests in the mountains, but the area is remarkably inaccessible. The vegetation is dense scrub 1-2 m high and would be almost impossible to walk through even if the ground was level. It is not quite thick enough to walk over. In fact it is mainly the precipitous mountain slopes which retain native scrub and have not been cleared by man. These are simply too steep and unstable to walk upon. By exploring the few tracks and stream beds which allowed access, wc were able to conduct a series of point counts and estimated diat there were about 370 Bullfinches in September, which would represent about 100 breed­ ing pairs. The range appeared to encompass a fragment of native vegetation of about 500 ha. Within this area, the density of Bullfinches is quite high com­ pared with numbers found in woodland in Britain. Apart from their colour, the birds resemble our own Bullfinch in general behaviour. They are detected by their piping calls and can be very approach­ able. The individual photographed by TDC was approached to 7 m in the open with a 500-mm lens. It is relatively easy to watch them feeding, though they often take flights of several hundreds of metres. The nest has never been described, or even apparently seen, by an ornithologist. We know, however, from the presence of brood patches on females, and the appearance of juve­ niles, that they breed from June to August. This could allow time for two broods. Clutch and brood sizes are probably small because the proportion of juveniles in autumn appears to reach only about 50%. The native forest cover of the Azores is greatly fragmented, but occurs more widely than the Bullfinch, both on Sao Miguel and on other islands. The Azores has 5(5 endemic plants, many of which occur in the native forest, so these areas arc of considerable interest and value in their own right. Much has been lost to agriculture and, in more recent years, to afforestation with an exotic conifer, Japanese red cedar Cryptomeria japonka. A more alarming threat is the extent to which introduced plants have been able to invade the laurel forests. Two exotic tree species, lily-of-the-vallcy-trce Ckthra arborea and cheese- wood Pittospomm unduktum, are locally abundant. Disturbed ground is densely carpeted by Kahili ginger Hedychium gardneranum, which already covers huge areas and is likely to take over when the conifer forests are felled. Thus, the native vegetation is under great threat in spite of being in a reserve and there is little chance of its regaining previously lost ground. In 1991, JR started an ecological study to enhance the meagre knowledge of the Azores Bullfinch. The critical question seemed to be why the bird is so localised, ft appears that it does not breed in a patch of laurel forest very close to its known breeding range, though we have seen the species there in au­ tumn. It is no longer a pest, but we do not yet know whether it any longer visits fruit orchards at all. Feeding observations, so far only in the summer, have shown a diet consisting of buds and seeds of weeds or fruits. The most remarkable feature has been the prevalence of introduced plants in the diet. Only in August, of the months studied so far, have endemic plants predomi­ nated. From May to July, the flowering heads of garden knotweed Polygonum capitatum are stripped voraciously. This is an introduced perennial herb of dis­ turbed ground and is most abundant along streamsides and beside roads. It Studies of Azores Bullfinch 679 316. Adult Azores Bullfinch Pyrrhula (fiyrrhulaj murina, Azores, July 1989 (T. D. Charlton) seems very odd that a very rare bird confined to native forest should rely so heavily on an introduced weed so largely associated with human disturbance. At the moment, we can only speculate that Azores Bullfinches might be dependent on native species at a time of food scarcity in winter. We hope that further study will assist the conservation of this very rare bird, whether it be a species or merely a distinct race. The government of the Azores has been quick in designating Special Protection Areas under the EC Birds Directive. The Bullfinch area is owned and managed by the Forestry Service, which has declared a reserve, and there is considerable local interest in furthering conservation. The Azores in general is not well known to ornithologists. Apart from its Bullfinch, the archipelago holds more than half the West Palcarctic population of Roseate Terns Sterna dougallii and enormous numbers of Cory's Shearwaters Calonectris diomedea, well in excess of those reported elsewhere in the North Atlantic. Madeiran Petrels Oceanodroma castro, Bulwer's Petrels Bulweria bulwerii and Little Shearwaters Puffinus assimilis also breed; and who knows what else? Acknowledgment Assistance with English names of plants was provided by the Royal Botanic Gardens, Kew. 680 Studies of Azores Bullfinch References BANNERMAN, D., & BANNERMAN, W. M. 1966. The Birds of the Atlantic Islands. Vol. 3. Edinburgh & London. BlBBY, C. J., & CHARLTON, T. D. 1991. Observations on the San Miguel Bullfinch. Acoreana 7: 297-304. KNOX, A. 1989. Proposed changes to the Voous list. Brit. Birds 82: 119-120. LE GRAND, G. 1983. Der wiederentdeckte Azorengimpel. Wir und die Vogel 15(1): 37-38. — -, 1985. O Priolo, a ave mais interessante de Sao Miguel. A(oreana 6: 195-211. Dr Colin J. Bibby and Trevor D. Charlton, RSPB, The Lodge, Sandy, Bedfordshire SCI 9 2DL Jaime Ramos, Edward Grey Institute, Department of 2fpohgy, South Parks Road, Oxford OXI 3PS .