ay ay Abe Beye ippeetn ty ee el Ain ie P € 4 ke Heid! ; f ; i , : : pi teat syy area ee Be FL : Parcs ite’ rf : ; 5 aay ; A i i , y z Tie! det HY EER a An SE yeunyaite he PENG soap ev td V3 ee antyryih etted? x eh Al Decne gh ete it re re : eth aw Os ae er, Siete mica WY) bss ghee d - ot ; ' hy ef ry s ih ' ce i i Petit ys rte ¥ is a 4 ) 14 Pr ‘ Pee Wi i . N rh tae i : 1 ' hy dy A j Pa ly " : ne a id ; hat 14 . pier MG hs Vol. 20 No. 1 ISSN 0036 9144 Scottish Birds The Journal of the Scottish Ornithologists’ Club Editor: DrS da Prato Assisted by: Professor D Jenkins, Dr J B Nelson, Dr | Bainbridge, Dr M Marquiss and R Swann Business Editor: The Secretary SOC, 21 Regent Terrace Edinburgh EH7 5BT (tel 0131-556 6042, fax 0131 558 9947). Scottish Birds, the official journal of the Scottish Ornithologists’ Club, publishes original material relating to ornithology in Scotland. Papers and notes should be sent to The Editor, Scottish Birds, 21 Regent Terrace, Edinburgh EH7 5BT. Two issues of Scottish Birds are published each year, in June and in December. Scottish Birds is issued free to members of the Scottish Ornithologists’ Club, who also receive the quarterly newsletter Scottish Bird News, the annual Scottish Bird Report and the annual Raptor round up. These are available to Institutions at a subscription rate (1997) of £36. The Scottish Ornithologists’ Club was formed in 1936 to encourage all aspects of ornithology in Scotland. It has local branches which meet in Aberdeen, Ayr, the Borders, Dumfries, Dundee, Edinburgh, Glasgow, Inverness, New Galloway, Orkney, St Andrews, Stirling, Stranraer and Thurso, each with its own programme of field meetings and winter lectures. The Waterston Library at the Club’s headquarters at 21 Regent Terrace, Edinburgh EH7 5BT is the most comprehensive ornithological library in Scotland, and is available for reference during office hours (Monday to Friday, 0930-1630 hrs). (But please phone beforehand). SOC annual membership subscription rates Direct Debit Other methods Adult £18.00 £20.00 Family (2 adults and any children under 18) living at one address £27.00 £30.00 Junior (under 18, or student under 25) £7.00 £8.00 Pensioner/unwaged £10.00 £11.00 Pensioner Family (2 adults living at one address) £14.50 £16.00 Life £360.00 £400.00 Life Family £540.00 £600.00 All subscriptions may be paid by Direct Debit and Covenanted. Subscriptions paid by Direct Debit greatly assist the Club. Please ask for a Direct Debit form by phoning the Secretary at the above address. Published by the Scottish Ornithologists’ Club, 21 Regent Terrace, Edinburgh EH7 5BT Printed by Meigle Print, Block 11, Units 1 & 2, Tweedbank Industrial Estate, Galashiels TD1 3RS Scottish Birds (1999) 20:1-5 The breeding birds of Auskerry, Orkney, 1969-1998 RG ADAM & C J BOOTH This paper documents changes that have taken place in the seabird population of Auskerry in the last 3 decades. In addition, other breeding birds are mentioned briefly. Auskerry is a small island, approximately 1.5 x 1 km, situated on the eastern side of the Orkney archipelago, 5 km south of Stronsay. lt has a rocky coasiline with some boulder and shingle beaches. There are cliffs on the west side which reach a height of 18 metres and where Kittiwakes, Guillemots, Razorbills and the majority of Shags nest. The island has long been used for grazing sheep. The vegetation is mainly Cal//unaheath and acidic grassland with Deschampsia tussocks on both eastand westcoasts. Itwas designated an SSSI in 1997 and classified as a Special Protected Area (SPA) in June 1998. Methods Information on the breeding birds from 1969 to 1998 has been obtained from a variety of sources including the Seafarer counts in 1969 (Cramp, Bourne and Saunders 1974). AD K Ramsay made almost annual visits in the 1970s into the mid 1980s, staying for several days on each occasion. It is due to his enthusiasm for recording the birds of Auskerry that we have so much data for these years. Both authors have visited the island on a number of occasions throughout the study period and particularly from 1992 to 1998, when 3 to 4 day stays were made annually. Apart from 2 visits in May and one in June the majority of counts have taken place in July, which means that some breeding birds that failed early will have been missed. Fora number of species this means that the counts will be minima. Breeding seabirds Fulmar Fu/marus glacialis A total of 150 apparently occupied nest sites was counted in 1969; then numbers ranged from 200 in 1971 to 121 in 1974 and 150 in 1980. Almost all of the inland nesting pairs that laid from 1974 to 1980 failed and the numbers at these sites dropped from 70 to 20 pairs. There was a large increase over the whole island in 1981 to 280 pairs and this level was maintained for several years. A decline has been noted in the last 5 years and a count in July 1997 found 110 apparently occupied sites on the west coast. There were very few inland nesting pairs in 1997 and 1998. It was thought that the number of nesting pairs in 1998 was very similar to the 1969 total. Storm Petrel Hydrobates pelagicus Before surveying methods had been established, it had been speculated that the breeding population was between 1,000 and 9,999 pairs. A Survey in 1995 using a daylight playback technique found an estimated 3,613 occupied nesting sites (Wood 1996). No obvious changes have 2 RG Adam & C J Booth SB 20(1) been noticed during the study period but it is possible that sheep have collapsed some burrows. Shag Phalacrocorax aristotelis Twenty nests were counted in 1969; numbers then increased to a peak of 75 pairs in 1974 before gradually declining to 30 pairs in 1983. A further decline has occurred although numbers have remained fairly stable in the last 5 years with 21 nests in 1994, 20 in 1996, at least 16 in 1997 and 20 in 1998. Arctic Skua Stercorarius parasiticus Only one pair was recorded breeding in 1969 but 3 pairs were present in 1970 and 4, possibly 5, in 1973. Four pairs bred from 1974 to 1977, then 2 to 3 pairs were noted between 1978 and 1983. Two pairs were present in 1988 and from 1992 to 1998. Great Skua Catharacta skua Single pairs bred in 1969, 1970 and 1972. There were 2 pairs in 1973 and 1974 then only one pair annually to 1983. Single pairs have been on territory from 1992 to 1996, with a nest and eggs found in 1993. A pair was present in 1998 but there was no indication of breeding. Common Gull Larus canus Eighty pairs were counted in 1969 and during the 1970s numbers fluctuated, declining to 20 in 1973 then rising to 40 pairs in 1976. There was a peak count of 90 pairs in 1983. Numbers have since fallen to only 25 pairs in 1996; 50 birds were recorded in 1997 and 51 in 1998. Lesser Black-backed Gull Larus fuscus This gull has declined dramatically in the last 15 years and it now appears to be extinct as a breeding species. One hundred pairs were found in 1969 increasing to 200 in 1973, numbers then fluctuated but gradually declined to 125 in 1983. In 1992 only 5 birds were counted; there were 2 pairs in 1993 but none have been found breeding since then. No birds were seen in 1998. Herring Gull Larus argentatus As with the previous species there has been a massive decline in recent years. From 450 pairs in 1969, numbers increased to a maximum of 900 pairs in 1976 then dropped to 600 pairs in 1983. In 1993 only 29 birds were counted, just 6 pairs in 1996 and 6 to 7 pairsin1997. Nine nests were found in 1998 but breeding success appeared to be low. Great Black-backed Gull Larus marinus Following an increase in the early 1970s numbers declined for a time but have since apparently stabilised. Only 20 pairs were noted in 1969 but there had been an increase to 125 in 1973 and a peak of 180 pairs was reached in 1974. Control measures in 1976 led to a decline to 60 pairs by 1983. One hundred and twenty five birds were counted in 1993 and 95 birds in 1996. In 1998 there were colonies of 80 and 25 birds, also 2 pairs nesting on inland peat banks and 7 pairs along the west coast. Kittiwake Aissa tridactyla The 1969 count was 81 pairs; numbers fluctuated through the 1970s with 60 pairs in 1973, 27 in 1974, 42 in 1976 and 82 in 1980. There has been a gradual decline since then with 55 to 60 pairs in 1983, 28 nests found in 1996 and 27 in 1997. In 1998 only 18 nests were counted including 6 in one area, where the young appeared to have been predated by anearby pair of Great Black-backed Gulls. Scottish Birds (1999) Sandwich Tern Sterna sanavicensis Two pairs attempted to breed in 1973 but were unsuccessful. On recent visits birds have occasionally been seen flying over the island but there has been no evidence of any further attempts at breeding. Arctic Tern Sterna paradisaea Breeding numbers and group sizes have fluctuated widely during the study period. Only 160 pairs were noted in 1969 but there were 700 pairs in 1973, just 60 pairs in 1976 then a huge increase in 1980 when 4,000 pairs were recorded, the peak count. There were 800 pairs in 1983 and 4,200 birds in 1988. Recently numbers have remained fairly stable with between 1,000 and 1,500 birds annually from 1992 to 1996. Counts of individual birds including failed and nonbreeders were 2,060 in 1997 and 2,330 in 1998. Since the introduction of the North Ronaldsay breed of sheep 5 years ago there has been predation of those colonies on or close to the beaches. In 1998 a successful attempt was made to keep these sheep away during the breeding season from the main colony near the lighthouse. Guillemot Uria aalge This species appears to have increased during the study period, especially in the 1970s. Unfortunately visits in July are too late to give meaningful counts. Six individuals were noted in 1969, then in July 1976 280 birds were counted but many of these were thought to be non breeders. Over 100 birds were present in the main colony in 1985 and 190 0n 15 June 1986. There were 109 birds, some with young, on 9 July 1998. Razorbill Alca torda As with Guillemot, July visits are too late for accurate counts of breeding numbers. In The Breeding Birds of Auskerry, Orkney, 1969-1998 3 1973, 6 sites and in 1976, 25 sites, were occupied. Acounton 15 June 1986 found 70 birds and on 9 July 1998 26 birds on ledges, some with young. Black Guillemot Cepphus grylle Breeding pairs have declined since the mid 1970s but appear to have remained fairly stable during the last few years. Seventy four nests were found in 1976 but only 35 in 1983 and 37 in 1988. In 1993 and 1994, 26 nests were located, 28 in 1996 and 31 in both 1995 and 1997. Only 25 nests were recorded in 1998 but some other sites, usually occupied, had been flooded by run off from heavy rain. There is a count of 205 birds around the island in April 1983 (Tasker & Reynolds). The Seafarer count was 342 individuals in 1969-70. Puffin Fratercula arctica The main colonies are along the west coast and in the north east of the island. Puffins increased in the 1970s and have remained fairly stable or even increased slightly since then. Further increases may be limited by lack of suitable breeding sites. Thirty four birds were recorded in 1969, 90 in 1974 then anincrease to 225in 1976. Further counts are 175 to 225 in 1983, 200 in 1994 and 250 birds in 1997. There were 140 birds on land on 9 July 1998. other species Shelduck T7adorna tadorna A pair was present in May 1993 but breeding was not confirmed. Mallard Anas platyrhynchus One or 2 pairs bred, although not annually, in the 1970s. No recent breeding records. 4 RG Adam & C J Booth Eider Somateria mollissima About 20 pairs were recorded breeding in 1977. Numbers remain about the same. Red-breasted Merganser Mergus serrator A pair was present in 1985 and a nest with 10 eggs was found in 1986. Oystercatcher Haematopus ostralegus Four to 9 pairs bred annually in the 1970s and between 4 and 20 pairs in the 1990s. Ringed Plover Charadrius hiaticula Two pairs bred in the early 1970s, 6 in 1984 and 5 to 8 annually since 1994. Lapwing Vanellus vanellus A pair was present and probably bred in 1992. Dunlin Calidris alpina A single pair bred in 1974 and 1976. Snipe Gallinago gallinago Fifteen to 20 pairs bred annually inthe 1970's. Numbers still about the same. Curlew Numenius arquata In 1993 one pair was present and a nest found. Single pairs were present in 1996 and 1997. Turnstone Arenaria interpres A single bird was displaying and defending territory in 1992. Rock Dove Columba livia Between 20 and 30 pairs bred annually in the 1970s. Only 25 birds were seen in 1998. Skylark Alauda arvensis At least 6 pairs were noted in the 1970s. Skylark have declined with none seen in 1998. SB 20(1) Meadow Pipit Anthus pratensis Several pairs were recorded in the 1970s, at least 2 pairs in 1997 and 3 in 1998. Rock Pipit Anthus petrosus About 25 pairs were present inthe 1970s. A minimum of 18 pairs was located in 1998. Pied Wagtail Motacilla alba yarrellii None were recorded inthe 1970s. Two pairs bred in 1992 and 1993, single pairs in 1994 to 1998. Wheatear Oenanthe oenanthe Eight to 12 pairs were present in the 1970s, one pair in 1994, 3 in 1995, 2 in 1996, 1 in 1997 and 1998. Hooded Crow Corvus corone A single pair was seen in 1974. There were 2 pairs in 1992 and 1993; single pairs 1996 to 1998. Raven Corvus corax A single pair has nested on the west side in most years and is usually successful. Starling Sternus vulgaris At least 10 pairs bred in 1976. Flocks of 20 to 50 birds are seen annually. Twite Carduelis flavirostris Six to 10 pairs bred annually in the 1970s. None were seen in 1997 or 1998. Discussion Some of the changes detailed above are due in part to more intensive grazing with subsequent depletion of ground cover. Sheep disturb ground nesting birds and may eat eggs and young. Apart from Arctic Terns, the seabirds most affected have been Common, Lesser Black-backed and Herring Gulls, although sheep are probably not the Scottish Birds (1999) main reason for the dramatic decline in the last 2 species. Numbers of both these gulls have fallen at colonies elsewhere in Orkney during the last 10 years (pers obs) and decreases have also been recorded from other parts of Britain (Thompson, Brindley and Heubeck 1998). The drop in Kittiwake numbers is in line with the decline that has occurred at some colonies throughout Orkney since atleast 1975 (Lloyd, Tasker and Partridge 1991). While a decrease in Black Guillemots has been recorded on Auskerry from the mid 1970s, a decline at other Orkney colonies does not appear to have been noted until the late 1980s (Orkney Bird Reports 1988 to 1990). Two of the species that have increased since 1969 are Great Black-backed Gull and Guillemot. The Great Black-backed Gull population, despite fluctuations, is now apparently fairly stable although decreases have been reported from some Orkney colonies since 1990 (Orkney Bird Reports 1990 to 1996). The increase in Guillemot numbers would seem to follow the trends at monitored colonies elsewhere in Orkney (Thompson, Brindley and Heubeck 1998). Acknowledgements We are very grateful to Simon and Teresa Brogan and their family, not only for additional information, but also for their help and The Breeding Birds of Auskerry, Orkney, 1969-1998 5 hospitality during our visits to Auskerry. Andrew Ramsay carried out nearly all the early survey work on which this paper is based. Our thanks to S da Prato and an anonymous referee for their comments on an earlier draft of this paper. We are also indebted to the boatmen who have transported us to and from Auskerry in a variety of conditions, particularly John Deerness, Harvey Groat and especially Smith Foubister. References Cramp S, Bourne W R P and Saunders D 1974. The Seabirds of Britain and Ireland. Collins. Lloyd C, Tasker ML and Partridge K 1991. The Status of Seabirds in Britain and Ireland. T and A D Poyser, London. Orkney Bird Reports 1988 to 1994. Chris Booth, Mildred Cuthbert and Eric Meek (eds). Orkney Bird Reports 1995 to 1996. Chris Booth, Mildred Cuthbert and Bob Adam (eds). Ramsay A D K 1978. Breeding Birds of Auskerry. Orkney Bird Report 1976-1977 D Lea 1978. Ramsay ADK 1984. Seabirds on Auskerry 1971- 1983, Orkney Ringing Group Report 1983:3-9. Tasker M and Reynolds P 1983. A survey of Tystie (Black Guillemot) Cepphus grylle distribution in Orkney, 1983. Report to Nature Conservancy Council. Thompson K R, Brindley E and Heubeck M 1998. Seabird numbers and breeding success in Britain and Ireland, 1997. JNCC Peterborough. Wood D 1996. An estimate of the number of Storm Petrels Hydrobates pelagicus breeding on Auskerry, Orkney. Seabird 19:40-46. R G Adam, Northolme, St Ola, Orkney, KW15 1SB C J Booth, Ronas, 34 High Street, Kirkwall, Orkney, KW15 1AZ Revised manuscript accepted December 1998 6 Scottish Birds (1999) 20: 6-13 SB 20(1) The use, abuse and misuse of crow cage traps in Scotland: a report on behalf of the Scottish Raptor Study Groups and the RSPB D DICK & A STRONACH The use of crow cage traps was investigated across Scotland during March-July 1998. All known records of crow cage trap catches received by the RSPB between 1985-1997 from members of the public were collated. A wide range of rare and fully protected species had been accidentally but sometimes deliberately trapped in crow cage traps. Following these findings, a 5 month study into the use and legal status of a sample of 36 crow cage traps in Scotland was conducted. At least 78% of these crow cage traps were being operated illegally, with most failing to provide water, shelter and food. Although it is believed that most illegal crow trap use was not malicious and that many traps had been left carelessly set by operators, there is some historical and current evidence of crow trap abuse which is considered to be deliberate. The impact that crow cage traps have on protected species was investigated and it is concluded that crow cage traps pose a significant threat to a wide variety of protected species in Scotland. Changes to the law are urgently required and a number of recommendations are suggested that would reduce the threat crow cage traps pose to many protected species of birds. Introduction The crow cage trap is a widely used tool in the uplands of Scotland. Ithas along history of use in this country and is recommended by game and agricultural advisors for multiple live trapping of Crows Corvis spp, Magpies Pica pica, Rooks Corvus frugilegus and Jackdaws Corvus monedula (Dick 1997). The legitimate use of these traps is covered by the catch all phrase ‘cage trap’ in section 5(5)(a) of the Wildlife and Countryside Act 1981. There is no definition or description of ‘cage trap’ in any British or Scottish legislation. This lack of definition has led to a proliferation of cage trap designs, and uncertainty over the terms of their use and maintenance by users and over the advice given by various agencies involved with enforcement of wildlife legislation (Dick 1997). There are 3common types of crow trap: funnel, ladder and Larsen traps. All 3 types are widely used across Scotland but it is the first 2 which are the subject of this paper. Operators of crow traps must comply with the conditions of the Protection of Animals (Scotland) Act (1912), the Abandonment of Animals Act (1960), the conditions of the Open General Licence (1997) and the Wildlife and Countryside Act (1981). These pieces of legislation require the responsible Scottish Birds (1999) person(s) to undertake certain tasks in relation to the use of traps and behaviour in relation to animal welfare by not causing unnecessary suffering. This includes inspecting crow traps daily when in use, unless there are problems with severe weather, and removing any birds caught at each inspection. Water and food must be provided in traps and it is likely that the provision of shelter in the harsh upland climate would also be required. All protected species must be released immediately. Crow cage traps are generally used in open country for the purpose of catching non territorial crows outside the breeding season, and are visited by a wide variety of protected species of birds which are sometimes caught inthem. The main aims of this paper are to detail the known cases of protected species caught in crow cage traps, investigate their general use and assess their potential impact on a range of species. Methods The first part of this study examines the records of known trapping of protected species in crow cage traps received by the RSPB between 1985-1997. A separate list is also provided which lists incidents of crow cage trap abuse which were considered to be deliberate. Incidents with evidence of alleged wrongdoing were reported to the police. From this appraisal, we considered that a more detailed study would help to determine the extent of the problem with such traps. The second part of the study details the results of a5 month investigation into the use and legal status of a sample of crow cage traps visited in spring and early summer 1998. Crow cage traps were visited across The use, abuse and misuse of crow cage traps in Scotland Vi Scotland at 26 estates and farms in Aberdeen, Angus, Banff, Dumfries, Inverness, Lanark, Lothian, Moray, Peebles, Perth, Roxburgh and Wigtown. Traps were selected by contacting local raptor workers and asking for details of all the known crow traps in their areas. From their responses these traps were then visited and checked and details of their size, structure and contents were recorded along with their legal status. Results The RSPB received 52 reports of incidents involving the trapping of protected species in cage crow traps in Scotland between 1985- 1997. Reports came from a wide variety of sources including members of the public, estate workers, gamekeepers, shepherds, members of RSPB staff and the police. Reports were received from 16 Districts across Scotland from the Outer Hebrides to Borders. Incidents of abuse of crow traps considered deliberate, 1985-1997 1 One Buzzard Buteo buteo killed by a keeper, March 1987, Renfrew hills. 2 Keeper told journalist that crow traps were being used to catch raptors for destruction including Hen Harrier Circus cyaneus and Golden Eagle Aquila chrysaetos, 1988, Dumfries. 3 One Golden Eagle November 1989 and 2 more 6 days later November 1989, Islay. 4 One live and one dead Buzzard and 3 dead (starved) Blackbirds Turdus merula, January 1990, Perth. 8 D Dick & A Stronach SB 20(1) Table 1 Reported records of crow cage trap catches received by RSPB between 1985- 1997. Abuse that was considered deliberate has been omitted. A few reports received concerned cage crow traps containing ‘many’ or ‘several’ of a particular species. For the purposes of this table ‘many’ and ‘several’ have been counted as 3, which is a minimum estimate. Species No of birds Buzzard 73 Fieldfare 25 Kestrel 28 Golden Eagle 15 Sparrowhawk 13 Raven 13 Tawny Owl Barn Owl Blackbird Unidentified finch Unidentified owl Short-eared Owl Goshawk Hen Harrier Red Kite Rough-legged Buzzard Wheatear —-=-=-=- PHO NMNWWHO O 5 One Goshawk Accipiter gentilis trapped live by keeper and illegally passed on to a bird of prey keeper, February 1992, Borders. 6 Two dead crows (starved), May 1995, Borders. 7 One live adult female Goshawk November 1995, Lanark. 8 Two Buzzards seized from a keeper who had removed them from a crow trap, put them in a sack and was returning to his vehicle, 1996, Inverness. No of traps No of regions 17 9 1 1 a 8 10 6 3 3 4 3 1 1 1 1 2 2 1 1 1 1 2 2 1 1 1 1 1 1 1 1 1 1 Crow cage trap survey 1998 The legal status of a sample of crow cage traps was investigated in 1998. 78% of the traps checked were being operated illegally (Table 2), with most failing to provide water, shelter and food (Table 3). Three of the crow cage traps that were used illegally also had no doors to facilitate the removal of non target species another illegally set trap had its door padiocked shut. Twelve crow cage traps were obviously not inspected daily because 8 contained dead, apparently starved Crows, onecontained adead Kestrel | | Scottish Birds (1999) The use, abuse and misuse of crow cage traps in Scotland 9 Table 2 Legal status of a sample of crow cage traps visited in 1998. Legal status Used legally* 6 Used illegally 28 Not in use, secured open 2 Total 36 Number of crow cage traps Percentage 17% 78% 6% 100% *Used legally if they were checked every 24 hours. Table 3 Criteria used for determining illegal use of the 28 crow cage traps. Criteria Number of crow cage traps Percentage No provision of water 18 64% No provision of food 17 61% No provision of shelter 17 61% Not visited daily 1Z 43% Falco tinnunculus, one a dead Sparrowhawk Accipiter nisus, along with a dead thrush Turdus sp and another contained a dead Merlin Falco columbarius which had been dead for some time. Two other traps contained a large number of Golden Eagle feathers, suggestive of a lengthy period of confinement with subsequent feather damage or destruction and removal of the trapped Golden Eagles. Without watching traps for 24 hour periods it was not possible to know how many traps were checked daily, but it is likely that if the welfare of the trapped birds was not paramount, as was the case in the majority of traps visited, then regular daily trap checking would also be a low priority in many cases. In most instances there were no signs of footprints or tracks on wet ground which would have been evident if crow traps were being closely checked daily. The wire mesh sizes of each trap were recorded. Mesh size varied between 1 and 2 inch diameter, with 1.25 inches being the most commonly used mesh size. The food/bait contents of each trap were recorded. Food was provided in 17 traps, which consisted of various lagomorphs in 15 (88%) traps, lambs/sheep in 3 (18%) traps, Pheasant Phasianus cholchicus in 3 (18%) traps and deer, grouse, duck, eggs and bread in one (6%) trap each. These bait items are main prey items for a number of raptors. The live contents of each trap were recorded. Of the 34 traps in use, 9 (27%) contained live birds. Six contained Crows, one contained 10 DDick &A Stronach SB 20(1) a live Rook, one contained 2 live Mistle Thrushes Turdus viscivorus and 2 Starlings Sturnus vulgaris which were released and one contained alive Kestrel Falco tinnunculus which was released. ALong-eared Owl Asio otus was found next to a funnel trap with several owl feathers in and around the trap. The bird had a broken left wing and had to be destroyed by avet. A post mortemrevealed that the owl had been shot. Discussion We have been unable to locate a single item of published research on the design, use and effects of cagecrowtraps. Thisis remarkable considering their widespread use in Scotland andbeyond. This paper therefore represents the first attempt to examine the effects of these traps on protected species. It is clear from the reports received and presented in this paper that crow cage traps are widely used across much of upland Scotland and that non target species are frequently trapped. A likely source of bias in the discovery and reporting of protected species by the public would be towards the finding of live birds in traps aS opposed to dead ones. People would be attracted to the flapping of these live birds and it is quite possible that many more traps contained dead birds than were reported. During this study atleast 20 protected species of bird were recorded trapped in crow cage traps, as well as several non specific reports concerning thrushes and finches. The reasons why these birds enter cage traps were not always Clear, but the great majority could be explained in terms of the food requirements of individuals. There also appears to be an element of mutual attraction between members of the same species, particularly Kestrels, possibly reinforced by family ties. The great majority of crow traps investigated in 1998 were being operated illegally across Scotland. Most crow traps investigated failed to provide water, food and shelter for their bait bird and potential victims. There is evidence that some crow traps were set deliberately to catch and destroy raptors; however, it is believed that most illegal trap use was not malicious but that many traps were carelessly set by operators. /fatrapis notinuse, then it must be rendered incapable of holding or catching birds (Open General Licence Conditions 1997). Seventeen (61%) of the traps being used illegally were probably not deliberately set to catch birds ie those without bait inside. These traps frequently had no food, water or shelter but had been left with the door securely shut and the funnel/ladder entrance open. In only 2 cases were crow traps secured open and rendered incapable of holding or capturing birds. Empty traps left with their door shut pose a significant threat to several species and have been responsible historically for the capture, and sometimes subsequent death, of Golden Eagle, Kestrel, Buzzard, Rough-legged Buzzard Buteo lagopus, Sparrowhawk and Short-eared Owl Asio flammeus. In this survey 3 protected bird species died as a result of neglect of crow traps: these were Merlin, Sparrowhawk and Kestrel. Although there were several examples of set traps being left unattended for extended periods, we believe that this is an under recorded and a widespread problem. Crow cage traps are primarily targeted at Carrion/Hooded Crows Corvus corone Scottish Birds (1999) The use, abuse and misuse of crow cage traps in Scotland 17 corone/C c cornix on the open hill, with statutory agencies recommending their targeted use in exposed positions well away from buildings (MAFF 1985 and ADAS 1987) such as prominent ridges or hill tops (FWAG 1988). However, there has been adisturbing increase in their use in woodlands particularly associated with Pheasant rearing, with traps found set with live bait in the form of pigeons or rabbits. There is evidence that these traps are specifically set for trapping raptors in a habitat that would not normally be associated with crows. The widespread use of small mesh wire in crowtraps increases the likelihood of trapping small birds and consequently some species of raptors, particularly Kestrel, Soarrowhawk and Merlin. The only people normally allowed to handle these wild birds are trained, licensed bird ringers. It seems an anomaly that untrained, unlicensed crow trap operators should be handling protected species. At present no one is yet calling for licensing of each trap, as is practised in other countries eg the Netherlands, but if crow cage trap abuse continues this may become unavoidable. As most crow cage traps are used in areas remote or deliberately concealed from the general public the evidence presented in this paper represents the tip ofthe iceberg. There is a clear need for statutory agencies and crow cage trap operators to consider ways to reduce the capture of non target species, whilst still allowing legitimate crow control. Potential impacts on a range of individuals and species Golden Eagle is of particular concern because the attraction of carrion such as lagomorphs, sheep and deer in traps is obvious; these items are a major food source for many eagles. It should be noted that eagles will also eat decoy crows. Dick (1997) considered that dispersing young birds are the most likely victims. Wide dispersal of juveniles into areas without resident eagles is a recognised pattern of both the Scottish and tiny English populations. Young birds do not breed until they are at least 5 years old and they are therefore at risk for a considerable period in areas where their presence may not be expected by the local vermin controller. Their loss to the small UK breeding population may also preventthe possible expansion ofthis species into former haunts. A further concern is the trapping of adult eagles during their breeding cycle when both adults are necessary to successfully incubate eggs or feed growing chicks; even a 24 hour absence could be fatal. Buzzard was the most commonly reported victim of trapping. One reason for this high level of reporting is undoubtedly that sucha large bird flapping in a trap will attract the attention of any passer by. However, their preference for carrion is undoubtedly a major factor. Most worrying was the number of detailed reports received of Buzzards with feather damage due to prolonged attempts at escape. Any bird with less than 100% ability to fly and hunt, particularly in winter, has to be seriously at risk. Buzzards also featured in a number of crow trap abuse cases which were considered to be deliberate. Kestrel Several reports were received of multiple trapping of this species. It is of interest that most reports concerned trapping during the summer months, after young have fledged. These birds do not normally eat 12 D Dick & A Stronach SB 20(1) SSS SS SR carrion. __It is more likely that a bird is attracted by small birds or rodents feeding on insects and maggots produced by the carrion. Once the first bird is trapped, others may be attracted by its movements and calling and a mixture of curiosity and familial attraction draws them into the same trap (Dick 1997). Kestrels cannot survive for as long as Buzzards and Golden Eagles without food and shelter and are likely to die from hunger and exposure if they remain trapped for long periods of time. Merlin This species does not normally eat carrion therefore Merlins are probably attracted by small birds feeding inside the trap. One adult male Merlin was found long dead inside a small mesh cage along with a live Kestrel. Sparrowhawk As this species feeds exclusively on live birds, the reason for its appearance in traps is due to attempts to catch small birds already trapped inside. The fact that all of the traps investigated used small mesh chicken wire means that many traps can hold small birds and, therefore, attract Soarrowhawks, Merlins and Kestrels. Goshawk There are a few reports of this scarce species entering cage traps to catch live prey. On at least 2 occasions it was believed that live prey (rabbits and pigeons) had been placed in the crowitraps specifically to catch Goshawks which would then be killed. Owls With several reports of trapped owls from members of the public it was not possible to identify individual species. However, confirmed reports of Tawny Owl, Short-eared Owl, Long-eared Owl and Barn Owl were received, with one report of repeat trapping of Barn and Tawny Owls. The birds are undoubtedly attracted by live rodents, or possibly small birds, in the cage traps. As with Kestrels and Sparrowhawks, these owls are relatively small birds of prey with high energy requirements and therefore need a steady supply of small fooditems. Prolonged periods of confinement would soon be fatal for these birds. The deliberate shooting of a Long-eared Owl suggests that not all deaths were accidental. Other bird species Although large crow cage traps can be constructed using wide mesh wire netting, all the crow traps examined during this study were constructed using fine mesh chicken wire and were therefore capable of trapping passerines. Once a small bird gains entry it is trapped as effectively as any crow. As mentioned previously, it is considered that these small birds are entering traps in order to feed on insects encouraged by bait carcasses. The reports of deaths of these birds should come as no Surprise given their need for continuous feeding, water and cover at night (Dick 1997). Recommendations “+ Continued regulation and increased enforcement of crow cage trap legislation. “* A licence condition that police/ statutory agencies are given locations of crow cage traps other than Larsen traps on all estates. “* A strict code of conduct to cover licensed use of crow cage traps should be produced by the Scottish Executive. ** Ban the use of small scale chicken wire mesh in construction and ensure that all traps have doors to help facilitate the removal of non target species. “+ Operators to record all non target trappings. Scottish Birds (1999) The use, abuse and misuse of crow cage traps in Scotland 13 “+ Crow cage traps are designed for use References on exposed positions in open hill country; their use in woodlands should be banned. Dick, D 1997. Crow cage traps — the case for tighter legislation. Scottish Bird News 45: 6-7. MAFF 1985. Leaflet 907 Crows. ADAS 1987. Leaflet P907 Crows. FWAG 1988. Information note S4. Foxes, Crows and the Hill Farmer. Dave Dick and Andy Stronach, c/o RSPB Scotland, Dunedin House, 25 Ravelston Terrace, Edinburgh EH4 3TP Revised manuscript accepted December 1998 Crow cage trap RSPB 14 Scottish Birds (1999) 20: 14-17 SB 20(1) The breeding status of the Spotted Crake in north east Scotland | FRANCIS & A THORPE Breeding season records of Spotted Crakes in north east Scotland from 1989- 98 are summarised, including the results of more intensive survey work during 1996-98. One to 5 calling males per year were recorded from 9 sites in total, almost all of which are extensive wet fens dominated by sedges Carex spp. population in some years. Introduction The Spotted Crake Porzana porzana is a rare breeding bird in Britain and Ireland (Stone et al 1997), though it is considerably commoner in central and eastern Europe, with 52,000-180,000 pairs in total (Tucker and Heath 1994). In Scotland, numbers have varied greatly from year to year, with more favoured areas historically being south west Scotland, Strathspey and Sutherland (Thom 1986). North east Scotland (Aberdeenshire and Aberdeen City) was not mentioned by Thom in connection with Spotted Crakes, Buckland et a/ (1990) note only 3 recent records (1977-88) and there was one breeding record mapped in the 1968-72 Atlas (Sharrock 1976). Since 1989, however, there have been many more records (NE Scotland Bird Reports 1989-97; RSPB and SNH unpublished information), with evidence of breeding on atleast 2 occasions. In 1990, the Scottish Ornithologists’ Club Grampian Branch organised a Nocturnal Survey (Bain 1991) which yielded much useful information on some less well recorded species (including Spotted Crake), particularly on lowland peat mosses and fens. The 1988-91 Atlas (Gibbons et al 1993) also recorded 4 x 10km squares with either breeding evidence or birds present. After 1991, however, there was no further This number of birds may represent 10-20% of the UK breeding systematic attempt to locate the birds. This paper documents the increase in records of Spotted Crakes during the breeding season over the last 10 years and reports on more systematic survey effort in NE Scotland from 1996-98. Methods We collated all records of Spotted Crakes from 1989-95 using records held by the NE Scotland Bird Recorders and RSPB and SNH unpublished information. From 1996- 98 we visited all sites known to have held birds in recent years, plus anumber of others judged to have potentiallly suitable habitat. In total, 21 sites were visited during darkness, usually from 2330-0200, between mid May and mid June in calm and, if possible, warm weather conditions. At least half an hour was spent listening for the characteristic ‘whip lash’ call at all sites. At some sites thought to be potentially suitable, but where birds had never been previously recorded, tape recordings of calls were played. However, tape playing did not yield any additional success and birds were always detected calling on arrival at occupied sites. Results Eight sites were visited in 1996, 18 in 1997 and 20 in 1998; 8 were visited in all 3 years. Scottish Birds (1999) The breeding status of the Spotted Crake in NE Scotland 15 Table 1 Breeding season records of Spotted Crakes in NE Scotland, 1989-98. Year Details 1989 1990 Bourne 1992) 1991 2 calling males at 2 sites 1992 4 calling males at 3 sites 1993 5 calling males at 3 sites 1994 5 calling males at 4 sites 1995 3 calling males at 2 sites 1996 5 calling males at 4 sites 227) 1998 The results from our survey are included in Table 1. lt should be noted that although the maximum number of sites at which birds called in any one year was 4, the total number of sites utilised by calling/breeding birds during 1989- 1998 was 9, plus 2 more from which possible records were reported. Of these 9 sites, 8 are extensive fens or marshy wetlands with high summer water levels, dominated by extensive Carex beds or other very wet fen vegetation, sometimes with patchy Phragmites reed and always with scattered willows Salix spp. The remaining site is a marginal waterside fen. Three sites have been used by calling birds once only, the remainder being more favoured and used for 2 ormore years in the last 10 witha maximum of 9, though one appears to have been deserted, with birds last recorded in 1992. Breeding was confirmed at one site in 1997 and suspected at another in 1990 (Bourne 1992). 3 calling males at 3 sites during breeding season 4 calling males at 3 sites; breeding probable (Suspected pair with young; 2 calling males at 2 sites; breeding proven (young birds seen) at another 1 calling male at 1 site (poor summer weather hampered survey results) Discussion From 1973-85 there were up to 12 calling males at up to 6 sites in Britain (Batten et a/ 1990). During the 1988-91 Atlas period (Gibbons et al 1993), 17-21 calling males were recorded at 10-14 sites in Britain and lreland. The most recent UK population estimate (Stone et a/ 1997), using data from 1989-92, is 1-20 pairs, though in Britain in 1993 there were 31 calling males at 19 localities, in 1994 11 males at 5 sites and in 1995, the most recent year for which information is available, there were 10 calling males at 8 localities (Ogilvie et al 1998). Clearly, the 1-5 calling birds per annum at up to 9 sites in total in NE Scotland are of considerable significance in both Scottish, British and UK terms — perhaps 10-20% of the recorded UK population in some years. C J Mead (in Gibbons et a/ 1993) considered that because of the penetrating call and precise habitat requirements, it is most 16 I|Francis & A Thorpe SB 20(1) unlikely that regular breeding sites are missed. This is not necessarily true in relatively under watched areas of Scotland. Certainly, an increase in the number of observers during the last 10 years might have contributed to the apparent increase in records in north east Scotland. However, it is interesting to note that even when a systematic effort was made to survey the birds in 1966-1998, we did not find more calling birds than in other years, and most records came from the same observers each year, rather than an increased number. The possibility remains that we under recorded the species on our visits, as some calling birds may have been missed. Apart from effects of bad weather and limited time spent at any one site, birds may be silent for part of the breeding season if pairing is successful (Cramp and Simmons 1979). The method we usedis similar to that adopted by the national Corncrake Crex crex survey (Green 1995; D W Gibbons in litt 1998 Corncrake Survey instructions). Corncrakes are known to sing most persistently between 2300 and 0200 hours GMT (0000 and 0300 BST), within which period males have been found to sing continuously during 70-80% of visits to home ranges. In the case of this species, the time spent singing at night varied little between late May and mid July (Green 1995). However, 2 visits are made for Corncrake survey work, which was not the Case in our Surveys. This also may imply that some Spotted Crakes were missed in NE Scotland. The RSPB, in conjunction with the statutory nature conservation agencies, will be coordinating a national census of Spotted Crakes in 1999. This will provide a better opportunity to assess the significance of NE Scotland in a year of enhanced fieldwork effort, standardised across the UK. We conclude that there has been a genuine increase in the presence of Spotted Crake during the breeding season in NE Scotland, with breeding proven or probable at least twice andlikely, as deduced from the numbers and regularity of calling birds, at several sites during 1989-98. Atleast 9 sites are suitable forthe species. The habitat occupied by the birds seems typical of that used elsewhere (Batten et al 1990; Tucker and Heath 1994), and there is no evidence of short term habitat change at any of them. Six of the sites are protected by SSSI status. However, long term vegetation succession will almost certainly lead to the drying out of some fens and SSSI status does not provide an easy solution to that potential problem. In addition, since most fens are not in agricultural use, measures such as the Scottish Countryside Premium Scheme are unlikely to be effective. Positive habitat management measures are needed, leading to the extension and rewetting of fens, and the creation of new wetlands liable to be colonised by wet fen vegetation, particularly Carex spp, in order to ensure that suitable habitat remains into the future. National and Local Biodiversity Action Plans may be the best way to achieve this, since fens are a UK priority habitat (Biodiversity Steering Group 1995). The requirements of Spotted Crakes should be built into this process. Acknowledgements We would like to thank Graham Rebecca and lan Bainbridge for their comments on the manuscript, and Keith Morton and staff of Scottish Natural Heritage for information and comment. References Bain D J 1991. The SOC Nocturnal Survey 1990. North East Scotland Bird Report 1990. Batten L A, Bibby C J, Clement P, Elliott GD & Scottish Birds (1999) The breeding status of the Spotted Crake in NE Scotland 17 Porter RE 1990. Red Data Birds in Britain. Poyser, London. Biodiversity Steering Group 1995. Biodiversity: the UK Steering Group Report. Volume 2. Action Plans. London, HMSO. Bourne W RP 1992. Alarm calls used by a presumed Spotted Crake in the Grampian Region. Scottish Birds 16: 220-221. Buckland S T, Bell M V & Picozzi N 1990. The Birds of North East Scotland. North East Scotland Bird Club, Aberdeen. Cramp S and Simmons K E L (eds) 1979. The Birds of the Western Palearctic. Volume 2. OUP, Oxford. Gibbons D W, ReidJB & Chapman RA 1993. The New Atlas of Breeding Birds in Britain and Ireland: 1988-1991. Poyser, London. Green RE 1995. The decline of the Corncrake Crex crex in Britain continues. Bird Study 42: 66- 765. Ogilvie M A & Rare Breeding Birds Panel 1998. Rare Breeding Birds in Britain in 1995. British Birds 91: 98-443. Sharrock JT R 1976. The Atlas of Breeding Birds in Britain and Ireland. Poyser, Berkhamsted. Stone B H, Sears J, Cranswick P A, Gregory RD, Gibbons D W, Rehfisch MM, Aebischer J A & Reid JB 1997. Population estimates of birds in Britain and inthe United Kingdom. British Birds 90: 1-22. Thom V M 1986. Birds in Scotland. Poyser, Calton. Tucker G M & Heath M F 1994. Birds in Europe: their conservation status. BirdLife International, Cambridge, UK. lan Francis, RSPB, 10 Albyn Terrace, Aberdeen AB10 1YP Andy Thorpe, 30 Monearn Gardens, Milltimber, Aberdeen AB13 OEA Revised manuscript accepted February 1999 |i if) 1 Hi Ahi Hi ll : | i ti iA Hef: | Pie ign: 1 Ki t! [ | SS Uy 4 Spotted Crake Mike Ashley 18 Scottish Birds (1999) 20:18-26 SB 20(1) Changes in the numbers and distribution of Mute Swans in the Lothians in spring from 1978 to 1998 A W BROWN & LM BROWN Between 1978 and 1998 the number of Mute Swans in the Lothians in April increased by 318%. This change occurred during 2 distinct periods, 1978 to 1985 when there was no significant increase, and 1986 to 1998 when the population increased exponentially with substantial increases apparent amongst both non territorial birds and territorial pairs. Additionally, from 1986 the percentage of non territorial swans was consistently greater than the percentage of territorial swans, in contrast to the earlier years of the study. The majority of non territorial swans tended to collect in flocks on still waters or at estuaries. However, numbers did not increase in all flocks nor did increases occur simultaneously between flocks. Whilst the number of territorial pairs on canal, still water and river habitats increased the number on rivers decreased as a percentage of the Lothians territorial population. Introduction This study commenced as a result of concerns expressed by local ornithologists regarding the number of Mute Swans Cygnus olor in the Lothians. The consensus was that numbers had declined substantially but there were too few data to substantiate the impression that the population had decreased during the 1970s. An annual census of the spring population was instigated to monitor trends. Changes which occurred in the structure of the Lothians population in spring during the past 20 years are presented in this paper, during which time the total population increased from 116 swans in 1978 to 485 in 1998. Since the total population comprised non territorial and territorial swans the changes in each were dealt with separately in order to investigate the processes by which such a dramatic increase occurred. Study Area and Methods The study covered an area of 2,000 sq km and comprised East, Mid and West Lothian in addition to the City of Edinburgh. It was anticipated that the majority of non territorial swans would be located in flocks at Linlithgow Loch, the Almond Estuary at Cramond, St Margaret's Loch in Edinburgh, Inverleith Pond in Edinburgh, the Esk Estuary at Musselburgh, the Tyne Estuary and associated fields near Tyninghame, and the Water of Leith at Leith and that small numbers would be present at additional waters throughout the study area (D G Andrew, AT McMillan, G L Sandeman and R W J Smith, pers comms). Non territorial swans were counted in early and late April from 1978 to 1984 (Brown and Brown 1984) and subsequently during one coordinated mid April weekend. Additionally, visits were Scottish Birds (1999) made to potential territorial sites between late March and early May in order to record the number of territorial pairs. Results Total population The total population increased by 318% between 1978 and 1998 (Appendix 1). Asa substantial increase occurred after 1985 (Figure 1) data were divided into 2 time periods, 1978 to 1985 and 1986 to 1998 (Brown, 1997). The mean annual increase was 1% from 1978 to 1985 but the trend was not significant (Spearman Rank Correlation Coefficient, r, = 0.667, n=8, p>0.05). A significant upward trend (r, = 0.989, n=13, p<0.01) occurred between 1986 and 1998 when the mean annual increase was 10%. Changes in the numbers of Mute Swans in Lothian 1978-1998 19 Non territorial population Amean annual decrease of 3% was recorded in the number of non territorial swans between 1978 and 1985, although this trend was not significant (r, = 0.000, n=8, p>0.05). A mean annual increase of 12% occurred during the period 1986 to 1998 which was significant (r, =0.984, n=13, p<0.01). During the early years the majority of non territorial swans occurred at the coastal site of Tyninghame which held a mean of 54 birds and 87% of the Lothians non territorial population during the first 3 years of this study (Figure 2). There was no significant trend in numbers there from 1978 to 1998 (r, = 0.296, n=21, p>0.05) with the mean fluctuating around 52 birds. During the last 4 years the site held a mean of 58 swans but Figure 1 The total (¢), non-territorial (o) and territorial (+) numbers of Mute Swans in the Lothians from 1978 - 1998. 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 Year 20 AWBrown & L M Brown SB 20(1) Figure 2 The number of non territorial Mute Swans present in flocks in April and their proportion (%) of the Lothians non territorial population (-e-) from 1978 - 1998. t Year Tyninghame St Margaret’s Near Linlithgow Year 100 con 80 ween 60 40 ——---——~ = 20 oe 78 SO. 82 £84°57986 Cramond Year Inverleith Scottish Birds (1999) Changes in the numbers of Mute Swans in Lothian 1978-1998 21 Musselburgh Leith only 19% of the non territorial population. St Margaret's Loch was continuously occupied from 1983 with a build up in numbers to a mean of 70 during the period 1995 to 1998 which constituted 23% of the non territorial population. Occupation of Linlithgow Loch commenced in 1986, followed by anincrease to a mean of 62 and 20% of the non territorial population during the last 4 years of the study. Small numbers were recorded intermittently at Cramond up to 1990; however, since 1991 numbers consistently exceeded 18, reaching a mean of 59 or 19% of the non territorial population from 1995 to 1998. Development of these 3 flocks contributed substantially to the significant increase in numbers in the Lothians between 1986 and 1997. Results show that these flocks rarely exceeded 70 swans and in the last 4 years each held about 20% of the non territorial population and around 60 birds. Each site appeared capable of holding many more swans, however, overall density may have SGn co , 90 92 94 9G 3S Year been less important in determining flock size, but rather individuals may have been less successful when competing for available resources, eg food, in a flock in excess of about 60 swans. Thus when a flock approached that size then swans tended to move to a smaller flock or to establish a new flock, as happened with the recent development of a flock at Inverleith Pond (Figure 2). Such behaviour may also account for the lack of an overall increase in numbers at Tyninghame. Although a flock occurred at Musselburgh in most years, numbers remained low during the last 4 years of the study when it held a mean of just 8% of the non territorial population. Permanent flooding of the Water of Leith at Leith and the associated demise of the flock there was documented by Brown and Brown (1984). Territorial population A significant upward trend occurred in the number of swans in the territorial population during the period 1978 to 1985 (r, =0.994, 22 AWBrown & LM Brown n=8, p<0.01) and during the period 1986 to 1998 (r, =0.981, n=13, p<0.01) when mean annual increases for both periods were 7%. Although the number of territorial swans increased, this section of the population decreased as a percentage of the total population during the study period. Overall a mean of 38% of the total population comprised territorial swans, the mean for the period 1978 to 1985 was 46% (range 34% to 54%) and the mean for the period 1986 to 1998 was 36% (range 29% to 43%). During the early period of the study the percentage of territorial and non territorial swans in the total population overlapped (Figure 3) and showed no significant trend (r, =+0.452, n=8, p>0.05). Acontrasting pattern was evident when the total population increased substantially between 1986 and 1998; the percentage of territorial swans SB 20(1) was consistently less than the percentage of non territorial swans with a significant divergent trend (r, =+0.791, n=13, p<0.01). Territorial habitats in the Lothians were categorised as canal, still water (ponds, lochs and reservoirs) or river. The mean annual increase between 1978 and 1998 on canal habitat was 13%, on still waters was 9% and on rivers 1%(Figure 4). In 1978 river habitat comprised 45% of the total number of territorial pairs but this had decreased to only 15% in 1998. At the same time the percentage on canal habitat increased from 5% to 15% and on still waters from 50% to 70% (Appendix 2). Thus, although the number of pairs increased the percentage change was not consistent over all habitat types and river habitat became less attractive to territorial swans. Figure 3 The percent of non territorial (.) and territorial (0) Mute Swans in the total population in the Lothians in April from 1978 -1998. 100.0 80.0 ee Sy Gy Sd fe) asl 20.0 0.0 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 Year Scottish Birds (1999) Changes in the numbers of Mute Swans in Lothian 1978-1998 23 Figure 4 The number of territorial pairs on still waters (¢), rivers (0) and canals (+) in the Lothians from 1978 -1998. Discussion The Mute Swan population in north west Europe increased by 60% between 1984 and 1994 (Rose 1995). Results from the Common Bird Census and Waterways Blrd Survey indicated that the British breeding population increased in number during the past 20 years (Marchant et a/ 1998) whilst results from the Wetland Bird Survey show that winter numbers increased also (Cranswick eta/1997). Although constituting only about 5% of the Scottish population (Rawcliffe 1958; Ogilvie 1981; Brown and Brown 1985; Brown and Brown 1993) the Lothians appear to have reflected the trend in the numbers in the wider population. Population growth theories (Begon et al, 1986) suggested that, because the Lothians population was low in the late 1970s any increase in numbers would have been slight during the early years of the study. Then, as the number of territorial pairs increased and by inference the number of births, a more rapid population increase would have followed. Density dependency would eventually cause the exponential rate of growth to cease and hence the population numbers to level off resulting in an upper asymptote and a sigmoidal growth curve, if environmental factors remained favourable. During the period 1978-85 there was no significant trend in the total population in the Lothians nor in the number of non territorial swans, although an increase was apparent in the territorial population. The upward trend in the territorial population continued during the period 1986-98 during which time the non territorial population, and consequently the total population, demonstrated a similar trend. However, the population continued to grow exponentially and there was no indication of density dependent growth by 1998. _ As yet it is unclear if the long term population change in the Lothians will conform to a sigmoidal pattern of growth. 24 A W Brown & L M Brown SB 20(1) Of the 7 sites which held flocks of non territorial swans, numbers increased initially at the inland sites of St Margaret’s Loch and Linlithgow Loch, slightly later at the coastal site of Cramond and most recently at the inland site of Inverleith Pond. The stepwise initiation of new flocks suggested site preference, although development of a new flock did not depend upon existing sites reaching capacity, except in the case of Inverleith Pond. A similar pattern was demonstrated earlier by Suter (1995) amongst Cormorants Phalacrocorax carbo wintering in Switzerland. New flocks of Mute Swans were located across the study area and their development reflected the exponential increase in non territorial and territorial swans throughout the Lothians. The divergent trend in the percentage of territorial and non territorial swans coincided with the period of exponential growth and was in part dependent upon the presence of suitable habitat across the Lothians. Increased urbanisation with its associated drainage schemes, modern efficient agricultural drainage systems and extraction of water from rivers for crop irrigation have all affected water levels in rivers. Widely fluctuating levels undoubtedly lead to changes in aquatic vegetation and to flooded nest sites and probably had an adverse effect on the number of territorial pairs of Mute Swans onrivers. Conversely, creation of farm ponds and reservoirs contributed to the increased number of pairs on still waters. The reason for the delayed increase in the number of pairs on canal habitat is unclear since water levels were maintained at a fairly constant height and aquatic vegetation was apparently abundant. Notall territories were occupied each year and consequently the territorial population has the capacity to increase further. Whilst some non territorial flocks became density dependent and numbers levelled off, the development of new flocks suggested that the non territorial population itself was not density dependent and likely to increase further. Similarly, an absence of density dependency in the territorial population was apparent and itis concluded that the Lothians Mute Swan population has the capacity to increase in future years. Acknowledgements We thank John Ford, Robin Henderson, Mick Marquiss, Malcolm A Ogilvie and William S Penrice for their invaluable advice and constructive comments on earlier drafts of this paper. Many people have contributed to this study by submitting their counts of swans and their assistance is greatly appreciated. In this respect we are especially grateful to John Helliwell, Derek Henderson, Jane MacGregor and Eric Wyllie. The cooperation of private landowners, local authorities and national agencies has been invaluable and much appreciated. In addition we are grateful to the Scottish Ornithologists’ Club, British Trust for Ornithology, Fife Council, Scottish Natural Heritage, Exxon Chemicals and Shell Expro for financial support. References Begon M, Harper J L & Townsend C R 1986. Ecology: Individuals, Populations and Communities. Oxford, Blackwell. Brown A W & Brown L M 1984. The status of the Mute Swan in the Lothians. Scottish Birds 13: 8-15. Brown AW & BrownLM 1985. The Scottish Mute Swan census 1983. Scottish Birds 13: 140-148. Brown AW & BrownLM 1993. The Scottish Mute Swan census 1990. Scottish Birds 17: 93-102. Scottish Birds (1999) Changes in the numbers of Mute Swans in Lothian 1978-1998 25 Brown L M 1997. Measurement of the demographic parameters of the Mute Swan Cygnus olor population in the Lothians. Unpublished PhD thesis, Napier University, Edinburgh. Cranswick P A, Waters RJ, Musgrove A J & Pollitt MS 1997. Wetland Bird Survey 1995-96: Wildfowl and Wader Counts. BTO/WWT/RSPB/JNCC, Slimbridge Marchant J, Wilson A & Glue D 1998. Changes in breeding bird populations, 1996-97. BTO News 216/217: 10. Ogilvie M A 1981. The Mute Swan in Britain, 1978. Bird Study 28: 87-106. Rawcliffe C P 1958. The Scottish Mute Swan census 1955-56. Bird Study 5:45-55. Rose P M (ed) 1995. Western Palearctic and South West Asia Waterfowl Census 1994. /WRB Publ. 35, \WRB, Slimbridge. Suter W 1995. Are Cormorants Phalacrocorax carbo wintering in Switzerland approaching carrying capacity? An analysis of increase patterns and habitat choice. Ardea 83:255-266. Allan W and Lyndesay M Brown, 61 Watt’s Gardens, Cupar, Fife, KY15 4UG. Revised manuscript accepted April 1999 . AQ Mute Swan with cygnet = 2 WN 7 Ww Mian 1 A \ “agg WBE RAW Soe GON ‘\ Wy” Andrew Dowell 26 AWBrown & LM Brown SB 20(1) Appendix 1 The number of Mute Swans Appendix 2 The number of terrritorial present in the Lothians each April pairs of Mute Swans by habitat in the from 1978 - 1998. Lothians from 1978 - 1998. Year veer Sons Aue meee Population Non territorial Occupied MOTTE Nem IROL YOTa EL : : aS pairs on __— pairs on pairs on phi lI i) terptories rivers canals still waters Wes} Wile 76 20 1978 9 { 10 1979 95 47 24 4979 8 { 15 19805 sits 65 26) 4980 7 { 18 1981 121 65 28 1981 7 1 19 1982 111 Sil 30 1982 7 { 24 1983 140 84 28 1983 9 ; 18 1984 127 65 31 1984 8 3 20 1985 124 60 32 1985 9 B) 4 1986 157 89 34 1986 7 3 24 1987 182 114 34 1987 9 4 21 1988 186 109 39° 1988 10 4 25 1989 220 130 45 1989 {2 3 30 19SOy S275 165 55 1990 13 8 34 1991 302 180 61 4991 14 8 39 1992 303 189 S/ 1992 12 9 36 99S moo 205 66 1993 10 9 47 1994 321 193 64 1994 8 11 46 1995 406 272 67 1995 8 11 46 1996 468 334 67 1996 8 11 48 1997 452 294 79 1997 12 Ss 52 1998 485 339 7/8} 1998 11 14 51 * The apparent discrepancies in the total population for 1980 and 1988 were due to 2 territories in 1980 being defended by single birds while in 1988 one territory was similarly occupied. Scottish Birds (1999) 20:27-45 728 SHORT NOTES Swifts nesting in Scots Pines at Abernethy Forest, Strathspey Generally, Swifts Apus apus nest in holes of old buildings in towns (Birds of the Western Palearctic Vol 4). However, natural nest sites, such as holes in large trees and on cliffs, are known to be used in a number of locations in Europe eg in trees in the Bialowieza Forest, Poland (Tomialojc et a/ 1984, Acta Orn 20: 241-310). Records of nest sites in trees in Britain are rare. Swifts were recorded nesting in old woodpecker holes in pines in ancient forest along the River Beauly around 1835, but, along with the decline of Great Spotted Woodpeckers Dendrocopus major and the ancient woodland, this habit was thought to have died out (Lack 1956, Swifts in a Tower, Methuen). However, there is arecent record of Swifts nesting in trees in Strathspey (Thom 1986, Birds in Scotland, Poyser). This note describes nest sites currently used by Swifts in the ancient native pinewood of Abernethy Forest, 57°15 'N, 3°40 W in Strathspey. Swifts occurred in the forest from May to August, and small groups were commonly seen flying over the canopy. Birds were observed entering holes of 10 Scots Pines Pinus sylvestris during 1996 to 1998, and were assumed to be nesting there. The mean nearest neighbour distances between nest trees was 0.82km (range 0.22—1.86km). Table 1 Descriptions of trees used by Swifts at Abernethy Forest. All the holes were woodpecker holes apart from 2(*) which were knot holes. DBH is the trunk diameter at breast height. Grid ref Tree DBH Live height (cm) or (m) dead NJ007174 8.8 39 D NJ008132 18.8 43 D NJ022163 24.7 fe LE NJ023141 20.2 58 D NJ024137 16.0 115 D NJ025143 15:2 86 D NJ026145 18.7 119 L NJ044150 21.0 142 I. NJ045154 19.5 97 D NJ052152 10.5 83 D Surrounding Number Mean tree of height density holes of holes (no/ha) (m) 71 1 6.6 309 3 133 138 ie 11.4 265 2 hes 605 13 10.4 301 5 11.0 442 5 9.1 51 16 10.2 118 V7 14.7 354 3 £5 28 Short Notes SB 20(1) Plate 1a A dead Scots Pines at Abernethy Forest where Swifts nest in old Great Spotted Woodpecker holes. Ron Summers Scottish Birds (1999) Short Notes 29 Plate 1b A live Scots Pine at Abernethy Forest where Swifts nest in old Great Spotted Woodpecker holes. Ron Summers 30 Short Notes The majority of the holes were old excavations of Great Spotted Woodpeckers (Plate 1a & 1b). Two sites were knot holes. The nests were either in the trunk or in large branches. Seven of the trees were dead. In one ofthe live pines, all 16 holes were in dead branches (Table 1). Tree height averaged 17.3m (measured with a clinometer) and the mean diameter at breast height (DBH) was 85cm (Table 1). The dead trees averaged 15.6m in height and 74cm in DBH. These dead trees are some of the largest in Abernethy Forest. Only 5.7% of a representative sample of standing dead trees in the old forest had a DBH of greater than 39cm, the minimum size in Table 1 (RSPB unpublished data). The number of holes in the trees ranged from 1 to 16, but it is not known how many were used as nest sites by the Swifts. The mean height of the holes was 10.2m (range 6.6 — 14.7, Table 1). The mean height of the holes excavated by woodpeckers was 9.4m. This was significantly higher than holes excavated SB 20(1) by woodpeckers in 36 trees not used by Swifts (the difference was 2.4m, t=2.3, P= 0.027), indicating that Swifts preferred the higher sites. In a wide ranging study of woodpeckers in Britain, Glue & Boswell (1994, British Birds 87: 253-269) found that the median nest height of Great Spotted Woodpecker nests was only 4.9m. The density of live and dead pines surrounding the Swift sites was 265 trees per hectare (range 51 — 605), typical of the old parts of the forest (Summers et a/ 1997, Botanical Journal of Scotland 49: 39-55). The low density of trees presumably makes it easier for Swifts to fly between the crowns of the trees and reach nest sites. Given that most of the Swifts’ nest sites were in old woodpecker holes, and woodpeckers are widespread in Britain, itis surprising that Swifts do not use them more commonly. Perhaps Swifts require old growth forest which provides the combination of high nest sites and alow density of trees. This habitat is rare in Britain. Ron Summers, Royal Society for the Protection of Birds, Etive House, Beechwood Park, Inverness, 1V3 2BW Revised manuscript accepted October 1998 Scottish Birds (1999) Size of Merlin pellets in winter and summer in Galloway There are, apparently, no published data on the dimensions of pellets ejected by Merlins Falco columbarius in winter or summer in Britain. In the North American subspecies, however, pellets average about 25 x 12 mm (Johnson & Coble 1967, Jack Pine Warbler 145: 97-98). Between 1970 and 1977, | collected 88 pellets from a winter roost and 76 pellets from breeding areas in Galloway. The pellets were measured when dry at their widest point and their length was measured to theirtapered ends. | amconfident that no other raptor species pellets were in these samples. Short Notes 31 The length of 88 winter pellets measured from 12mm (min) to 41mm (max) with a mean length of 23.6mm + 6.72mm SD; their width measured from 6mm (min) to 16mm (max) with a mean width of 11.8mm + 1.87mm SD. The length of 76 breeding season pellets measured from 12.5mm (min) to 53mm (max) with amean length of 29.3mm + 8.84mm SD; their width measured from 8mm (min) to 16.5mm (max) with a mean length of 11.6mm +2.09mm SD. Thus the most common size of 42 (48%) winter peilets at 20-29mm was slightly longer than the size of 31 (40%) breeding season pellets, but no distinction was made from adults and young birds pellets. The most common size of 57 (65%) winter pellets at 10-12mm was wider than the size of 32 (42%) breeding season pellets (Figures 1 and 2). Fig. 1. Frequency of Distribution of size of Merlin pellets in winter and summer 50 30 20 Number of pellets (mm) Go) 2D Se 32 Short Notes SB 20(1) a SSS SS Fig. 2. Frequency of Distribution in the width of Merlin pellets in winter and summer Number of pellets (mm) Winter Summer 46 7-9 10-12 13-15 16-18 Fig. 3. Frequency of Distribution in the length of Hen Harrier pellets in winter So 10 Ps eee AIA Scottish Birds (1999) Short Notes 33 Fig. 4. Frequency of Distribution in the width of Hen Harrier pellets in winter = ‘Wl. In comparison, Merlin pellets are about the same length as those from Kestrels Falco tinnunculus at 20-30mm but are smaller in width at 12-17mm (Village 1990, The Kestrel, London). The commonest size of Merlin pellets are slightly larger than Sparrowhawks Accipiter nisus which seldom exceed 20mm x 10mm (Newton 1986, The Sparrowhawk, Calton). During the same period | measured 100 Hen Harrier Circus cyaneus pellets from their winter roosts, the species most likely to = nes be mixed with Merlin pellets as they often share the same roosts. They measured between 46mm x 8mm with a mean length and width of 26.25 + 7.46mm SD and 14.21 + 2.37mm SD. _ Interestingly, the majority of Hen Harrier pellets at 20-29mm (45%) were the same length as the majority of Merlin pellets (48%) (Figures 1 and 3); Hen Harrier pellets are generally larger in width at 13- 15mmin winter (Figure 4) comparedto Merlins which are smaller at 10-12mm (Figure 2). R C Dickson, Lismore, New Luce, Newton Stewart, Wigtownshire DG8 OAJ Revised manuscript accepted November 1998 34 Short Notes Corn Bunting decline in Easter Ross We report a decline to apparent extinction of Corn Buntings Milaria calandra in Ross in 1991-97. We saw birds from near the coast between Shandwick and Tarbat Ness, west to Arabella and Fearn Station. Allbut3 were east of a line from Easter Arboll through Meikle Rhynie to Loans of Tullich. We putin most effort there, but also searched each year from Tainto Nigg. Almost all summered north of Tarrel(A) and south of Lower Pitkerrie (B). Although this was a small area, density was locally high, with adjacent cocks’ song posts as near as 100m. Winter. In March 1991, 65 from Hilton north to Wilkhaven (35 north of Seafield and 30 at Hilton), allin cereal stubble bar 6 eating grain at sheep troughs on grass. In September 1991, one, one and 8 in cereal stubble in A. In January 1992, 40 at Balmuchy crossroads and 7 at Seafield, on cereal stubble. In November 1993, 5 in A including 2 in cereal stubble and 1 in turnips. In February 1997, 7 at Cadboll Mount. None was seen in winter 1997-98. Summer. 1991, cocks were found singing from Shandwick north to Templecroft (15 in A and 15inB including one at Glastullich and one at Arabella). These 30 were just under half the 65 in March, as expected from a SB 20(1) presumed sex ratio of about 1:1. Out of 23 cocks where we noted crop type, 7 sang above spring barley, 6 winter rape, 6 grass, 3 winter barley and one winter wheat. 1992, 27 cocks were recorded from Shandwick north to Templecroft (16 in A and 11 in B including one in Arabella). Out of 25 cocks where we noted crop type, 11 sang above spring barley, 6 winter rape, 5 grass, 2 winter barley and one in oats. 1993, 24 cocks were recorded from Hilton to Templecroft (11 in Aand 13inB). Out of 16, 5 sang above spring barley, 4 peas, 3 set aside, 2 winter barley, one winter rape and one in potatoes. Turnips, potatoes, peas and rape were in 20% of fields, grass 40% and cereals 40% (HM). A car killed a hen carrying food. 1994. In April, 4 sang in a sunny spell at Seafield where we saw 2-4 in 1991-93, and no others were seen but weather was bad. Later, one cock was in B, 2 at Bindal, where AW and MNT saw 1-2 in 1991-93, and one each at Lochslin and Loans of Tullich (0 and 0-1 in 1991-93). This suggested no decline up to 1994. 1995-98. In 1995 a singing cock was at Bindal and one at Nigg. In May 1996 one was at Wilkhaven and no birds were seen in summer 1997 or 1998. A Watson*, R L Swann, H McGhie & M Nethersole-Thompson *c/o Institute of Terrestrial Ecology, Banchory AB31 4BY Accepted December 1998 Scottish Birds (1999) A further survey of Twites wintering in Caithness Over the 3 winters between 1992 and 1994 we carried out 5 surveys of Twites Carduelis flavirostris wintering in Caithness (Clark & Sellers 1997, Distribution and abundance of Twites wintering in Caithness, Scottish Birds 19: 1-9). These showed winter numbers to be typically around 2,000-3,000 birds with a peak of about 7,000 birds, the latter coinciding with an abundance of unharvested Oil Seed Rape Brassica napus. The surveys also showed that the main habitats used were weedy Turnip 8 napa fields, uncut rape and rape stubbles when available, andto alesser extent other stubbles, weedy areas such as roadside verges and Marram Grass Ammophila arenaria. With the passage of several years since the completion of this work and against the backdrop of aperceived continuing decline in the species’ numbers (Clark & Sellers 1997, Twite Winter Habitats in Scotland: Results of the 1997 questionnaire Short Notes B15 survey, unpublished report) it seemed of value to carry out a repeat survey to see whether any significant change in numbers has taken place. We present here the results of such a survey carried out in December 1998. Survey methods were as described in our earlier paper, except that we were fortunate enough to secure assistance from 3 other counters, rather than 2 as before. Habitat types follow those listed in Clark & Sellers (1998, Winter habitats of Twites in Scotland, Scottish Birds 19: 262-269). This new Survey was conducted over the weekend of 12-13 December 1998 and attempted to locate all Twite flocks in Caithness. The weather for both days was good with mainly cloudy bright conditions with sunny spells, light winds, minimal rain (one or two short showers only) and maximum daytime temperatures of 7°C on 12 December and 3°C on 13 December. Table 1 Summary of results from 12-13 December 1998 survey of Twites in Caithness. Turnips Rape & rape stubbles No of flocks 21 9 (49%) (21%) No of birds 18802 3280° (32%) (55%) Other Uncut Others stubbles oats 8 3 2 (19%) (7%) (5%) 598 146 23° (10%) (2%) (<1%) a_ Includes a flock of 150 birds on turnips and barley stubbles, and one of 300 birds on turnips and fodder rape. b Includes flocks of 120 and 220 birds on rape and barley stubbles c One flock of 15 on potatoes and one of 8 on setaside 36 Short Notes SB 20(1) The results obtained are summarisedin Table 1. A total of 5927 birds was recorded in 43 flocks ranging in size from 8 to 1,500 birds, with a mean of 138 birds. The very large flock of 1,500 birds had been present at South Murkle between Thurso and Castletown for some weeks prior to the survey and, so far as we are aware, is the largest ever recorded in Scotland. There was a second large group numbering 1050 birds (flocks of 300, 300 and 450) on 8 adjacent fields of rape and rape stubbles at Nipster, roughtly midway between Thurso and Wick. Overall, the 7 flocks of ©300 birds accounted for 60% of Twites located, and flocks of ©100 birds for 85%. The flocks were distributed throughout the normal winter range in Caithness, the larger ones being associated mainly with the intensively farmed area between Thurso and Wick, and the smaller flocks with more peripheral areas, especially the north west of the county. The majority of birds and most of the larger flocks were found in fields of uncut rape or rape stubbles, whilst turnip fields accounted for almost half of flocks and almosta third of birds. The only other habitats from which they were recorded were barley and oat stubbles, uncut oats, fodder rape, setaside and weedy potato fields. There had been difficulties with the harvest of crops, especially rape, in Caithness in 1998 because of bad weather, and much rape seed had been spilt from the seed pods before they could be gathered in. Furthermore water logging of the ground had prevented many of the rape and other stubbles being ploughed in, such that by the time of our Survey there was much more food available for the birds than was usual. Asin December 1992, another year when there was a superabundance of rape seed, the number of birds found was at the upper end of the range of variation so far recorded, and the results reaffirm the importance of rape seed to Twites and that the number of birds in Caithness in mid-winter is related to the abundance of food (cf Table 2). Twites wintering in Caithness originate from Caithness itself, the NW Highlands, the northern part of the Western Isles and the Northern Isles (Clark & Sellers 1998, Movements of Twites in Scotland, Scottish Birds 19: 270-279). We suspect that the variable numbers in Caithness are a result of passage birds from the Northern Isles, and possibly those from the Caithness breeding population, staying in Caithness when food Table 2 Winter numbers of Twites in Caithness in relation to availability of rape. Date No of birds 1 14-15 Nov 1992 2368? 2 28-29 Dec 1992 6882 3 4-5 Dec 1993 2136 4 3-4 Dec 1994 3036 5 29-30 Dec 1994 2873 6 12-13 Dec 1998 5927 a No of Availability flocks of rape 19 high 20 high 26 low 30 low 19 low 43 high This count was made before all wintering birds had arrived or passage was complete Scottish Birds (1999) is abundant, but moving further south when itis not. It follows that surveys such as ours are at best only a very crude indicator of the general fortunes of Twites in Scotland; however, it seems reasonable to conclude _that there has been no marked reduction in the northern Scottish populations over the past 6 years. Short Notes S74 We are indebted to Peter Miller, Donald Omand and Julian Smith for their unflagging help with the surveys, and Mark Hancock, lan Bainbridge and RSPB Scotland for their encouragement and support throughout. Hugh Clark, 3 Lindsay Place, Wick, Caithness KW1 4PF Robin M Sellers, Crag House, Ellerslie Park, Gosforth, Cumbria CA20 1BL Revised manuscript accepted January 1999 Goldcrest nest sites in witches’ brooms The nests of Goldcrests Regulus regulus are typically suspended in twigs near the end of conifer branches but, where typical sites are not available, in the forks of branches (Birds of the Western Palearctic vol. V1 p.682). In north west Sutherland there are no coniferous woods north of Laxford Bridge and Gobernuisgach Lodge except for small plantations at Eriboll Lodge, Inverhope and Durness. There are old and often degenerate Birch Betula woods along sheltered glens, ~ eg in Strath Beag, Strath More and along the shores of Loch Hope and Loch Eriboll; otherwise the land is virtually treeless. Goldcrests are uncommon (Gibbons et al/ 1993 The New Atlas of Breeding Birds in Britain and Ireland: 1988-1991, Poyser) in what appears to be an unsuitable habitat, though in Sutherland generally they are described as “usually in good numbers, breeding in native and plantation woodland” (Angus 1993, Sutherland Birds Northern Times). At Eriboll in summer 1989 a Goldcrest was seen Carrying food to its young in a nest at 3m in a witch’s broom, a circular twiggy outgrowth on the branches up to 1m in diameter but more usually 0.5m, in an 11m high old Birch tree in a small area of Birch and Willow Salix. Droppings at the nest entrance in February 1990 suggested roosting by Goldcrests or Wrens Troglodytes troglodytes. Two further Goldcrest nests were found in witches’ brooms in Birch in 1989, one at Eriboll and one near Loch Loyal. In March 1990 23 Birches with witches’ brooms large enough to contain Goldcrest nests were examined at Eriboll. Eleven trees were in a gully, widely spaced, in a grazed pasture; 12 in a small old birch wood. In a witch’s broom in the isolated trees there was an old Chaffinch Fringilla coelebs nest and in the wood single nests of Chaffinch and Wren or Goldcrest. Casual searches subsequently produced an old Song Thrush Turdus philomelos nest 1.3m above the ground in a witch’s broom on a substantial limb of an old Birch and an unlined Wren nest at 2m, again in an old Birch. 38 Short Notes Ten birch trees holding 24 witches’ brooms, 19 of them big enough to contain a Goldcrest’s nest, were examined at Laxford Bridge in 1992-93. Only one partly built but unidentifiable nest was found. Further casual searches revealed no Goldcrest nests in Birch. This was to be expected as there were conifer plantations in the area. However, a partially built Goldcrest’s nest was found on 25 June 1997 in a witch’s broom in a Birch in a small native wood; the SB 20(1) young fledged in the first week of August. A large area of conifer nearby had been recently felled and this may have led to the use of Birch for nesting. Goldcrest nests in witches’ brooms were suspended beneath slender branches similar to the outer branches of conifers. Those of Chaffinch and Song Thrush were in witches’ brooms growing on stouter branches and were on the upper side of these. Ray Hewson, Department of Zoology, University of Aberdeen, Aberdeen AB24 2TZ Tom Talbot, Laxford Cottage, Laxford Bridge, Sutherland, 1V27 4SH Accepted January 1999 Goosander killing a frog At 12.15 hours on Monday, 15 February 1999 | was in the hide at a small pond adjacent to the loch at the James Hamilton Heritage Park in East Kilbride. A male Goosander Mergus merganser suddenly surfaced immediately in front of the hide holding a large frog Rana sp. By the large size of the amphibian | took it to be a female. The Goosander held the frog by one thigh and proceeded to swing it round and round in a circular fashion in what | presume was an attempt to subdue it. After a few seconds of this treatment the Goosander deftly turned the frog and swallowed it head first with apparentease. | canfindno specific mention of frogs as a food for Goosanders in any of the current ornithological literature. One gets the impression that they feed almost exclusively on members of the salmon family. Frances Gatens, 94 Telford Road, East Kilbride G75 OBX Revised manuscript accepted March 1999 Dr M Marquiss has commented: This note is interesting because it draws attention to the varied diet of these ducks. Though not reported in the general bird literature, to judge from the contents of Goosander stomachs(Marquiss et al 1998 Fish eating birds and salmonids in Scotland 156 pp. The Scottish Office, Edinburgh) it is clear that frogs are regularly taken in February and March when many Goosanders switch from feeding on rivers to still waters (Marquiss & Duncan 1994 Seasonal switching between habitats and changes in abundance of Goosanders Mergus merganser within a Scottish river system. Wildfowl 45, 198-208). Scottish Birds (1999) Short Notes 39 Merlin’s sunning behaviour in summer | read with interest the note (Dickson 1998 Scottish Birds 19:176) on the sunning behaviour of Merlins Falco columbarius in winter. Apparenily information onthe sunning behaviour of Merlins in summer in Britain is generally lacking and so | have been encouraged by R C Dickson to submit this note. On 5 May 19839, | visited a Merlin’s nest site in Dumfriesshire on a bright, warm, sunny day. The female was apparently nest selecting and attimes nest scraped at various places on a steep heather clad hillside, and on one occasion she was accompanied in this by the adultmale. Onseveral occasions, however, she lay spreadeagled on top of the heather with fully outstretched wings and partly fanned tail, sunning, sometimes for up to a minute before changing position. At other times she sat erect and faced the sun as if sun basking, behaviour which | have often observed in female Merlins prior to commencement of laying. Some falcons will spread their wings fully while sunning in the prone position on the ground, but apparenily this has only been recorded hitherto of birds in captivity (Simmons 1986 The Sunning Behaviour of Birds, Bristol). Dickson (1995 Scottish Birds 18: 58-59) has recorded the full spread open wing sunning position on the ground by Peregrine Falcons Falco peregrinusin winter in the wild, but the full spread open wing sunning position on the ground by Merlins in summer has not been documented previously. C J Rollie, 22 Main Street, St John’s Town of Dalry, Galloway DG7 3UW Hen Merlin Accepted April 1999 40 Short Notes SB 20(1) Interlocking of talons between Common Buzzards Further to R C Dickson (Scottish Birds 19:166), | have also observed Buzzards Buteo buteointerlocking talons during display manoeuvres. However, in this case no food was passed suggesting this was an aggressive action. On6 March 1999 | noted 5 Buzzards soaring over an area approximately 1°km? of farmland and woodland beside Auchendores Reservoir near Kilmacolm, Inverclyde. From aroadside vantagepoint | watched 2 adult birds soaring in close proximity, which then unexpectedly flew steadily towards each other, interlocked talons and began cartwheeling, completing several turns, and dropping an estimated 10 to 15m in altitude from a soaring height in the region of at least 40m above ground level. After this encounter, one of the birds flew a comparatively short distance westward and Grey Heron killed by Great Skua At about 16.45 hours on 1 October 1998, whilst scanning the Bay of Firth at Finstown, Mainland, Orkney, | observed a Bonxie Catharacta skua cruising around the bay, at a height of about 19 metres. It made 2 complete circles of the inner bay and appeared to be looking for food; there were a few ducks, waders and gulls around. Since 28 September | had noticed a number of local Grey Herons Ardea cinerea which had the habit of flying across the bay. As the Bonxie was cruising about, | saw one of the herons, which had a white crown and was therefore an adult bird, flying slowly across the bay some 2m above the water. | was disappeared over a small belt of woodland. The other individual continued soaring in an eastward direction towards a slope covered in Hawthorn Crategus monogyna and Gorse Ulex euopaeus scrub interspersed with various tree species, where at least one juvenile Buzzard was calling noisily. In an area in which Buzzards are numerous, this is the first occasion on which | have noted this particular behaviour. J Towill, 82 Sempill Avenue, Bargarran, Erskine, Renfrewshire PA8 6DG Accepted April 1999 R C Dickson has commented that aggres- sive talon grappling is quite common in terri- torial Buzzards but cartwheeling has rarely been reported in Britain though a record of a courting pair cartwheeling is in Bird Study 1975 (22:261). amazed to see the Bonxie approaching it from above and behind, unnoticed, dive directly onto its back and knock the heron into the water, whereupon the Bonxie moved its feet onto the heron’s head and forced it under the water until it drowned. During this time the Bonxie continued to remain air borne flapping its wings to gain the necessary force to keep the heron’s head under the water. Having killed the heron, which appeared to be healthy insofar as its ability to fly was concerned, the Bonxie spent about 10 minutes trying to turn the corpse over onto its back. Itthen proceeded to pluck and eat the largerbird. Thistook some hours to complete before the skua finally and rather ponderously Scottish Birds (1999) took off and flew off in the direction of Hoy. Whilst | have occasionally seen skuas and indeed large gulls attack herons, | have never seen them killed in so efficient and rapid a manner as | witnessed that day. D H Lawson, 25 Kirkland Avenue, Blanefield, Glasgow G63 9BY Revised manuscript accepted May 1999 Pomarine Skua kleptoparasitising an inland gull roost Winter records of Pomarine Skua Stercorarius pomarinus in Scotland are scarce (Thom 1986 Birds in Scotland, Poyser, Furness 1987 The Skuas, Poyser) but showing increasing frequency in recent years (Scottish Bird Reports). Inland records are also rare (Thom 1986), especially away from autumnal gales, and long staying individuals inland are largely unknown. lt came as a great surprise when a first winter bird was discovered in the Baron’s Haugh and Strathclyde Park area of Lanarkshire on 31 December 1994 and stayed until 23 January 1995. This represented the first and only Lanarkshire record to date. With the exception of the first sighting at Barons Haugh, the bird was not seen away from Strathclyde Loch (an 8tha artificial loch situated between Hamilton and Motherwell) throughout the rest of its stay. Throughout much of the day the bird rested on Strathclyde Loch, rarely flying, presumably to conserve energy. Most of its activity was noted during the last 2 hours or so of daylight. At this time gulls from a wide area of central Lanarkshire converged upon the loch to roost. Short Notes 41 Professor R W Furness has commented that there are previous records of Great Skuas killing herons (Furness, R W 1979. Foods of Great Skuas Catharacta skua at North Atlantic breeding localities. Ibis 121: 86— 92). However, there are few descriptions of how such kills are made and this is a nice description of the method. My experience of herons in Foula is that most are juvenile birds that tend to weaken and are killed by skuas as they near death by starvation. The most numerous species were Herring Gull Larus argentatus (up to 10,000) and Black-headed Gull Larus ridibundus (up to 5,000) with smaller numbers of Common Gull Larus canus, Lesser Black-backed Gull Larus fuscus and Great Black-backed Gull Larus marinus also present. As the gulls flighted into the roost, the skua would take to the air and patrol up and down the loch, seeking out potential victims. As the gulls began to settle on the water the skua would fly towards the flock and flush them into the air. Once the flock was airborne, the skua would fly around underneath the flock to keep the birds aloft. It would then fly very fast almost vertically into and through the flock, executing a vertical stall followed by a high speed falcon like stoop and rapid pursuit behind the selected gull. The pursuit flights were variable in length but most lasted less than 30 seconds (based on approximately 60 chases observed over 5 different evenings). Although successful pursuits were recorded against all5 gull species present, most (>80%) were against Herring Gulls. Approximately 20-25% of pursuits were successful in forcing the gull to regurgitate its last meal. Once the gull had regurgitated, the skua quickly fell to 42 Short Notes SB 20(1) the water surface to retrieve its reward, and alighted on the water to feed. The gulls then settled on the water again, but leaving a large space around the feeding skua, approximately 100m in diameter. Once finished feeding the skua would take to the air and flush the roosting gulls again until the daylight faded. The skua ceased activity around one hour after sunset. During the 90-120 minutes of activity each evening the skua would secure at least 5 or 6 regurgitated meals. The skua eventually roosted on the loch with the gulls again giving it a wide berth of approximately 100metres. This level of disturbance was clearly unsettling to the gulls arriving to roost and, during the course of the skua’s stay, the number of Herring Gulls using the loch to roost fell steadily from approximately 10,000 to about 1,500, whereas at this time of year gull numbers usually remain high at this site. Furness (1987) records kleptoparasitism as a minor feeding strategy in Pomarine Skuas, compared to Arctic or Great Skuas. He comments ‘Pomarine Skuas which winter in the warm, but rich upwelling areas, off West Africa, are thought to feed largely by predation of small seabirds such as phalaropes, and by scavenging’. Kleptoparasitism of the gull roost appeared to be a successful short term strategy for a young skua inland in mid winter. However, it was clearly unsustainable, in that the prey species, Herring Gull, quickly learned to avoid the skua, with approximately 85% switching to use alternative roosting sites during the duration of its stay. Thanks to Bob Furness, lain Gibson and Stan da Prato who kindly read an earlier draft of this note. Chris Waltho, 73 Stewart Street, Carluke, Lanarkshire ML8 5BY Revised manuscript accepted June 1999 Rit To Pomarine Skua Angus Hogg Scottish Birds (1999) Large Kestrel clutch sizes in south west Scotland 1997-98 In 1997, Short-tailed Field Vole Microtus agrestis numbers peaked in Ayrshire as part of a regular 3 to 4 year cycle. The following season this occurred in West Galloway witha crash in vole number in late June. The abundance of voles as a prey item in both areas coincided with Kestrel Falco tinnunculus females laying large clutches and fledging large broods. The average clutch size in 1997 inthe Ayrshire study area was 5.8 (n=26), the highest recorded since work began in 1972. Eighteen nests contained clutches of 6 and one of 7 eggs. The average number of young produced per successful pair was 4.6 and one pair reared a brood of 7.1n 1998, asmaller sample of 5 pairs was monitored in West Galloway and proved to be very productive with an average Clutch size of 6.2 with one 7 and one 8. Fledging success was very high at 5.2 young per pair and again one pair reared 7 young. Instances of Kestrels in Britain laying 7 or 8 eggs are not common even when weather Short Notes 43 conditions are favourable and food plentiful. One other instance of 8 eggs being laid had been recorded in the Ayrshire study area in 1991. In the Birds of the Western Palearctic Vol 2, clutch size for Eurasian Kestrels is given as one to 7, and data from British Trust for Ornithology Kestrel nest record cards between 1950 and 1987 (Shrubb 1993, The Kestrel, Hamlyn) record only 3 clutches of 7 in Scotland and 8 clutches in England. Burton 1997 (Birds of Prey on Farmland The Raptor 24:16-20 Spring 1997) in southern England recorded only 4 broods of 7 in 14 years. By comparison, in a 12 year study in Finland, (Karen eta/ Hatching asynchrony in Eurasian Kestrels in relation to the abundance and predictability of cyclic prey, Journal of Animal Ecology 67:908-917) recorded 16 clutches of 7 and 3 of 8 eggs. In our study in south west Scotland both observers were confident that in each case of 8 eggs the record clutches had been laid by single hens. This was based upon regular visits to the nesting territories, seeing only one female, and also the fact thatthe markings on each clutch were relatively uniform in colouration and marking indicating that one female had laid the entire clutch. Gordon Riddle, Swinston, Culzean Country Park, Maybole, Ayrshire KA19 8JX Geoff Sheppard, The Roddens, Leswalt, Stranraer DG9 0QR Revised manuscript accepted June 1999 44 Short Notes SB 20(1) Recent fluctuations in a Common Sandpiper breeding population The Breeding Birds of south east Scotland : a Tetrad Atlas 1988-1994 (Murray, Holling, Dott & Vandome, 1998) states that the Common Sandpiper Actitis hypoleucos population in the Lammermuir Hills, SE Scotland had undergone a significant decline in recent years, and this prompted T W Dougall to analyse his data for the nearby Moorfoot Hills from 1993 to 1998. This study was initiated largely to colour ring as many chicks as possible to complement the long term population studies in the Peak District by D W Yalden and colleagues, and to increase the national sample of the species ringed, surprisingly few of which are handled annually, given its abundance: 196 fully grown birds and 108 chicks in 1997. In 1998 A Mee began a research project involving the genetic fingerprinting of sub populations of TWD’s study population. Effort was concentrated along roadside stretches of the Blackhope and Dewar Waters near Garvald and Ladyside, and along the Glentress and Leithen Waters, where a car could be used as a hide, or where cover allowed a close approach to chicks. In addition, adults were mist netted in the early part of each season, and throughout the season in 1998 in AM’s sub populations using mist nets, nest traps and particularly effectively, baited spring traps. In this way, 10.13km of burns were sampled consistently over the 6 year period (AM’s additional length is not treated here). From observations of adult behaviour and the distribution of nests and broods, it was possible to estimate the number of territories held, and their success. For AM’s more intensive study in 1998, breeding territories were accurately defined using the territory mapping method (Table 1). In case the intensive effort of daily site visits by AM in 1998 introduced a bias, the data for this area only (as Surveyed in previous years to TWD) were compared on an annual basis with the data from the remainder of the monitoring sites. In each year except one, where there was no change for the remainder sites, the direction of population change was the same, although the magnitude differed. It seems likely that AM found more territories than TWD would have with his less intensive effort, but only in the region of 2 territories, one successful (ie for 1998 in Table 1, if TWD and not AM had been involved, n = 27 and successful = 19). Table 1 Common Sandpiper territories on monitored sites in the Moorfoot Hills, 1993-1998. Year n n/km successful 1993 31 3.06 23 1994 22 PATE 15 1995 25 2.47 14 1996 21 20% 14 1997 22 a7 15 1998 29 2.86 20 successful/km % successful PIT 74.19 1.48 68.18 1.38 56.00 1.38 66.67 1.48 68.18 1.97 68.97 Scottish Birds (1999) lt would appear, therefore, that over the 6 year period the Common Sandpiper population of part of the Moorfoot Hills fluctuated between 31 and 21 territories (3.06 and 2.07/km) overall, with the largest annual decrease of 29% between 1993-94 and the largest annual increase of 32% between 1997- 98. There is no indication of an overall decline or increase in the population over the study period. In the Peak District (sampling area = 10.05 linear km) DWY found that, over the same years, the population increased from6.5 to 14 territories (0.65 to 1.39 /km) but these figures were depressed following a population crash of 37.5% in 1988-89 due to severe weather in the spring of 1989, followed by seasons of poor recruitment. Territories from 1988-92 had been 20, 12.5, 13, 10 and 7. At the start of the Peak District study in 1979-82 there were 20-22 territories annually (1.99-2.19/ km, Holland & Yalden 1991 Bird Study38:151- 159). Both the Moorfoots and Peak District studies show that local Common Sandpiper populations can suffer large and sudden decreases, but can recover well over a few years if conditions improve, though what are presumably sub optimal sites can remain without sandpipers for a number of years. Nationally, the BTO’s Waterways Bird Survey (WBS) index values (Mountford method) for the years 1993-98 are, respectively, 93, 93, 82, 85, 88 and 86; these are part of a recent downward trend from peak values of 121 and Short Notes 45 125 in 1984 and 1985. However, it is thought that the 20 year trend is one of little change (Marchant, Wilson & Glue 1998 BTO News 216-217). DWY (in Gibbons, Reid & Chapman 1993 The New Breeding Atlas) has pointed out that most WBS census plots are not of optimum Common Sandpiper habitat as only 20% of plots and 19% total river length surveyed were fast flowing (Marchant ef al 1998) so the fluctuations and long term trend may not be representative of the whole Common Sandpiper population. There is thus perhaps a need for increased monitoring of upland riparian bird populatins, which could be addressed by more Breeding Bird Surveys (BBS) transects. 3 Although the overall population as expressed by territories / km, described in Table 1, ranges from 2.07 to 3.06, for the monitoring sites in the Garvald area only, it ranged from 2.80 to 4.80. The latter figure compares favourably with the maximum in any year of 4.70 reported from the prime stretch in the Peak District study area (Yalden 1986 Bird Study 33: 214-222) and would greatly exceed the Peak District figure if only the most productive stretch at Garvald was involved. We thank the BTO for supplying the national WBS indices. We are grateful for help in the field to Lynn Campbell, Alan Lauder, Tony O’Connorand W Underwood. Weare indebted to the various landowners and their agents for their interest andco operation with our studies. Especially Blackhope, Raeshaw, Rosebery Estates and M Cotterill, G Wainwright and the National Trust. AM’s study is funded by NERC and SNH. Tom W Dougall, Allan Mee and Derek W Yalden, c/o 62 Leamington Terrace, Edinburgh EH10 4JL Revised manuscript accepted June 1999 46 Obituary SB 20(1) OBITUARY Archibald Gordon Stuart Bryson 1912-1999 ‘Six keen naturalists met at 27 Inverleith Terrace to consider the proposal of starting a society for those interested in nature study and especially ornithology.’ These words, from the record of a meeting of 6 Edinburgh Academy schoolboys on 6 November 1929, signalled the start of a whole stream of organisations leading on, one from another, which have provided a framework for amateur ornithology in Scotland ever since. The meeting at the start of that extraordinary process was of the Inverleith Field Club and the list of its 6 members was headed by A G S Bryson, the others being H F D Elder, J H B Munro, H Simpson, J G Stewart and the Secretary, G Waterston. Members of the Club took to making excursions to Aberlady Bay or Portmore and Gladhouse, or the coast along to Cramond. Notes of these were kept in a Club diary, written in black ink with the bird names in red. There were no field guides and precious few other books to guide them. They had to find their own way, which, by degrees, they did. ‘12 January 1930. One rather outstanding observation this month was that of Bryson and Simpson who saw a Great Crested Grebe at Granton.’ Individualism was valued. ‘Rule 6. lfa Member discovers a Bird’s Nest, he is entitled to the sole proprietorship of the same, and Members wishing to photograph the nest or in any way molestit, must obtain the ‘Owner's’ permission.’ This exclusivity was extended to those beyond the Club and, while based on a genuine fear of egg collectors, was a characteristic which tinged some people’s perception of the group and its successor bodies. ‘Rule 7. Club members shall not divulge any information accumulated by other members of the Club to any person outside the Club.’ Such information came to include the discovery of a diver and chick at Arisaig by Archie in 1930, which, after a subsequent visit to the Royal Scottish Museum, he was convinced was the first breeding record for Scotland of the Great Northern Diver. By 1933, these teenagers were off into the wider world in their twenties and were more outward looking. Archie and the others resolved to found a larger group beyond the Academy, the Midlothian Ornithological Club, Scottish Birds (1999) 20:46-48 47 because ‘There are in Edinburgh, young energetic bird enthusiasts among whom there is little or no cooperation.’ And ‘The remarkable success of the Oxford Ornithological Society in accumulating scientific data on these lines (through corporate observation) is well known to all modern research workers’. Much effort in the Club was spent in arguing about subspecies, an obsession of the time, and also in field observation whenever time allowed. Archie, now an apprentice CA, could be spotted of a Saturday lunch time after leaving the office, taking the train to Cobbinshaw and birdwatching round the reservoir equipped with rolled umbrella, bowler hat and binoculars. Ringing birds was a favourite activity of the Club, but in those pre misinet days migrants were hard to catch and a suitable place had to be found for the construction of a Heligoland trap. There was an ideal site not far away at the mouth of the Firth of Forth, and the Isle of May Bird Observatory came into being the following year, founded by this same company of friends — the first cooperatively managed bird observatory in Britain. The object was to have as much fun as possible while contributing something to the sum total of scientific knowledge. It should be remembered that at this time the amount of professional field ornithology was almost nil. With only a small proportion of migrants caught in the single trap, identification of subspecies remained a vexing question and traditional means of obtaining specimens were not scorned. As Archie remarked laconically, though not without a touch of melodrama, in the Observatory log for 4 October 1935: ‘The Chiffchaff was shot at dawn. He was certainly not Siberian and his large measurements suggested Scandinavian. Fifty two species were seen during the day.’ The Observatory was closed during the war and reopened after it, with Archie giving faithful service to its committee, including chairing it, for more than 30 years. Back on the island the fun continued. 21 May 1947. ‘JHBM (lan Munro) and AGSB spent some time attending the hens’ feeding times in the hope of rushing the House Sparrows into the Heligoland trap. Nil return.’ 14 May 1948. ‘A late drive of the trap resulted in a Whitethroat and a hen being incarcerated in the box together. While AGSB picked up the hen, JHBM was able to rescue the Whitethroat undamaged. It was felt that there was no interest or scientific value in a hen brooding a Whitethroat and the light was too short to allow us to see whether the Whitethroat would brood the hen.’ Two years after the start of the Isle of May Observatory, members of the MOC were instrumental, with others, in setting up the Scottish Ornithologists’ Club in 1936. Archie served the Club loyally as its Honorary Treasurer from 1946 to 1958 and as a Council member until 1963. In 1935, he had visited Fair Isle with George Waterston and discerned its potential as a bird observatory site, even better than the Isle of May. Both of them were back there in 1946, George brooding once again about the bird observatory which actually came into being 2 years later. Archie was one of its Trustees for nearly 40 years until 1985. Nor did the repercusssions of the early days of the Inverleith Field Club end there. In 1964 7 people, atleast 5 of them members of the SOC, set up the Scottish Wildlife Trust. 48 Obituary As well as studying birds Archie had to earn a living. Following his father into the profession, he qualified as a CA in 1935 and in that year went to join his Uncle David in the family firm, Blackwood Bryson and Company, in Calcutta. He remained in India during the war, serving in the Royal Indian Navy Volunteer Reserve as a signals instructor in Bombay and as astaff officer in Delhi. There he welcomed as a guest on leave his fellow MOC member Donald Watson, knowing that he was posted to the Burma front but unable to tell him. After the war he returned to Edinburgh, joining his father’s firm, and taking over from him a place on the Boards of the Royal Blind Asylum and the Scottish National Institution for the War Blinded, where he served for some 37 years. His father was the first Auditor of the National Trust for Scotland and Archie gradually took over the Audit. His firm continued as Auditors until 1963, by which time they had held the appointment for 32 years, relinquishing it only when the growing task became too large for the resources of the firm. Back in the voluntary sector, Archie had great sympathy for the work of the SSPCA and served on its board of Directors from 1953 to 1987. His interest in ornithology continued unabated and he was elected to the British Ornithologists’ Union in 1948. He served on its committee and became its SB 20(1) Honorary Treasurer in 1956, continuing in that office for 20 years. The Minute of the AGM at which he retired records that he had ‘meticulously, but often in new and imaginative ways, controlled and made the best possible use of the funds of the Union , which had benefited enormously from his advice and skill.’ That advice and skill characterised his whole life, to the great benefit of many individuals as well as organisations. But the cautious and perceptive precision of ‘a typical Edinburgh CA’ as one of his MOC colleagues described him, should be seen in the context of a warm personality full of sparkling good humour. Known as The Thresher in his young days for his dynamic enthusiasm and leadership, he put these boisterous days behind him. He became a quiet, reassuring presence, never pushing himself forward, but as an ace committee man influencing many organisations in their best interests without ever seeking the limelight for himself. He was a passionate gardener, with a keen eye for alpines and Rhododendrons. Buthis greatest love was for his family, his wife Eleanor and his daughter Catherine and granddaughter Mary. Archie was one of the real originals of field ornithology in Scotland, an amateur naturalist in the best sense of the term. We mourn his passing and we salute his contribution to the legacy which all of us enjoy. John Arnott Advice to contributors Authors should bear in mind that only a small proportion of the Scottish Birds readership are scientists, and should aim to present their material concisely, interestingly and clearly. Unfamiliar technical terms and symbols should be avoided wherever possible and, if deemed essential, should be explained. Supporting statistics should be kept to a minimum. All papers and short notes are accepted on the understanding that they have not been offered for publication elsewhere and that they will be subject to editing. Papers will be acknowledged on receipt and are normally reviewed by at least 2 members of the editorial panel and, in most cases, also by an independent referee. They will normally be published in order of acceptance of fully revised manuscripts. The editor will be happy to advise authors on the preparation of papers. Reference should be made to the most recent issues of Scottish Birds for guidance on style of presentation, use of capitals, form of references, etc. Papers should be typed on one side of the paper only, double spaced and with wide margins and of good quality; 2 copies are required and the author should also retain one. We are happy to accept papers on computer discs; however, please state the type of word processing programme used. Contact Sylvia Laing on 0131 556 6042 if you wish further information on this. Headings should not be underlined, nor typed entirely in capitals. Scientific names in italics should normally follow the first text reference to each species unless all can be incorporated into a table. Names of birds should follow the official Scottish list (Scottish Birds 1994 Vol 17:146-159). Only single quotation marks should be used throughout. Numbers should be written as numerals except for one and the start of sentences. Avoid hyphens except where essential eg in bird names. Dates should be written:...on 5 August I9Ote but not on the 5th (if the name of the month does not follow). Please do not use headers, footers and page numbers. Please note that papers shorter than c700 words will be treated as short notes, where all references should be incorporated into the text, and not listed at the end, as in full papers. Tables, maps and diagrams should be designed to fit either a single column or the full page width. Tables should be self explanatory and headings should be kept as simple as possible, with footnotes used to provide extra details where necessary. Each table should be on a separate sheet. Maps and diagrams should be either good quality computer print outs in black and white (please do not use greyscale shading) or in black ink and be camera ready, but drawn so as to permit reduction from their original size. Scottish Birds Volume 20 Part 1 September 1999 Contents Main papers The breeding birds of Auskerry, Orkney, 1969-1998. RG Adam & C J Booth 1 The use, abuse and misuse of crow cage traps in Scotland: a report on behalf of the Scottish Raptor Study Groups and the RSPB. D Dick & A Stronach 6 The breeding status of the Spotted Crake in north east Scotland. | Francis & A Thorpe 14 Changes in the numbers and distribution of Mute Swans in the Lothians in spring from 1978 to 1998. A W Brown & L M Brown 18 Short notes Swifts nesting in Scots Pines at Abernethy Forest, Strathspey. R Summers 27 Size of Merlin pellets in winter and summer in Galloway. R C Dickson 31 Corn Bunting decline in Easter Ross. A Watson, R L Swann, H McGhie & M Nethersole- Thompson 34 A further survey of Twites wintering in Caithness. H Clark & R Sellers 35 Goldcrest nest sites in witches’ brooms. R Hewson & T Talbot o/ Goosander killing a frog. F Gatens 38 Merlin’s sunning behaviour in summer. C J Rollie 39 Interlocking of talons between Common Buzzards. J Towill 40 Grey Heron killed by Great Skua. DH Lawson | 40 Pomarine Skua kleptoparasitising an inland gull roost. C Waltho 41 Large Kestrel clutch sizes in south west Scotland 1997-98. G Riddle & G Sheppard 43 Recent fluctuations in a Common Sandpiper breeding population. T W Dougall 44 Obituary Archibald G S Bryson. J Arnott 46 Advice to contributors Inside back cover Published by the Scottish Ornithologists’ Club, 21 Regent Terrace, Edinburgh EH7 5BT. © 1999 — LY 0 ey SY i, Bid 0000 90 HV Vol. 20 No. 2 ISSN 0036 9144 Scottish Birds The Journal of the Scottish Ornithologists’ Club Editor: DrS da Prato Assisted by: Professor D Jenkins, Dr J B Nelson, Dr | Bainbridge, Dr M Marquiss and R Swann Business Editor: The Secretary SOC, 21 Regent Terrace Edinburgh EH7 5BT (tel 0131-556 6042, fax 0131 558 9947). Scottish Birds, the official journal of the Scottish Ornithologists’ Club, publishes original material relating to ornithology in Scotland. Papers and notes should be sent to The Editor, Scottish Birds, 21 Regent Terrace, Edinburgh EH7 5BT. Two issues of Scottish Birds are published each year, in June and in December. Scottish Birds is issued free to members of the Scottish Ornithologists’ Club, who also receive the quarterly newsletter Scottish Bird News, the annual Scottish Bird Report and the annual Raptor round up. These are available to Institutions at a subscription rate (1997) of £36. The Scottish Ornithologists’ Club was formed in 1936 to encourage all aspects of ornithology in Scotland. It has local branches which meet in Aberdeen, Ayr, the Borders, Dumfries, Dundee, Edinburgh, Glasgow, Inverness, New Galloway, Orkney, St Andrews, Stirling, Stranraer and Thurso, each with its own programme of field meetings and winter lectures. The Waterston Library at the Club’s headquarters at 21 Regent Terrace, Edinburgh EH7 5BT is the most comprehensive ornithological library in Scotland, and is available for reference during office hours (Monday to Friday, 0930-1630 hrs). (But please phone beforehand). SOC annual membership subscription rates Direct Debit Other methods Adult £18.00 £20.00 Family (2 adults and any children under 18) living at one address £27.00 £30.00 Junior (under 18, or student under 25) £7.00 £8.00 Pensioner/unwaged £10.00 £11.00 Pensioner Family (2 adults living at one address) £14.50 £16.00 Life £360.00 £400.00 Life Family £540.00 £600.00 All subscriptions may be paid by Direct Debit and Covenanted. Subscriptions paid by Direct Debit greatly assist the Club. Please ask for a Direct Debit form by phoning the Secretary at the above address. Published by the Scottish Ornithologists’ Club, 21 Regent Terrace, Edinburgh EH7 5BT Printed by Meigle Print, Block 11, Units 1 & 2, Tweedbank Industrial Estate, Galashiels TD1 3RS Scottish Birds (1999) 20:49-62 49 Raven population size and distribution on the Isle of Islay in winter M MADDERS & F M LECKIE The numbers and distribution of Ravens was studied on the Isle of Islay between October 1996 and April 1997. Survey work identified 35 Raven territories, of which 32 were occupied by breeding pairs in April. Territorial pairs were located in 2 additional areas where no suitable nest sites were apparent. Most territories were located around the coast, with the main concentration in the west and southwest, notably The Oa. The estimated breeding density across the whole island was 0.04 pairs km’ ie close to the Scoitish mean. Eight territories were monitored regularly throughout the winter. These territories were occupied continuously by Raven pairs. The estimated size of the nonterritorial population, based on counts of Ravens at a communal roost, varied from 149 to 285 birds. Daytime sightings of Raven were clustered around 2 axes in the centre and northwest of the island. These axes intersected at the communal roost. Raven abundance in different parts of the island was investigated in relation to various measures of topography and habitat. When distance from the roost was taken into account, Raven distribution was best explained by the cover of improved grassland and proximity to the municipal refuse tip. Introduction In recent years, licences have been issued to shoot limited numbers of Ravens Corvus corax on the Isle of Islay, in response to a perceived increase in the Raven population and because they are widely believed to damage and kill domestic stock. This situation has given rise to concern among some conservationists that our knowledge of Raven ecology on Islay is insufficient for the effects of licensed killing on the Raven population to be predicted. This paper presents the results of research to estimate the size of the breeding and non breeding Raven population in 1996-7. A further aim of the study was to map the distribution of Ravens in winter, and relate Raven abundance to various measures of land cover and terrain. Islay is located some 24km west of the Kintyre peninsula and covers approximately 615km? (Fig 1). Inland, it is characterised by unproductive bogs and mires, heaths and rough grasslands, improved and semi improved grasslands, arable habitatand even aged conifer plantations. The coastline measures c220km and comprises mainly steep rocky cliffs, dunelands and intertidal habitats. The island is exposed and affected by a maritime climate (Boyd 1983). Methods Territorial Raven population size Systematic searches for territorial Raven pairs were carried out in October 1996 and April 1997. The whole coastline of Islay was walked during each survey period. All inland 50 M Madders & F M Leckie SB 20(2) crags, steep gullies and stands of mature woodland marked on 1:25000 scale OS maps were visited during April. Information on areas where breeding had been proven or suspected was sought from _ local ornithologists, land managers and stalkers. These areas were visited in both October and April. Details of the number, location and behaviour of all Ravens seen were plotted on to 1:25000 scale maps. Crags and suitable trees (>3m tall with a trunk diameter at breast height > 0.5m) were scanned carefully for Raven nests. Territories were defined as follows: Breeding territories. Areas containing a nest, or cluster of nests, that were occupied by a pair of Ravens. Apparently vacant territories. Areas containing one or more nests but which were apparently unoccupied by Ravens. Non breeding territories. Areas where no nests were found in which territorial behaviour by a pair of Ravens was observed. Follow up visits were made to all apparently vacant and non breeding territories, in order to confirm status. To test whether pairs roosted within their territories at night, a random sample of 8 breeding territories identified in October 1996 were visited at monthly intervals Figure 1 Map of the Isle of Islay, showing place names mentioned in the text. —~3 em. a f F 7 j iA c ( pat ; \ aN —* } : . ia Port Askaig f f { & ( Saligo : Ballygrant » ie \ iy Bridgend \ Bruichladdich ; » X ( dees \ Port Charlotte, _-{ Bowmore } STAR ee a % RINNS / Cy ) a ¥: \, Ardtalla f = 4 ee é (7 < or \ Ss ee : Portnahaven ~~ \ S 7 Vi ae / Port Ellen pte if Dp KS THE OA Ya ag aN > ne a [ ] Refuse tip MOkxs Scottish Birds (1999) Raven population size & distribution on the Isle of Islay in winter 51 until March 1997. During each visit, evidence of recent roosting (eg presence of Ravens at dawn / dusk, fresh faeces below roost ledges and branches) was recorded. Non territorial Raven population size The size of the non territorial Raven population was estimated from counts at a communal roost near Bridgend. A roost in this area had been known for at least 25 years. In 1996-97 it was located in tall mainly coniferous, trees at the edge of extensive policy woodland. Ravens were counted simultaneously from 2 vantage points (A and B), located 0.35km northwest and 1.25km south of the roost, respectively. Between them the vantage points afforded a comprehensive view of all the routes by which Ravens could enter or leave the roost area. Vantage point A was located at the same elevation as the tops of the trees used for roosting, whilst vantage point B was below tree top height. As a result, even quite low flying Ravens were silhouetted against the sky and therefore relatively easy to detect. Two evening counts and one dawn count per month were undertaken. Counts were carried out during nil precipitation in a variety of wind conditions. Dawn counts were undertaken on the first suitable morning following a dusk count. Data from the April dawn count was excluded from analysis because count accuracy was probably affected by human disturbance. Evening counts commenced at least one hour before sunset and continued until no further Raven movements could be detected. On arrival, the observer at vantage point A counted any Ravens already present within the roost. Morning counts commenced one hour before dawn and continued until all Ravens had apparently left the roost. The roost was then entered in order to check that no further birds remained. During counts, the number and flight direction of Ravens approaching or leaving the roost were recorded to the nearest minute. Observations from each vantage point were carefully synchronised so that the overall count could be adjusted to take account of any duplication. In cases where there was a discrepancy between observers in the number of Ravens involved in a particular movement, the higher figure was adopted. At dusk, Ravens that had flown to roost sometimes flew out again later, either to interact with incoming birds or forage briefly before returning to the roost. Birds leaving the roost in this way were deducted from the total. A number of additional sites where local information and our own observations suggested that Ravens might roost communally were checked for evidence of occupation (eg faeces under potential roost trees / ledges, birds gathering at dawn / dusk). We made several observations in the late afternoon and early evening from various vantage points on the east coast of Islay, in order to confirm that Ravens from Jura did not routinely fly to roost at Bridgend. Raven distribution Information on the distribution of Ravens during daylight was gathered between October 1996 and April 1997 during 2 weekly systematic counts of wintering geese carried out by Scottish Natural Heritage. These were conducted from vehicles by teams of 2 observers following fixed routes within 6 areas (Fig 2). Between them, the routes covered virtually all roads and driveable tracks on the island. Each route was covered 13-16 times. Counters were asked to record the number and location of all Ravens seen on 1:25000 scale maps. Sightings were later transferred 52 M Madders & F M Leckie SB 20(2) onto amap depicting the island as an array of grid cells, each cell representing 2x2km on the ground. Cells containing land which was >50% visible from the count routes (‘observable cells’) were determined in the field. The number of times each observable cell was surveyed (counting each cell once for every route) was calculated. Finally, an index of Raven abundance was determined, by dividing the number of Ravens recorded in each observable cell by the number of times that cell had been surveyed. Habitat data were taken from a map of land cover compiled by Lyndsey Kinnes (Scottish Natural Heritage) using 1:24000 scale aerial photographs (Scottish Development Dept 1988), provisional soil maps (Macaulay Land Use Research Institute), 1:25000 Ordnance Survey maps and field surveys. Within each observable grid cell, the cover of 6 habitat types was recorded by eye to the nearest 5 percent. The habitat types recorded were: Woodland. Heaths, bogs and montane habitats. Rough grasslands and bracken. Improved grasslands. Coastal dunes and cliffs. Other, including marshes anthropogenic features. CPN > CIN and In addition, 3 topographic variables that were considered to have a possible influence on Raven use were recorded, measured at the centre point of each cell. These were altitude (m), distance from the communal roost (km), Figure 2 Routes used to count wintering geese. Gruinart vs Kilmeny { NI Oa & Ardtalla Vi ae \ = = a a Scottish Birds (1999) Raven population size & distribution on the Isle of Islay in winter 53 and distance from the municipal refuse tip near Bowmore (km). Results Territorial Ravens A total of 32 breeding territories, 3 apparently vacant territories and 2 non breeding territories were identified. Incubation was confirmed in all 32 breeding territories. On the basis of these observations, the density of breeding Ravens on Islay was estimated to be 0.04 territorial pairs km. The distribution of breeding territories was predominantly coastal (Fig 3). Of 32 nests used in 1997, 75% were located on sea cliffs. The number of occupied nests per km of coast varied from 0.43 on The Oa to less than 0.07 along the south and north east coasts. Inland breeding territories were widely spaced. The mean nearest neighbour distance for 7 inland nests used in 1997 was 4.5km (range 2.4-8.2km). Five nests (16% of total) were located in trees, including Scots Pine, Beech and Oak. All tree nests were greater thanikm from the coast. Figure 3 Raven territories located on Islay in 1996-97. @ Occupied territory © Apparently vacant territory © Apparently territorial non-breeding pair 54 M Madders & F M Leckie SB 20(2) Figure 4 Numbers of Ravens entering the communal roost at dusk, 1996-97. 30 Count Of the 8 territories selected for monthly monitoring, 7 were used for roosting by pairs of Raven throughout the winter. The remaining territory was used until February, when the pair disappeared and could not be relocated nearby. This territory was subsequently classified as vacant. Non territorial Ravens The number of Ravens counted entering the communal roost at dusk ranged from 149 to 285 (median 223.5, interquartiles 180-258.5, n = 14). From Oct-Dec numbers fluctuated between 230-280 birds (Fig 4). Numbers declined sharply in January, and only 149 Raven were present by early February. Thereafter, numbers increased steadily to nearly 240 by late April. There was a high level of agreement in the number of Ravens counted at dusk and dawn in 5 out of 6 pairs of counts (Table 1). In the remaining pairing (Feb 9/10), 40% fewer Ravens were counted at dawn than the previous evening. However, it was strongly 0 = | | 200 - | é 100 | Oct Nov Dec Jan Feb Mar Apr suspected that some birds had already left the roost before dawn observations began. Following dawn observations, no unaccounted Ravens were flushed when the roost was entered to check for birds remaining inthe roost. Although this does not necessarily mean that all birds present in the roost at the beginning of dusk observations were visible from vantage point A, it is nevertheless reassuring. Prior to roosting, up to 150 Ravens usually gathered at a hill summit 0.5km NW of the roost. These birds tended to enter the roost after most other Ravens, often in groups of 5- 20 (exceptionally, 83). Of 1569 Ravens that flew directly to the roost, 49% comprised single birds and 37% were apparently paired. With the exception of a flock of 15 and another of 22, the remainder comprised groups of 10 or less. A consistently large proportion of birds approached the roost from the east, although a greater proportion arrived from the south and southwest on some count evenings (Fig 5). There was a statistically significant difference in the proportion of birds Scottish Birds (1999) Raven population size & distribution on the Isle of Islay in winter 55 Table 1 Numbers of communally Table 2 Mean number of Ravens (+SE) roosting Ravens recorded during pairs recorded from 6 routes used to count of counts made at dusk and dawn. wintering geese. Date Dusk Dawn _ Difference Route No of counts No of Ravens Al-23) OC» 256 25 i -2% Mean + SE 25-26 Nov 232 256 +10% 6-7 Dec 266 263 -1% Gorm 16 13.88 2.66 5-6 Jan 215 242 +12% Gruinart 13 11.54 3.29 9-10 Feb 149 90 -40% Kilmeny 16 inleOO 2198 10-11 Mar 181 170 -6% Laggan & Glen 15 26 Aan Gre2 Oa & Ardtalla 13 7.46 1259 Mean* 12% The Rinns 16 Sel 225 * Ignoring direction of difference soe Be ease ea Figure 5 Flight direction of Ravens observed entering the communal roost at dusk on 14 evenings. The compass rose shows the number of birds that approached from each cardinal compass point as a % of the total that flew direct to the roost. Black squares identify the group median. Other squares indicate the upper (grey) and lower (open) quartiles. N NVV NE SW bee S : 56 M Madders & F M Leckie SB 20(2) that approached from each cardinal compass direction (Kruskal-Wallis one way anova: X*_= 17.89, P = 0.012). Observations of Raven movements across the island early and late inthe day were consistent with the assumption that Bridgend was the only communal roost site. Distribution of Raven activity A total of 1,143 Ravens were recorded from the routes used to count wintering geese. Ravens were recorded most often from the Laggan and Glen route, and least often from the Oa and Ardtalla, and The Rinns routes (Table 2). However, there was no statistically significant difference between routes in the number of Ravens recorded per count (K-W one way anova: eS 8.38, PO: 1S6 sis) Of 203 2x2km grid cells containing land, 30% were classified as unobservable from the count routes and excluded from further analysis. These cells were mostly distributed in the northeast and southeast of the island and tended to be occupied by ground above 100m. Allbut 8 Ravens (0.6%) were recorded in grid cells classified as observable. Cells Figure 6 Relative abundance of Ravens recorded during 2 weekly counts of wintering geese, 1996-97. Areas marked “X” were not visible from count routes. 20 80 70 aa -+00®@ Vv on ° 30 40 Scottish Birds (1999) Raven population size & distribution on the Isle of Islay in winter 57 Table 3 Correlation coefficients between variables representing Raven abundance =and various measures of habitat and topography. Correlations with P<.001 are shown in bold. Variables were used in a multiple regression model! to predict Raven abundance (see Appendix). Altitude -0.19 Heath, bog & montane 0.00 +0.61 Improved grassland +0.42 -0.38 -0.37 Rough grassland +0.18 70.30 -0.18 +0.43 Woodland -0.08 -0.11 -0.17 +0.07 +0.20 Roost distance -0.31 +0.10 . -0.09 -0.30 -0.16 -0.04 Tip distance -0.35 +0.24 +0.01 -0.36 -0.10 -0.05 +0.55 Index of Raven Altitude Heath Improved Rough Wood Roost abundance bog & grassland grass -land distance montane -land tT Log transformed. Table 4 Differences in habitat and terrain between 2x2km cells where Ravens were recorded, and where they were not. The table shows mean values (+SE). Ravens recorded Ravens not recorded Mean + SE Mean et Percentage cover of: Woodland 8.1 LES: 9.6 1.9 Heath/bog 30.6 72208 38.2 4.6 Rough grassland 13.9 10. tite? i et Improved grass 18.5 1-3: 6.3 9, Other habitats a5 0.8 7.0 1 a Altitude (m) af 71 5 1 Distance to roost (km) tt.0 SEL 0.6 0.8 Distance to refuse tip (km) 10.1 13e1 0.4 0.7 with most Raven sightings were distributed along 2 broad axes. These extended from Bridgend towards Saligo, and from the refuse tip towards Port Askaig, respectively (Fig 6). The axes intersected at the communal roost. This analysis indicated that Ravens were least abundant in the south of Islay and The Rinns. Habitat and topographic variables differed between the observed cells in which Ravens were recorded and those in which they were not (Table 4). These differences were Statistically significant in respect of improved grassland (Mann-Whitney U = 834.5, P < 0.0001) and distance from the refuse tip and roost (Student's t= 3.68, df. 139, P <0 .001 and t= 1.94, df. 139, P = 0.05, respectively). Other variables did not differ significantly between the 2 groups of cells (altitude: U = 1179.5, P = 0.41: woodland: U= 1917.5, P= 0.25; heath /bog /montane: U = 1920.5, P = 0.29; rough grassland: U= 1933.5, P =0.31; other: U= 1891.5, P = 0.22). 58 M Madders & F M Leckie SB 20(2) After controlling for the effects of roost distance, variables representing the percentage cover of improved grassland and distance from the refuse tip were each found to have a significant effect on the number of Ravens recorded (See Appendix). The other variables tested did not have significant effects. Improved grassland was negatively correlated with roost and tip distance (Table 3). Despite this, improved grassland had a positive effect on the Raven index, whereas roost and tip distance each had negative effects. This suggests that the cover of improved grassland had a greater effect on Raven abundance as distance from the roost and tip increased. Taken together, the cover of improved grassland and distances from the roostand tip were relatively poor predictors of Raven abundance, between them accounting for only around one quarter of the variability in the Raven index. Limitations of analysis Few areas above 300m were visible from the count routes. It is possible that these areas were occupied regularly by Ravens that were not recorded. However, there is circumstantial evidence to suggest that this was not the case. Firstly, relatively few Ravens were seen flying over the skyline of ground within grid cells classified as unobservable. Second, few Ravens were encountered in upland areas during systematic searches for territorial pairs in October and April. Anecdotal information from stalkers and shepherds suggested that groups of up to 25 Ravens often made use of ephemeral food sources such as deer grallochs and carcasses, but that these birds were not present continuously. Goose counters spent more time observing areas of improved grassland than other habitats, where there were fewer geese. It might therefore be expected that they were more likely to detect Ravens in improved grassland. Furthermore, an individual Raven may have been counted twice if it moved from view at one counting station to reappear at another counting station in the time it took to counta large goose flock. However, observers were highly experienced counters and aware of these potential sources of bias. Care was therefore taken to standardise search effort and take account of bird movements. Discussion The density of breeding Ravens on Islay does not appear to be exceptional in comparison with other parts of upland Britain. For example, Cross and Davis (1986) estimated there were 0.11 pairs km? in central Wales, ie 3 times the density on Islay. The density found in 11 Scottish studies reported by Ratcliffe (1997) ranged from 0.01 pairs km? in Sutherland to 0.10 pairs km? in Shetland, with a mean of 0.05 pairs km? (ie similar to that found in the present study). It is possible that we failed to identify some territories (especially inland), due to the difficulty of locating nests in wooded areas and the sometimes cryptic behaviour of territorial birds. However, all potentially Suitable habitat was searched thoroughly, and we therefore consider that few territories were overlooked. Further survey in 1999 identified 2 additonal breeding territories that may have been occupied in 1997. The large number of non territorial Ravens, many of which were paired and presumably seeking territories, suggests that the breeding population is close to the maximum that the island can support. The breeding population is probably limited by the availability of nest sites, especially inland where there are few trees or crags suitable for nesting. The Scottish Birds (1999) Raven population size & distribution on the Isle of Islay in winter 59 presence of 3 apparently vacant territories is consistent with studies elsewhere (eg Davis & Davis 1986, Ellis et a/ 1994), which have shown that all territories are seldom occupied in any one year. It is possible that the vacant territories found on Islay held insufficient food to maintain a resident pair of Ravens. This might be due to long term decline in territory quality, or to a temporary shortage of food in marginal territories. Alternatively, Ravens may have been victims of persecution by humans or territorial disputes with Golden Eagles Aquila chrysaetos. The latter is thought to account for the disappearance of Ravens in late winter from one of the 8 monitored territories. We found no evidence that territorial Ravens used the communal roost, or that there were other communal roosts on Islay. Numbers at the roost were lowest in late winter, when it might be expected that the population as a whole was at its annual minimum due to winter mortality. However, numbers increased in early spring, indicating that birds which had roosted elsewhere, on Islay or beyond, began to enter the communal roost. Some pairs may have attempted to set up territories in late winter, when carrion was presumably relatively abundant, and then returned to the communal roost during spring as they lost interest in territorial behaviour and food became more ephemeral. Ratcliffe (1997) noted that, in some districts, non territorial birds become more dispersed in winter, and that single birds and pairs sometimes detached themselves from the main group. It is possible that some Ravens used one or more temporary roosts elsewhere on Islay during late winter. Although we found no evidence to support this, such behaviour has been noted in previous years (see Argyll Bird Report 1991). Alternatively, some non territorial birds may have left the island in late winter, and returned, or were replaced gradually during spring. The latter hypothesis is consistent with the findings of studies in Maine, US (Marzluff et a/ 1996), where the Raven population was shown to be extremely fluid, with groups moving large distances to establish new roosts in response to changes in food availability. If this occurs in Scotland, non territorial Ravens on Islay might be part of a population that covers most of Argyll and includes parts of Northern Ireland. However, this would require frequent large scale movements across the open sea, and current evidence suggests that this only occurs sporadically (eg Elliot 1989). Roosts that consistently attract over 200 Ravens are scarce in Britain, although over 1,000 birds occupied one in north Wales in recent years (Ratcliffe 1997). In Argyll, roosts of up to 100 or so birds have been reported at Oban, Tobermory (Mull) and on the Kintyre peninsula (Argyll Bird Report 1987-92, 1996). Many of these roosts are associated with refuse tips. However, these roosts have a much larger hinterland from which to draw birds than the one at Bridgend. The size of the non territorial Raven population is of concern to stock managers, since the biras are potentially highly itinerant and communally roosting birds may share information on the location of food resources (Marzluff ef a/ 1996). Thus, it might be possible for large numbers of Ravens to rapidly locate temporally and spatially restricted food sources, such as deer grallochs in winter or vulnerable young stock in spring. Similarly, concern has been expressed in some quarters that Ravens might affect bird species perceived to be of high conservation importance (eg some wading birds and seabirds, Chough Pyrrhocorax pyrrhocorax, and Hen Harrier Circus cyaneus), through the predation of eggs and young, and 60 M Madders & F M Leckie SB 20(2) disruption of nesting and foraging behaviour. However, there is as yet no unequivocal evidence for this. It is important that losses of stock due to Ravens are rigorously investigated and quantified in order that current perceptions of predation, widespread among shepherds, can be tested. The non territorial population is large in relation to the size of the island, and these birds are probably most frequently involved in reports of stock losses. Ravens are wary birds and difficult to approach closely. As a result, licensed killing and illegal persecution may be directed mainly at territorial birds, which canbe ambushed more easily near their nest sites. Overall, the pattern of Raven distribution was associated with improved grasslands and distance from the refuse tip. The data suggest that territorial Ravens may have been relatively scarce in areas where overall abundance was greatest. This was probably because improved grasslands in general, and the refuse tip in particular, were located in low lying and treeless areas that provided few nesting opportunities. This is consistent with Davis & Davis (1986), who found that nonterritorial Ravens in central Wales tended to frequent places lacking nesting pairs. Non territorial Ravens may have simply avoided areas occupied by territorial pairs. However, we Saw groups of non breeding Ravens close to nesting pairs on several occasions, with the resident birds apparently tolerating this intrusion. Clearly, food resources are central to the question of why Ravens preferred improved grasslands. There are no published data on the diet of Ravens on Islay, and this should be a priority for future research. Elsewhere, Ravens feed on awide range of prey, including sheep carrion, Rabbit Oryctolagus cunniculus and hare spp, small mammals, birds and various invertebrates (Ratcliffe 1997). Studies of winter diet based on analysis of regurgitated pellets have shown a preponderance of sheep (eg Bolam 1913; Newton, Davis & Davis 1982; Ewins, Dymond & Marquiss 1986) or lagomorph prey (eg Mattingley 1995; Marquiss & Booth 1986). On Islay, sheep Carrion is more abundant in lowland pasture than elsewhere (D Carss, pers comm), and this is likely to be an important factor in explaining Raven distribution. It is also possible that foraging Ravens selectimproved grassland because invertebrate prey is abundant and carcasses of rabbits and hares are sometimes present. It is widely believed that Raven numbers have increased throughout west Scotland in recent years. Possible reasons for this include increased availability of sheep and deer Carrion, and increased scavenging Opportunities at refuse tips. There are no reliable estimates of the size of the Raven breeding population on Islay prior to our study, although a partial survey in 1994 located c 20 territories (J Gordon, perscomm). Previous ad hoc counts of Ravens roosting communally near Bridgend indicate that at least 100 birds have roosted there since the early 1970s (M A Ogilvie, pers comm). However, the difficulties involved in accurately counting Ravens using the roost mean that earlier counts may have under estimated the number of birds actually present. Consequently, there is no conclusive evidence that the Raven population on Islay has increased, although it may have done so. Conclusions This study shows that the wintering Raven population on Islay comprises at least 64 Scottish Birds (1999) Raven population size & distribution on the Isle of Islay in winter 61 territorial and up to 285 non territorial individuals. Territorial birds were mainly distributed around the coast, where nesting opportunities were greatest. Areas that were used most intensively by Ravens tended to be occupied by improved pasture habitats, or were located close to the refuse tip. Acknowledgements The data presented were gathered as part of a wider study of Raven ecology on Islay commissioned and funded by the Scottish Office Agriculture, Environment and Fisheries Department (SOAEFD) and Scottish Natural Heritage (SNH). We thank J Campbell (SOAEFD) and R McKenzie (SNH) for their support during this project. We are indebted to D Hayward, MA Ogilvie, C Cronin, R Lilley and the 1996-97 SNH goose counting team for assistance with fieldwork. Julia Welstead extracted the habitat data. The cooperation of Islay’s farmers, landowners, estate managers and _ staff is gratefully acknowledged. M Marquiss and an anonymous referee suggested significant improvements to an earlier draft. References Argyll Bird Reports 1987-92 & 1996. Argyll Bird Club. Bolam G 1913. Wildlife in Wales. Frank Palmer, London. Boyd J M 1983. Natural environment of the Inner Hebrides — an introduction. In Boyd J M & Bowes D R (eds) Natural environment of the Inner Hebrides. The Royal Society, Edinburgh. Cross AV & Daves P E 1986. Monitoring of Ravens and land use in central Wales. Unpublished report NCC. Davis P E & Davis J E 1986. The breeding biology of a Raven population in central Wales. Nature in Wales 3: 44-54. Elliot RE 1989. Birds of Islay. Christopher Helm, London. Ellis P M, Okill JD, Petrie GW & Suddaby D 1994. The breeding performance of Ravens from asample of nesting territories in Shetland during 1984-1993. Scottish Birds 17: 21-34. Ewins P J, Dymond J N, & Marquiss M 1986. The distribution, breeding and diet of Ravens Corvus corax in Shetland. Bird Study 33: 110-116. Marzluff J M, Heinrich B & Marzluff C S 1996. Raven roosts are mobile information centres. Animal Behaviour 51: 89-103. Mattingley W A 1995. Winter diet of Ravens in Perthshire. Scottish Birds 18: 71-77. Newton |, Davis PE & Davis JE 1982. Ravens and buzzards in relation to sheep farming and forestry in Wales. Journal of Applied Ecology 19: 681-706. Ratcliffe D A 1997. The Raven. T & A D Poyser. London Mike Madders, Carnduncan, Bridgend, Isle of Islay PA44 7PS Fiona M Leckie, ITE, Hill of Brathen, Banchory, Kincardineshire AB31 4BY Revised manuscript accepted August 1999 62 M Madders & F M Leckie SB 20(2) Appendix: Statistical regression of habitat and terrain variables on Raven abundance. Least squares multiple regression was used to explain variation in the index of Raven abundance?’ (the continuous dependent variable). The independent variables were the percentage cover of 5 habitat types, altitude, and distances to (a) the communal roost, and (b) the refuse tip (see Table 3). The variable representing roost distance was forced into the regression equation, regardless of whether or not it was significant. This was justified because it was expected to be a nuisance variable with an important effect on Raven distribution. Other variables were entered using a forward selection procedure, with an F to enter value of 0.05 (entry) and 0.10 (removal). Scatterplots of residuals were inspected in order to check the data met the necessary assumptions. The model selected 2 variables, representing distance to the refuse tip and percentage of improved grassland. Summary statistics are shown below. B Beta Ip Signi Roost distance -0.005488 -0.074586 -0.785 0.4337 Tip distance -0.020422 -0.222643 -2.3911 0.0181 % improved grassland 0.011366 0.340916 4.351 0.0000 Intercept 0.442828 4.192 0.0000 a = 0.23 Thus, log y = 0.44 = 0.01 roost distance __ 0.02 tip distance ae 0.01 improved grassland With the variables representing roost and tip distance in the equation, R=0.37,F = 11.16, P < 0.0001. After the addition of improved grassland, R=0.49,F =14.72"P < 0.0001. an 2 Log transformed. Raven Andy Dowell Scottish Birds (1999) 20:63-72 63 Barnacle Geese on the Solway: 1990-1996 JM BLACK"’, D PATTERSON”, P SHIMMINGS?° & E C REES? There have been exceptionally large counts of Barnacle Geese wintering on the Solway Firth since 1994. The highest figure comes from 2 coordinated counts carried out during the 1996-97 winter, which gave an average of 23,000 birds. The possibility that the “extra” birds may have come from one of the other Barnacle Goose populations is considered. Ring resightings did not provide evidence for substantial immigration, but a short term movement of birds to the Solway Firth may have gone undetected due to the difficulty of reading rings on Rockcliffe Marsh. Mass autumn migration has occurred earlier in recent years but Barnacle Goose counts, both on the Solway Firth and in northeast Scotland, also indicate that some geese arrive later in the season. Moreover, it is possible that birds are still present on uninhabited islands off the west coast of Norway in autumn and early winter. This suggests that recent censuses on the Solway Firth have taken place prior to the arrival of the entire Svalbard population. More frequent coordinated censuses are needed to monitor any build up in numbers during the winter, and to describe any short term immigration from other populations. Introduction Record numbers of Barnacle Geese Branta leucopsis were counted on the Solway Firth during the 1996-97 winter. Two coordinated counts, carried out in December 1996 and March 1997, found an average of 23,000 individuals, a substantially higher figure than the 18,100 birds recorded during a more extensive coordinated census of the same area in October 1996. The coordinated census in the previous year came to only 12,700 (on 20 October 1995), although other counts in autumn and later in the season gave higher figures (Table 1). Normally estimates of population size, as determined by the coordinated censuses, give much smaller differences between years, averaging 668 birds (SE 113, n = 24 years), and ranging from 100-2,100 individuals, for consecutive seasons up to and including the 1993-94 winter. The first major rise in numbers occurred between the 1993-94 and 1994-95 seasons, with population estimates of 14,350 and 17,900 geese respectively. The largest increase recorded before 1994-95, when numbers rose from 8,400 in 1984 to 10,500 in 1985, was attributable to a massive recruitment of first year birds; 26% of the population (2,730 of the 10,500 geese recorded in 1985) were juveniles. The increases in 1994 and 1996 could not be explained by similarly exceptional breeding seasons, however, nor by exceptionally high survival rates in these 2 years (see below). 64 JM Black, D Patterson, P Shimmings & E C Rees SB 20(2) Table 1 Svalbard Barnacle Goose population parameters on the wintering grounds, 1990- 1996. Asterisks denote calculated values (see text). Number of counts are indicated in parentheses. Final assessment indicates estimated population size determined by the coordinated censuses and coordinated counts. Year Autumnassessment Winter-Spring Final assessment Proportion (%)of assessment young: autumn (final) 1996-97 19,600" 23,000 (2) 23,000 Well (IS) 1995-96 13,645? (2) 17,445 (2) 20,450 15.7 (20.9) 1994-95 13,450? (3) 17,900 (2) 17,900 Tee) {Qs )}) 1993-94 14,350? (4) 10,900 (2) 14,350° 11.8 1992-93 13,200 9,140 (2) 13,200 5a 1991-92 13,300 10,100 (4) 13,300 14.0 1990-91 12,100' 9,900 (6) 12,100 12.0 ’ = Coordinated autumn census across entire Solway. In 1996-97, an additional 1,500 birds thought to be from the Svalbard population, counted in other parts of Scotland at the time of the autumn census, are included in the autumn figures. = Mean of coordinated census total and of other autumn counts that exceeded numbers recorded during the census; ie favours higher value. In 1995-96, an additional 530 birds thought to be from the Svalbard population, counted in other parts of Scotland at the time of the autumn census, are included in the autumn figures. 3= A further 119 Barnacle Geese were reported to have “wintered in Norway in 1993”; from Myklebust et al (1994). These were not included in the final assessment value. This paper documents the pattern of growth in the Svalbard Barnacle Goose population, and investigates the disparity in numbers in relation to the timing of the annual coordinated censuses, to determine if the censuses were being undertaken before all birds had arrived on the Solway Firth. It considers whether, as the Svalbard population grew, an increasing number of birds arrived later in the season, and also the possibility of immigration from one of the other Barnacle Goose populations in recent winters. Background and counting techniques The Wildfowl and Wetlands Trust (WWT) has been monitoring the size and breeding success of the Svalbard population of Barnacle Geese since 1959. Daily monitoring started in 1970, with twice daily counts from the tower at Eastpark Farm and annual co- ordinated censuses of the Solway Firth. The Svalbard population is one of 4 Branta leucopsis populations that occupy different breeding and wintering areas in western Europe. Its tendency to winter almost exclusively on the Solway Firth gives it one of the smallest ranges for any goose population in the world, making It easier to determine total population size through co- ordinated censuses of the birds in the wintering range. Disturbance and exploitation of geese on the Solway Firth, where the saltmarshes were used as a firing range, resulted in the Scottish Birds (1999) Barnacle Geese on the Solway: 1990-1996 65 Svalbard Barnacle Goose population declining in the early part of the century to only 300 birds by 1948 (Owen & Norderhaug 1977). The population responded to several conservation measures and increased in a step wise fashion over a 50 year period (Black 1998). Due to the increase in numbers, a proportion of the population has begun to nest and feed in areas that are beyond the limited traditional distribution (Owen et al 1987, Prestrud et a/1989, Black et a/ 1991). This has caused concern amongst local farming communities, both at staging sites and in the wintering range (Black 1998). Methods used to count the geese have varied slightly over the years, mainly in response to the increasing number of birds wintering on the Solway Firth. In the early years of the study, total population size was determined by a coordinated census in October. Initially this involved counting geese at Caerlaverock, then visiting other sites in the area to check for additional birds. More systematic coverage of other sites was made from the mid 1980s, however, following the increase in population size to > 10,000birds in 1984 (Black 1998). The census now involves around 12 people counting birds at all major sites on the Solway Firth within a one-hour time period, with any goose movement between neighbouring sections being recorded and taken into account when determining the total numbers present. This is considered the most accurate way of assessing population size from the ground. From 1994-95 onwards, less extensive but more frequent coordinated counts were undertaken, in addition to the coordinated censuses, to verify census results and monitor changes in distribution. These counts ideally require 5 or 6 people covering all major sites used by the geese within a 2 hour time period, but with some less frequently used areas being omitted. Thus coordinated counts may miss a few goose flocks and, since a longer time period is involved, are more susceptible to double counting. Coordinated counts so far have been rather accurate, however; counts made in autumn 1995 and autumn 1996 were, on average, within 4% of the coordinated census figures recorded the following day. The reassessment The disparity in numbers between the 1995/ 96 and 1996/97 seasons could not be explained by a good breeding year with a large recruitment of goslings (Table 1), nor to a substantial reduction in adult mortality, which is already low (<10% per annum, see Pettifor et a/ 1998). Two further explanations therefore were considered, which might both be influencing the numbers recorded: mass immigration from one of the other 3 Barnacle Goose populations, and the possibility that the coordinated censuses were being conducted before all birds had arrived on the Solway Firth. Potential immigration: the evidence Each of the Barnacle Goose populations has a sample of marked birds, making it possible to assess the level of movements of birds between populations. Since the early 1990s, WWT has recorded more than 25,000 ring resightings of over 3,000 individuals on an annual basis; some 95-98% of marked individuals identified in one season are resighted in the next. Ring reading effort is greatest in the wintering grounds, particularly at Caerlaverock and Southerness, but annual expeditions to the spring staging grounds and biannual trips to the breeding grounds also have been undertaken. 66 JM Black, D Patterson, P Shimmings & E C Rees The Russian and Baltic breeding populations, which winter in Germany and The Netherlands, were most recently estimated at 236,000 birds in January 1994 (B Gunter pers comm). Around 10,000 of these are from the Baltic breeding population, of which some 20% are ringed (K Larsson pers comm), giving c2,000 ringed birds in the Baltic breeding population in 1994. Of the remaining 226,000 Russian birds, approximately 89 should still be ringed, based on an annual mortality rate of 12% derived from 1978-84 data. Thus, only about 0.04% of the Russian birds were ringed in 1994. This gives an estimated 0.89% ringed birds for Barnacle Geese wintering in continental Europe. lf the 5,000 extra birds seen on the Solway Firth in 1996-97 was due to immigration of geese from the continent, we would have expected to detect c1,000 extra rings if movement was mainly from the Baltic breeding population, 2 rings if they came from Russia, or 57 rings if there was a homogeneous mix of the 2 populations in the wintering range. Given the high resighting rate for the 3,000+ ringed birds at Caerlaverock, it is likely that any new rings would have been read amongst geese using the reserve. Only 2 Baltic ringed birds were detected, however, and none of the birds ringed in Russia. The 2 Baltic breeding birds were first seen inthe Svalbard population during the 1992-93 winter, and have been recorded on the Solway Firth each year since then. No other Baltic ringed birds have been identified in the Svalbard population, even during the 1994-95 winter, when the first substantial increase innumbers was recorded. Studies of the Greenland breeding population of Barnacle Geese have included catching and ringing the geese wintering at RSPB SB 20(2) Gruinart, Islay, in recent years. By 1994, 5% of 7000 birds at RSPB Gruinart were ringed, together with 1-2% of c20,000 geese wintering elsewhere on Islay. About 1 % of the remainder of this population (c18,000 geese in western Scotland and Ireland) have also been fitted with plastic rings (S Percival pers comm). We therefore would have expected to have recorded at least 50 extra rings on the Solway if the influx of 5,000 birds in 1996-97 was due to immigration from the Greenland population. Geese from the Greenland breeding population were recorded at Caerlaverock in 1965 (M A Ogilvie) and 1976 (M Owen), as were 5 birds in 1991-92, one in 1994-95 and 2 birds in 1995-96. Some immigration therefore seems to occur, but not on a scale that explains the recent population increase. The goose identified in 1976 remained in the Svalbard population, but all the other birds were recorded back on Islay the following winter (S Percival pers comm). Thus, although some movement between populations has been recorded, the small number of birds identified on the Solway that had been ringed outside the range does not provide evidence for large scale immigration from the Baltic and Greenland populations. Difficulties in reading rings on the saltmarsh at Rockcliffe, however, means thatthe arrival of new birds to this part of the Solway would go undetected, particularly if it were for short periods late in the season. Some tentative support for the immigration hypothesis comes from counts of Barnacle Geese on Islay, which show that numbers there decline during the winter when the autumn population exceeds 20,000, and not when it falls below this level, but there were insufficient data for closer investigation of mortality or emigration with respect to initial population size (Choudhury & Owen 1993). Moreover, the Scottish Birds (1999) fact that few individuals in the Russian population carry rings means that we cannot readily dismiss the possibility of a large scale movement of these birds to the Solway Firth. The Russian population has been monitored closely in recent years so, if 5,000 birds were to move to Scoiland for a prolonged period, the annual census of the Russian population would have indicted a phase of reduction or stability in numbers, rather than the continual increase in recent years (B Gunter pers comm). There remains the possibility that geese may arrive on the Solway Firth direct from Russian breeding grounds, but this is impossible to prove in the absence of a more extensive ringing programme. Annual censuses It was thought initially that the increase in goose numbers recorded during the 1990s could be attributed to the different counting methods used (described above), and particularly to bird movements influencing the coordinated count totals. The sustained increase between 1990-91 and 1996-97, however, indicates genuine growth resulting in a near doubling of the population over this period (Table 1). The high figures returned from the coordinated census in autumn 1996 confirmed 2 high coordinated counts made the previous spring, when an average of 17,445 birds were recorded in April. These counts awaited confirmation because the vast majority were recorded on Rockcliffe Marsh, which is alarge expanse of saltmarsh and difficult to cover. The coordinated census in October 1995 had recorded only 12,700 birds, with coordinated counts at the time giving only slightly higher figures (averaging 13,115 with the census total; Table 1). Armstrong & McCaul (1995, 1996) had reported 2 high counts of the area in the Barnacle Geese on the Solway: 1990-1996 67 previous winter, averaging 17,880 in February 1995. Two opportunistic (partial) counts by WW also found an average of 15,100 in February 1995, which were similar to Armstrong & McCall’s figures, but were again higher than WWT’s coordinated census the previous autumn. The opportunistic and coordinated counts therefore indicated that, at least from 1994-95 onwards, good numbers of birds were arriving after the date of the coordinated census in autumn (Table 1). Lower mid winter and spring assessments before 1994-95 can be attributed to only parts of the Solway Firth being counted from mid winter onwards in earlier years. Table 1 summarises the autumn and mid winter/spring counts, and indicates the final population estimate for each season, up to and inciuding the 1996-97 winter. The coordinated census in autumn 1996 was 18,100. The coordinated counts obtained later in the year were averaged, giving a total of 23,000. This indicates that 4,900 birds arrived after the initial census. Certainly some 1,500 were recorded in other parts of Scotland in the third week of October, including: 800 at Loch of Strathbeg, 300 at Meikle Loch, 200 at Findhorn Bay, Moray, and 112 in the Montrose Basin. The proportion of juveniles also increased during the winter, from 12.1 % (n= 11,098 sampled in October) to 16.1 % (n=10,089 sampled in December and January), indicating the late arrival of some family parties. The proportion of juveniles recorded in 1994 and 1995 therefore was revised, adding the same proportional increase (of 33.06%) to the autumn values that had been observed between the autumn and mid winter estimates in the 1996-97 season. The final population estimate for the 1995-96 winter was also revised (Table 1), based on a population of 23,000 including 16.1% 68 JMBlack, D Patterson, P Shimmings & E C Rees SB 20(2) a Figure 1 Arrival dates for Barnacle Geese at Caerlaverock: a) first 1000 birds b) first 50% and c) 75% of final population value. Days since 20 September onSuUSRBRERSRBSAS Days since 20 September ounsansrnsnsasRsans Days since 20 September onGanSnsRSRESRBRS g 968 1972 41976 1980 1984 1988 1982 1996 Year —_ —_ 968 1972 1976 1980 1984 1988 1992 1996 Scottish Birds (1999) juveniles in 1996-97, and an estimated annual survival rate of 94% (Pettifor ef al 1998), since the spring 1995 counts were considered conservative and were made late in the season. Moreover, the difference in counts recorded in 1995-96 and 1996-97 had not been explained by mass immigration (see above), nor by observed breeding SUCCESS. Timing of arrival What further evidence is there to support the suggestion that Barnacle Geese have been arriving later on the wintering grounds in recent years? The timing of autumn migration was investigated by plotting the arrival dates for the first 1,000 birds, and for the first 50% and 75% of the final population, each year from 1970 to 1996 (Fig 1). Results indicated that over this period most of the population arrived earlier, rather than later, at Caerlaverock (r= -0.552, P<0.01, n=23; r=- 0.482, P<0.02, n=25 and r= -0.470, P<0.02, n=25 for the first 1,000 birds, first 50% and first 75% of the population respectively, Pearson correlations; Fig 1).Even when the exceptionally late 1970 season was excluded, the inverse correlation of arrival dates over time remained significant for the first 1,000 birds and for 75% of the population (r= -0.431, P=0.05, n=22; r= -0.368, P=0.08, n=24 and r= -0.400, P=0.05, n=24 in each case). An assessment of the Wetland Bird Survey counts from November to March (ie months beyond the normal migratory period) indicates that the number of flocks in north eastern Britain outside the Solway Firth region has increased during the study period, with a mean of 5.8 (SD +/-3.87) other sites being used each year from 1971-1979 inclusive, 12.1 (+/- 3.93) other sites from 1980-81 to 1989-90 and 15.4 (+/- 3.82) other Barnacle Geese on the Solway: 1990-1996 69 sites from 1990-91 to 1996-97. Moreover, the numbers recorded at these other sites has increased, particularly since 1984. At least 1,500 were still on migration in October 1995 and 1996, atthe time of the coordinated census, and a few had still not arrived by December (Fig 2). This suggests that, although most of the Svalbard population has arrived earlier on the Solway Firth in recent years, thereis an increasing tendency for some birds to lag behind and to trickle into the Solway during the mid winter months. Conclusion There have been exceptionally large counts and censuses of Barnacle Geese on the Solway Firth since 1994. Two coordinated counts returned an average figure of 23,000 birds in the 1996-97 winter. The low frequency of recording ringed birds from the other Barnacle Goose populations did not provide evidence for mass immigration, but it is possible that short term movements to Rockcliffe Marsh may have gone undetected. Counts of Barnacle Geese during mid winter months in northeast Scotland indicate that birds have stayed at autumn migratory sites for longer periods in more recent years. Unknown numbers of birds may also be lagging behind on uninhabited offshore islands on the west coast of Norway. Delayed arrival or immigration is not unprecedented for the Svalbard population; G Harrison, counting the geese on the Solway Firth in 1964, reported that Barnacle Goose numbers at Rockcliffe Marsh in spring can, indeed, be higher than counts undertaken in autumn at Caerlaverock (Harrison 1974; see also Roberts 1966). More frequent coordinated censuses will be undertaken in future, with a view to monitoring any build up in numbers during the winter, and recording any short term immigration from other populations. 70 JMBlack, D Patterson, P Shimmings & E C Rees SB 20(2) Figure 2 Number of geese recorded in northeast Scotland (excluding the Solway Firth) in October, November and December. 1600 lll - — Lie 1400 | 1200 1000 800 Number of birds 600 400 200 Ni N N N Sy N N BN a Ss . N N N N N N AN N N N N 8 1A (2913 74 75 1677 78 79580 181, 82 83484565 186 IC, B N N N N 7 88 89 90 91 92 93 94 95 96 Year Ra Oct a Nov ~~ Dec The continued growth in the Svalbard population has been attributed to an increase in newly established colonies (Prestrud et a/ 1989), and to additional nests at long established colonies (Drent et a/ 1998). This, in turn, is linked with an increase in the breeding ratio (the proportion of successful breeders to potential breeders), and the proportion of juveniles seen in winter flocks (Pettifor et a/ 1998). There is some evidence to suggest, however, that increasing density in the breeding range is having a deleterious effect at some of the older breeding colonies (Black 1998). Several demographic changes have been noted, including an older age of first breeding, an increase in the number of non breeders, and an increase in gosling and adult mortality during migration. There is also evidence that the growth rate and final body size of geese reared at some colonies has declined over the study period - probably through increased competition for limited food on the brood rearing areas (Black et al 1998, Loonen et a/ 1997). Acknowledgements We thank WWT staff and volunteers for making coordinated counts and censuses Scottish Birds (1999) possible, in particular Steven Cooper, Carl Mitchell and Peter Williams, who helped to ascertain recent population sizes. Richard Phillips kindly prepared the figures and, with David Cope and Myrfyn Owen, made helpful comments on a draft of the text. References Armstrong R & McCaul C 1995. Distribution of Barnacle Geese around the Solway Firth during the winter of 1994-5. Report to SNH, Dumfries. Armstrong R & McCaul C 1996. Distribution of Barnacle Geese around the Solway Firth during the winter of 1995-6. Report to SNH, Dumfries. Black J M 1998. Flyway conservation and management plan forthe Svalbard Barnacle Goose population. DN Report 22:1-100. Directorate for Nature Management, Trondheim. Black J M, Deerenberg C & Owen M 1991. Foraging behavior and site selection of Barnacle Geese ina traditional and newly colonized spring staging area. Ardea 79: 349-358. Black J M, Cooch E G, LoonenMJJE, Drent RH & Owen M 7998. Body size variation in Barnacle Goose colonies: evidence for local saturation of habitats. Norsk Polarinstitutt Skrifter200:129-140. Choudhury S & Owen M 1993. Migratory geese wintering on Islay: assessing the impact. Report to the Scottish Office. Wetlands Advisory Service Ltd., Slimbridge. Drent RH, Black J M, Loonen M J J E & Prop J 1998. Barnacle Geese Branta leucopsis on Nordenskidldkysten, western Spitsbergen - in thirty years from colonisation to saturation. Norsk Barnacle Geese on the Solway: 1990-1996 7A] Polarinstitutt Skrifter 200:105-114. Harrison J M 1974. Caerlaverock. Conservation and wildfowling in action. WAGBI Conservation Publication, WAGBI. Loonen M J JE, Oosterbeek K & Drent R H 1997. Variation in growth of young and adult size in Barnacle Geese Branta leucopsis: evidence for density dependence. Ardea 85:177-192. Myklebust M, Byrkjeland S, Gylseth P H & Storkensen OR 1994. Fugleri Norge 1994. Rapport fra Norsk faunakomite for fugl (NFKF). Var Fuglefauna 18:303-322. Owen M & Norderhaug M 1977. Population dynamics of Barnacle Geese Branta leucopsis breeding in Svalbard, 1948-1976. Ornis Scandinavica 8:161-174. Owen M & Shimmings P 1992. The occurrence and performance of leucistic Barnacle Geese Branta leucopsis. Ibis 134:22-26. Owen M, Black J M, Agger MC & Campbell CRG 1987. The use of the Solway Firth by an increasing population of Barnacle Geese in relation to changes in refuge management. Biological Conservation 39:63-81. Pettifor R A, Black J M, Owen M, Rowcliffe J M & Patterson D 1998. Growth of the Svalbard barnacle goose Branta leucopsis winter population 1958- 1996: an initial review of temporal demographic changes. Norsk Polarinstitutt Skrifter200:147-164. Prestrud P, Black J M & Owen M 1989. The relationship between an increasing population of Barnacle Geese and the number and size of their colonies in Svalbard. Wildfowl! 40:32-38. Roberts E L 1966. Movements and flock behaviour of Barnacle Geese on the Solway Firth. Wildfowl 17:36-45. 72 JM Black, D Patterson, P Shimmings & E C Rees SB 20(2) ’The Wildfowl & Wetlands Trust, Slimbridge, Gloucestershire, GL2 7BT 2 The Wildfowl & Wetlands Trust, Caerlaverock, Eastpark Farm, Dumfries, DG1 4RS 3 The Wildfowl & Wetlands Trust, Martin Mere, Burscough, Ormskirk, Lancashire, L40 OTA 4 Department of Wildlife, Humboldt State University, Arcata, California 95521-8299, USA °> Findon Cottage, Clashmore, Dornoch, Sutherland, 1V25 3RG ® Postboks 97, N-8860 Tjotta, Norway Revised manuscript accepted September 1999 Barnacle Geese Myrfyn Owen Scottish Birds (1999) 20: 73-80 73 Breeding population estimates for Lapwing, Oystercatcher, and Curlew in Scotland: results of the 1998 BTO Lapwing Survey A M WILSON & S J BROWNE Volunteer fieldworkers surveyed 142 tetrads in Scotland for breeding Lapwings during spring 1998, 118 of which were surveyed for other wader species also. Breeding populations of 69,800 pairs of Lapwings, 36,200 pairs of Oystercatchers and 32,000 pairs of Curlews were estimated in Scotland. The Lapwing population estimate is slightly higher than that for England and Wales in 1998, suggesting that the British population may now be less than 140,000 pairs. The estimates for Oystercatcher and Curlew are conservative. This suggests that previous estimates of the British population of Curlews may be too low. Introduction Lapwings Vanellus vanellushave been known to be declining in the British Isles for several decades, especially in England and Wales (Shrubb & Lack 1991, O’Brien & Smith 1992). This decline appears to have accelerated recently in England and Wales where there was a 49% drop in numbers between 1987 and 1998 (Wilson et alin press). There is little evidence for such large declines in Scotland where the RSPB and SOC surveys revealed anon significant 138% decrease in numbers in the Scottish lowlands between 1992 and 1997 (Tharme 1998). Since the Lapwing was once a common and widespread bird on lowland farmland, its decline is important since it may reflect wholesale changes in farming practice in both arable and pastoral farming areas. Against this background, the British Trust for Ornithology (BTO) organised a survey of breeding Lapwings in Scotland in 1998 in conjunction with the BTO/RSPB survey of Lapwings in England and Wales. The aim was to provide a baseline population estimate comparable with that made for England and Wales. There have been several estimates of breeding Lapwing numbers on Scottish lowland farmland in recent years (Galbraith et al 1984, O’Brien 1996) but there has not been an attempt to estimate the population of the country as a whole. During the survey, other wader species were also counted as some of these also have substantial breeding populations in Scotland. Methods Field methods followed those used in the Lapwing survey of England and Wales in 1987 and 1998 (Shrubb & Lack 1991, Wilson et al in press), where volunteers surveyed a single randomly selected tetrad (2x2km square) in each 10km square of the national grid. In addition to counting Lapwings, observers were given the opportunity to count all other wader species, although this was optional. As it was not feasible to achieve as high a level of coverage in Scotland, which is less densely populated with volunteer fieldworkers than England and Wales, it was decided to restrict the survey sample size. 74 AM Wilson & S J Browne SB 20(2) This was done by selecting a higher number of random tetrads in regions where occupancy by Lapwings was high during the 1988-91 Breeding Atlas (Gibbons et a/ 1993), than in areas where occupancy was low. The Stratification was done using the 29 BTO regions of Scotland, which are loosely based on the old counties of Scotland, or subdivisions thereof. A Regional Index of Lapwing Occupancy (Regindex) was calculated from the 1988-91 Breeding Atlas data as follows: Regindex = tetrads where Lapwings were seen / tetrads visited. It was thought that a survey sample size of 300 selected tetrads would be sufficient. The number of tetrads selected in each BTO region was calculated as follows: Number of tetrads selected in each BTO region = ( Regindex / Totindex) x 300. Totindex equals the sum of Regindex across 29 BIO regions. One tetrad was selected randomly from each 10km square and then the appropriate numbers of tetrads was randomly chosen from these for each BTO region. This resulted in a total of 281 tetrads being selected, some of the initial 300 were excluded as they were entirely inthe sea. The surveys were carried out on a single day during April, to coincide with the peak egg laying period of Lapwings. Apparently breeding pairs of waders were plotted on maps and the total number of pairs estimated for each of the 4 x 1km squares in the tetrad. Observers were asked to visit all suitable breeding wader habitat within the tetrad, most of which was covered by scanning with binoculars from roads and paths. As wader densities are more likely to be affected by landscape features such as land use, gradient and altitude than by geographical location, it was decided to analyse the data using Institute of Terrestrial Ecology (ITE) Landscape Types (Bunce et a/ 1996). This system classifies each 1km square into one of 4Landscape Types: Arable, Pastoral, Marginal Upland and Upland. As ITE Landscape Types are given by 1km square rather than by tetrad, it was decided to analyse the bird data on a 1km square level. Separate population estimates were made for each of 4 Landscape Types within each BTO region as follows: Population = mean count within 1km squares covered x total number of squares This resulted in population estimates for each BTO region and Landscape Type; 108 strata in allas some BTO regions do not contain all 4 Landscape Types. However, as coverage was sparse in some regions, populations were estimated for only 60 of the possible 108 strata. Population estimates for strata where no 1km squares were covered were made by calculating a mean density in each Landscape Type across all regions covered and multiplying this by the area of each Landscape Type in regions not covered. There was no systematic bias in those regions where coverage was low. The Pearson Correlation Coefficient between the number of tetrads selected and the number of tetrads covered was not significant (r=0.10392). Ninety five percent confidence limits were derived using the proportionately random bootstrap method to produce the upper and lower 2.5 percentiles of the frequency distribution estimates derived from 999 repeated simulations (Greenwood 1991). Scottish Birds (1999) Results A total of 142 tetrads was covered for the survey representing 513 x 1km squares (55 squares were entirely in the sea). The proportion of squares covered falling into the Pastoral and Marginal Upland squares closely reflected the proportion of those Landscape Types in Scotland, but Arable squares were over represented and Upland squares under represented. However, as estimates were calculated by Landscape Type, an uneven spread of squares would not bias the population estimates. While all observers counted Lapwings, 17% of observers made no attempt to count other wader species, resulting in a smaller sample size of 424 x 1km squares for other waders (Table 1). Oystercatcher Haematopus ostralegus and Curlew Numenius arquata were found in a high proportion of squares, although there was insufficient coverage of other wader species to produce population estimates. Oysiercatcher was found to be the most widely distributed wader species with breeding birds found in 21% of 1km squares, marginally more than Lapwing which was found in 20.1% of squares (Table 2). Breeding estimates for Lapwing, etc in Scotland 1998 75 Curlew was found in 19.3% of 1km squares but was the most widely distributed wader in Upland squares. Overall population estimates were 33,554 pairs of Oystercatcher, 64,408 pairs of Lapwing and 32,413 pairs of Curlew (Table 2). The Uists in the Western Isles are well known for their especially high concentrations of breeding waders (O’Brien 1996) but, unfortunately, no squares were surveyed there as part ofthe 1998 BTO Lapwing survey. Populations of Oystercatcher and Lapwing have been well studied on the Uists where the most recent estimates were of 2,726 pairs of Oystercatchers and 5,410 pairs of Lapwings in 1993 (O’Brien 1996). The populations of both species may have actually increased on the Uists since then (Fuller & Jackson in press). As the Uists form less than 4% of the land area of the Marginal Upland Landscape Type, it is reasonable to assume that excluding these islands from the analysis would have little effect on the population estimate for that Landscape Type. Adding the Uist populations to the totals in Table 2 would give Scottish population estimates of around 36,200 pairs of Oystercatcher and 69,800 pairs of Lapwing. Since Curlew have only recently colonised the Uists and remain Table 1 Numbers of 1km squares covered for the 1998 Lapwing Survey in each of the ITE landscape types in Scotland. Landscape type Number of 1km squares in which Lapwings were surveyed (% of total) Arable 123 (24) Pastoral 62 (12) Marginal upland 109 (21) Upland 219 (43) Number of 1km squares in which other wader species were surveyed (% of total) % of Scotland falling into each landscape type 119 (28) 16 61 (14) 13 84 (20) 23 160 (38) 48 SB 20(2) AM Wilson & S J Browne 76 Table 2 Population estimates of breeding Lapwing, Oystercatcher and Curlew in Scotland in 1998. Species & % squares Pairs found Density (pairs Population Lower 95% Upper 95% Landscape type occupied per 1km?) estimate (pairs) confidence limit confidence limit Oystercatcher Total 21.0 186 0.38 33,554 21,810 48,226 Arable Zoe 54 0.45 6,596 SOM UA 10,208 Pastoral ZnS 14 0.23 Zioem 998 4,168 Marginal Upland 23.8 55 0.65 11,660 5,124 18,380 Upland 16.3) 63 0.29 W277 1 9,623 16,071 Lapwing Total 20r 338 0.66 64,408 33,809 104,567 Arable 27.6 98 0.80 11,463 6,036 19,738 Pastoral 19.4 40 0.65 S702 1,026 8,178 Marginal Upland Zi\eal 103 0.94 21,740 7,665 40,726 Upland W258 9 0.44 27,474 9,542 48,136 Curlew Total 19:3 180 0.42 32,413 22,083 43,085 Arable 1 27 0:23 2,628 12055 4,248 Pastoral Zilles 22 0.36 3,046 WOU 5, 1300 Marginal Upland 23.8 59 0.70 12,265 6,025 19,281 Upland 19.4 U2 0:33 14,474 WOAOEKE 18,789 Scottish Birds (1999) scarce around 26 pairs were located in 1998 (Rabbitts 1999), no adjustment to the Curlew population estimate from Table 2 was necessary. Discussion Population estimates for Oystercatcher, Lapwing and Curlew in lowland Scotland in 1992-93 were 82,500 pairs, 92,000 pairs and 35,000-55,000 pairs respectively (O’Brien 1996). The Scottish population estimates published here are lower than those for the Scottish lowlands in 1992-93 for all 3 species, although confidence limits overlap for Lapwing and Curlew. Populations of these waders showed no significant change in the Scottish lowlands between 1992-93 and 1997 (Tharme 1998) but Lapwings decreased by a statistically significant 28% on Breeding Bird Survey squares in Scotland between 1994 and 1998 while Oystercatchers decreased by a non significant 18% and Curlews by a non significant 10% over the same period (Noble ef a/ 1999). The 1992-93 estimates were derived from an RSPB/SOC survey of waders in lowland Scotland. The definition of lowland Scotland used in the RSPB/SOC survey was that from the Macauley Land Capability for Agriculture classification system and is impossible to compare with the 4 Landscape Types used in this paper. The Upland Landscape Type for example, includes most ofthe Scottish islands while the Marginal Upland Landscape Type also includes a large areas of lowland (Figure 1). The estimates of lowland populations from the RSPB/SOC survey was based ona definition of lowland Scotland covering 46.7% of the total area. The Arable and Pastoral Landscapes together make up just 29% of the total area of Scotland, soclearly, lowlands as defined by the Macauley system include Breeding estimates for Lapwing, etc in Scotland 1998 DE considerable areas of Marginal Upland and Upland as defined by ITE Landscape Types. Since large parts of the Upland and to a lesser extent the Marginal Upland Landscape Types hold very low Lapwing densities, it could be argued that the RSPB/SOC survey did in fact cover a substantial proportion of the Scottish Lapwing population. We suggest that the final estimate of around 69,800 pairs of Lapwings in Scotland from the 1998 BTO survey is not significantly biased. This total is more than the estimate of 63,000 pairs in England and Wales in 1998 (Wilson et al in press). This suggests that the British population of Lapwings is now less than 140,000 pairs - considerably less than the 185,000 -238,000 pairs estimated for 1988-91 at the time of the New Breeding Atlas (Gibbons ef a/ 1993). The population estimates for Oystercatcher and Curlew should, however, be considered too low. The estimate of 36,300 pairs of Oystercatchers is particularly low compared with the estimated 82,500 pairs inthe Scottish lowlands in 1992/93 (O’Brien 1996). It should be noted thatthe RSPB/SOC surveys involved 3 visits to each 1km square and the estimates of Oystercatcher and Curlew numbers were taken as the maximum number on any one visit. The 1998 BTO survey involved just one visit, which would naturally lead to lower population estimates. Also, the BTO survey was designed specifically to locate breeding Lapwings and as such, underestimates of numbers of other species may have occurred for a number of reasons. Most importantly, the visits were timed to coincide with the peak period of nesting activity of Lapwings, which are earlier than that for Oystercatcher or Curlew (Cramp & Simmons 1983). It is also possible that observers put more effort into locating Lapwings than other waders and 78 A M Wilson & S J Browne SB 20(2) Figure 1 Distribution of ITE Landscape Types and tetrads covered in the 1998 Lapwing survey. Black = Landscape type. White squares = tetrads covered in each Landscape type. Arable Pastoral Upland Marginal Upland Scottish Birds (1999) they may have not surveyed some habitats that are unsuitable for Lapwings but are used by other species, notably shorelines and river channels which are favoured by Oystercatchers. It is also likely that some tetrads were not covered as they were deemed unsuitable for Lapwing, high moorland being one example. Scotland does not appear to have suffered from the widespread loss of breeding waders on lowland farms that has been encountered in England and Wales, although there can be little doubt that large decreases have occurred in the more intensively farmed areas (Stan da Prato pers comm) and there is now good evidence of a decline in Lapwing numbers in Scotland during the 1990s (Noble ef a/ 1999). However, due to the more rapid and long standing decline further south, Scotland holds an increasingly high proportion of the British population of several wader species, including over half the Lapwing population. Since the estimate of 32,000 pairs of Curlews in Scotland in 1998 is undoubtedly too low, the most recent British population estimate of 33,000-38,000 pairs (Reed 1985) may be an underestimate since other parts of Britain, especially northern England also support large populations of breeding Curlews (O’Brien & Murray 1998, O’Brien ef a/ 1998). Furthermore, Britain is estimated to support at least 35% of the European breeding population of this species (Piersma 1986), placing great importance on the Scottish population in particular. The maintenance of mixed farming systems and traditionally managed grasslands are essential if numbers of these and other waders species are to be maintained. The designation of Environmentally Sensitive Areas in Scotland has the potential to benefit breeding Breeding estimates for Lapwing, etc in Scotland 1998 79 waders (Picozzi eta/ 1996) butsuchschemes would have to become more widespread if the rapid decline in breeding Lapwing numbers evident in much of England is to be prevented from occurring throughout Scotland. Acknowledgements First and foremost we would like to thank the volunteer fieldworkers and regional organisers for completing the fieldwork. Many thanks to Simon Gillings, Su Gough, Dr lan Henderson, Mark O’Brien, Dr Stan da Prato, Nicki Reed and an anonymous referee for many useful comments on drafts of this paper and Dr Rob Fuller and Dr Juliet Vickery for steering the BTO’s Lapwing Survey from its inception. The 1998 Lapwing Survey in Scotland was funded by the Helen and Horace Gillman Trust and the BTO’s Save Our Skylarks Appeal for research into farmland bird populations. References Bunce R GH, Barr C J, Clarke R T, Howard D C & Lane A M J 1996. ITE Merlewood land classification of Great Britain. International Journal of Biogeographry 23:625-634. Cramp S & Simmons KEL 1983. The Birds of the Western Palearctic Volume 3. Oxford University Press. Oxford. Fuller R J & Jackson D /n press. Changes in populations of breeding waders on the Machair of North Uist. Scotland, 1983-1998. Wader Study Group Bulletin. Galbraith H, Furness R W & Fuller R J 1984. Habitats and distribution of waders on Scottish agricultural land. Scottish Birds 13:98-107. Gibbons D W, Reid JB & Chapman R A 1993. The New Atlas of Breeding Birds in Britain and Ireland: 1988-1991 T & AD Poyser, London. Greenwood J J D 1991. Estimating the total number and its confidence limits. Appendix in Shrubb & Lack 1991. (see below). 80 AM Wilson & S J Browne SB 20(2) Noble D G, Bashford RT, Marchant J H, Baillie S R & Gregory RD 1999. The Breeding Bird Survey 1998. Report number 4. British Trust for ornithology, Joint Nature Conservation Committee and Royal Society for the Protection of Birds. O’Brien M & Smith K W 1992. Changes in the status of waders breeding on wet lowland grasslands in England and Wales between 1982 and 1989. Bird Study 39:165-176. O’Brien M 1996. The numbers of breeding waders in lowland Scotland. Scottish Birds 18:231-241. O’Brien M, Green M, Harris A & Williams | 1998. The numbers of breeding waders in Wales in 1993. Welsh Birds 2:35-42. O’Brien M & Murray 5 1998. Estimating the breeding wader populations on farmland in northern England in 1993. Wader Study Group Bulletin 85:60-65. Picozzi N, Cait D C & Cummins R P 1996. Breeding waders in the Cairngorm Straths ESA in 1995. Scottish Birds 18:197-204. Piersma T 1986. Breeding waders in Europe: A review of population size estimates and a bibliography of information sources. Wader Study Group Bulletin 48, Supplement. Rabbitts B 1999 (editor). Outer Hebrides (Western Isles) Bird Report for 1998. Reed T 1985. Estimates of British breeding wader populations. Wader Study Group Bulletin 45:11- IZ. Shrubb M & Lack P C 1991. The numbers and distribution of Lapwings nesting in England and Wales in 1987. Bird Study 38:20-35. Tharme A 1998. RSPB SOC Lowland breeding wader survey of mainland Scotland 1997. Scottish Bird News 52:13. The Scottish Ornithologists’ Club. Wilson A, Vickery J & Browne S. The numbers and distribution of Lapwings nesting in England and Wales in 1998. In press. Bird Study. A M Wilson, British Trust for Ornithology, The Nunnery, Thetford, Norfolk, [P24 2PU. S J Browne, Game Conservancy Trust, Fordingbridge, Hampshire, SP6 IEF Revised manuscript accepted September 1999 Biss: Fal ae oe BE > aa Mijares ZS Cy . .< Oystercatcher David Mitchell Scottish Birds (1999) 20: 81-93 81 Bird numbers in an Aberdeenshire glen in March-June 1987-99 D JENKINS & A WATSON Singing birds were counted in March-June 1987-99 by mapping territory locations. Birchwood held more birds per hectare than Pine or scrub, and woodland beside short grassland held most. Big areas of woodland or scrub had more species, but fewer birds per hectare, than smaller areas. Permanent pasture held more Lapwings than Oystercatchers, and arable land similar numbers of both. All species fluctuated in number from year to year, but 3 showed long term declines and 10 long term increases. Decline and recent increase in 3 other species were associated with stream pollution from acidification and later improvement in stream quality. Year to year change was related negatively in Wrens to snowfall (ie fewer after much snowfall), in Robins and Stonechats to the number of mornings with snow lying, and in Mistle Thrushes and Oystercatchers to frost. Several species, including Dunnock and Song Thrush, reported as declining in the UK, did not do so on the study area. Both it and nearby wintering areas lacked the profound farm and forestry changes typical of most of the UK. Introduction For 13 years the same observer (DJ) mapped singing birds in a narrow glen not subject to intensive farm and forestry practices. The aims were to record bird species and numbers in different habitats and to detect possible trends in numbers. Little is known of these topics in Scottish glens, save for grouse and raptors. Because hard winters can reduce bird numbers (Greenwood & Baillie 1991) and the study area is snowier and frostier than most parts of the UK, we compared changes in numbers with local records of snow and frost. Other workers’ research on water acidity and invertebrates in our study area provided an opportunity to compare their data with stream bird numbers. As in most studies of bird numbers, our data are indices, not censuses. Study area This was 40 km west south west of Aberdeen in a narrow glen bottom of the Water of Feugh, a tributary of the River Dee. It lay over granite, the parent material for the glacial deposits and soils. Moorland above it was used for deer and grouse shooting, and some sheep in summer. It ran 9km along the Water of Feugh and Burn of Corn at 134-280m altitude. Five routes, each a day’s walk, totalled 12 km, divided into 15 sections of 500m length and 7 other areas, which alone or combined formed habitat areas A-V (Fig 1, Appendix). All Pine was Scots Pinus sylvestris, for brevity called Pine below, and all Birch Betula pendula. Many trees regenerated after 1945, as shown by maps, aerial photographs, and a survey in 1985 (Watson & Hinge 1989), but 82 D Jenkins & A Watson SB 20(2) Figure 1 Study area showing habitat areas A-C, D, E, F-J etc described in the Appendix. The study area comprised 15 x 500m sections A to O, and 7 other areas P and Q to V, which alone or combined formed habitat areas A-C, D etc. checks since 1985 showed no material change during the period of the present study. Most woodland and all scrub were natural, with more tree and scrub species, more varied age and understorey, and less bare ground than modern plantations. Understorey plants were mainly Gorse Ulex europaeus, Broom Cytisus scoparius, Bracken Pteridium aquilinum, Heather Calluna vulgaris, Blaeberry Vaccinium myrtillus and some Juniper Juniperus communis. Natural pinewood grew north of A-E, K and L, and south of A-D, and planted Pine and Larch Larix decidua north-east of A. Young trees grew on all areas except fields, but browsing and fire prevented regeneration locally. Rabbits Oryctolagus cuniculus, Roe Deer Capreolus capreolus and Red Deer Cervus elaphus grazed all areas, with many Rabbits in fields. Sheep and cattle grazed M- O’s fields all years, Q’s in Summer, and T’s irregularly. Fires burned 2-3 ha of heather on F-J in March 1991 and 1996, after which a Wheatear Oenanthe oenanthe sang in 1996 and probably nested in 1999. Some scrub was Cut on negligibly tiny parts of E and S. Methods Birds studied in all areas and years DJ considered the Common Birds Census (CBC, Marchant 1983) unsuitable for this study in a long narrow area running through many habitats. While walking slowly with many stops, he plotted on 1:10 000 maps the locations of all singing birds heard up to 150m_ on either side, on each of 5 walk routes. The 300m width broadly coincided with the glen bottom. Some farmland was wider, but open terrain and few birds made it feasible to cover all of it. He began each walk within an hour of dawn in March-April and at 0600 in May-July. Mostly he walked the road from Finzean Sawmill west to Ballochan, but covered A-C and M-O by circuits. He used a telescope to check walk counts on M-O’s fields and to count on Q’s fields from a vantage point. He walked each route on 4-5 days in 1987, 8- 12 in 1988, and c17 annually in 1989-99 by weekly counts in 1 March-30 June. Even the 4-5 greatly exceeded the 2 now recommended (BTO 1996) for the Breeding Bird Survey (BBS), and the usual 17 exceeded the 10 whichis the minimum of the CBC, so fieldwork was more time consuming than in the BTO national surveys. Also, the timing within an hour of dawn or at 0600 in summer was more stringent. We regarded a species as territory holding if song was heard in 3 or more weeks in the same 500m section of a walk route. In a Scottish Birds (1999) preliminary account ofthe early years, Jenkins (1995) based each species’ spring/summer total on the highest number repeated on at least 3 days on each route (eg 4, 6 and 5 would give 4), assuming that transients caused the peak of 6. This assumption could not be tested, and some routes covered several different habitats. Here, therefore, we use a different method: the highest spring/ summer count for each species in each of the 15 sections and 7 other areas. It gives some confidence that totals by both methods are closely related, eg in Wren Troglodytes troglodytes, Robin Erithacus rubecula, Willow Warbler Phylloscopus trochilus, Blue Tit Parus caerulea and Chaffinch Fringilla coelebs (n= 13 years, r,=0.72 - 0.88, P< 0.02 - < 0.001). Birds not studied in all areas and years Birds in an Aberdeenshire glen in March-June 1987-99 83 Notes below Table 1 show data on species not counted in all areas and years. No counts were made of: Siskins Cardueélis spinus, often seen over all habitats but rarely heard singing; Crossbills Loxia spp in years when all were in flocks in June; the mainly crepuscular or nocturnal Tawny Owl Strix aluco, Woodcock Scolopax rusticola and Snipe Gallinago gallinago; Pheasants Phasianus colchicus and Grey Partridges Perdix perdix because gamekeepers released them; and Carrion Crows Corvus corone and Magpies Pica pica because keepers killed them. Apart from Kestrel Falco tinnunculus, Buzzard Buteo buteo and perhaps Sparrowhawk Accipiter nisus, raptors nested outside the area, and were counted by other observers. From 1992 the landowner forbade entry to wood P, but from nearby fields the observer noted birds singing in P, except Goldcrests Regulus Table 1 Annual total number of territories, based on numbers of singing passerines, lekking Blackcock Tetrao tetrix, duck pairs, and nest sites in a gull colony to nearest 5. or 6e 69 90 91 Mallard 10 8 4 8 Goosander 5 7 8 5 9g Black Grouse” stewie. 19 26. 17 Oystercatcher Mpetidea 15 -14 13 Lapwing eames talS 155. 19 Curlew eG tf 919-43 Common Sandpiper 1 1 3 3 0 Common Gull Z0eeSO 50° 55° 25 Cuckoo 7 5 7 10 5 Tree Pipit 0 3 1 2 0 Meadow Pipit+ - - + 19 - Grey Wagtail 3 9 14. AS 7 Pied Wagtail 4 3 4 10 7 Dipper 5 6 8 6 9 Wren 25> AB 53 76 40 Dunnock Peet AO. AS AS Robin Phe At cof. .46 35 ee GS Ce CS be 7 Gi be 22S V6 ib. “AS 9 oil 18 Seu 1OF tS 12h oe 0 2 0 Or 2 1 OO iS 10 5 ae: O 0 O-.0 f ih 2 6 3 3 SMR - 2 2 1 0 0 owe + + 16 ee. 5220 925 7 7 3 Desi“ G14 ott 6 6 9 ie sali Pt ee, toc ahs + 6 - 2 eS. Ole 64). 59) 43. 51.43. 63.81 92 12, 1k. 23 16 6 9 23 24 43, 40 32 44 29. 55 ..76 7/6 84 DJenkins & A Watson SB 20(2) Redstart Zi 4 5 8 5 6 2 5 Sma") 7 STE Whinchat Lal IS ween ae B2s Salat 9 10: 9 10 °S ie Stonechat 0 1 3 7 + 4 5 6 oa Ti S785 Blackbird 3 4 5 6 6 2 2 3 4 32 5 4. tws8 Song Thrush 16 6 21 138° 238 18 12 12 “ISS OR oeeoe Mistle Thrush 10 hs} ors 2c ne 8 Ti vq 6 oilSraat2 Willow Warbler 56 56 53 638 #68 59 48 49 68) Gh) (6SiomaiGG Spotted Flycatcher 3 3 5) 8 4 67 a0 5.) 1S io 5S a iGeestis Long-tailed Tit 0 4 4 2 4 4 2 2 Orns 3 corre Blue Tit 3 5 Sryeisenda , tO Wile ai 15) 154 AO M2Zaaeatis Great Tit 4 5 S19 as 9 14 £416 12) SISmeaeiee2 Treecreeper 10 5 7 3 8 ala 10:°)12q0 Al 2s 9 14 14 Chaffinch 50060 708” 725e7S AOWGASIEAl 69° (69 -* «69 Raia Yellowhammer 8 4 6 Uh 8 3 5 3 SES 2 SeeS Goldcrest except P 0 0 0 S 0 1 5 2 4 1 12) Galea Coal Tit except A-C 5 5 if Gi a0 8 2 7 114 14 oe a lc8} Greenfinch except P 0 0 1 1 2 0 1 2 Py eg? 3 Ze * In the early 1990s a secondary lek became established nearby but outside the study area, and birds abandoned the study area lek in May 1999 after flooding. + Birds present but not counted. Others: Redshank Tringa totanus 7 in 1991 and 95; Short-eared Owl 1 in 89 and 92; Great Spotted Woodpecker Dendrocopus major 7 in 89, 93 and 95, 2 in 96 and 97, and 5 in 98; Skylark Alauda arvensis 7 in 90 and 95; Sand Martin Riparia riparia 3 in 95, 7 in 96; and 6 in 99; Wood Warbler Phylloscopus sibilatrix 2 in 92 and 96; Chiffchaff Phylloscopus collybita 1 in 99; House Sparrow Passer domesticus 17 in 97, and 2 in 98 and 99; Linnet Carduelis cannabina 2 in 90, 1 in 92 and 95, 4 in 98 and 10 in 99; Redpoll Carduelis flammea 2 in 90, 1 in 91 and 98; Crossbill 1 in 97 and 3 in 98; Bullfinch Pyrrhula pyrrhula 7 in 94 and 2 in 98. Others present each summer but counted only in a few years were Woodpigeon Columba palumbus 12 in 98 and 9 in 99; Swallow Hirundo rustica in buildings 3 in 88, 7 in 98, and 6 in 99; Jackdaw Corvus monedula 12 in 97 and 15 in 98; and Starling Sturnus vulgaris 3 in 90, 4 in 93 and 96, 6 in 97, 10 in 98 and 2 in 99. Former residents were Red Grouse Lagopus scoticus on F-J until the mid 1980s and Grey Partridges on fields. regulus and Greenfinches Carduelis chloris | hypoleuca and others sang in more than 3 which were not loud enough. Of 35 species recorded irregularly, we regard those heard singing in only 1-2 weeks as not territory holding. A few Green Woodpeckers Picus viridis, Pied Flycatchers Ficedula weeks, but are excluded below because no females were seen. A pair of Short-eared Owls Asio flammeus that nested nearby in 4 years often hunted the area, and Wheatears sang on the area at times though mostly nesting outside. Scottish Birds (1999) Weather Mr F Sheridan ran a station for the Meteorological Office at Bridge of Dye at 160m, where weather resembled that on our study area 7 kmto the west. From his readings at 0900 daily we obtained maximum and minimum air temperature, precipitation, snow depth, and snowfall since the previous morning. We calculated the longest run of days with snow lying all day, given that snow one morning must lie all day if snow is lying next morning in the absence of a fresh fall. Because this proviso did not always apply, the longest run is a minimal value. Birds in an Aberdeenshire glen in March-June 1987-99 85 The range of November-April values for cumulative snowfall was 14-110cm, mornings with snow 4-34, longest run 1-10, and cumulative frost 58-168 day °C below 0° (eg treating -2° on one morning as 2, and -1° next morning as 1, adding to 3 day °C below). Snowfall was related to the number of mornings with snow and /n longest run (r= 0.81 and 0.78, P< 0.002 and < 0.003), and the last 2 were related (r= 0.94, P< 0.001). Cumulative frost was poorly related to snowfall, mornings with snow, and /n longest run (r= 0.10, 0.21, and 0.14), as expected because frost often occurs without snow during winter anticyclones. Table 2 Mean annual densities (number of territories of each species per 100 ha) in 10 woodland or scrub areas A-C to V, mean total density of all species, and number of species, with densities rounded to nearest whole number except in cases <0.5. A-C D CE Area size in ha 30:,, 159, 15 Woodpigeon” 8 i iO Green Woodpecker 0 Ue aC) Great Spotted Woodpecker 2 Za 30) Tree Pipit 4 Zita Wren 52 “Psy Sis) Dunnock 2 2 8 Robin 52is 25). 30 Redstart 15 4 1 Whinchat 0 OZ Stonechat 0 0 1 Blackbird 10 1 0 Song Thrush 14 Y © Mistle Thrush ; 10 6 4 Wood Warbler 1 1 0) Willow Warbler 45. Soin 50 Goldcrest 0.3 ay Spotted Flycatcher 15 6 4 Long-tailed Tit 6 Die a2. Coal Tit Wie ts slit Blue Tit 29 ee sabe Great Tit 25 3 a F-J K L P Ss U V ice) —_—h a | —_— Ol — NO —k 0 0 2 1 0 0 1 ZS) I, S22 8 15) 1 5 Th 1 23 0 2 5 7 3 31 0 86 DJenkins & A Watson SB 20(2) Treecreeper 19 Soe 1 8 2 5 0 0) 8 Jackdaw+ 42 0 O 0 0 0 0 0 0) 0 Starling+ 11 a © 0 0 0 2 3 0 0 House Sparrow 1 OO 0 0 0 0 0 0 ) Chaffinch (5 ASSO 16 62, 55° 1105, 27aiiaaeenS Greenfinch 6 OO 0 0 Ono 0 0 0 Linnet 0 Ono 0.2 0 0 0 1 0 0 Redpoll 0 One® 0.1 0 1 0 1 8 0 Crossbill 0.3 1 0 0 0 0 0 0 0 4 Bullfinch 0.3 On” eel 0.1 0 0 0 0 0 0 Yellowhammer 1 oy 0) 0 0 2 33>) AA aero 0 Total density 462 198 222 108 306 306 426 146 792 400 No. of species Zi (28) 710) 22 19 19 19)... Ae _ lA "1998 and 99, * 1988-99 bar 1990, * 1988-93, and + 1998, 99 and some data in other 1990s years. The number of species was related to the area’s size in ha (both|n, n = 10, r= 0.64, P = 0.045). The number of species per ha varied 48-fold in different areas. It was related negatively to the area’s size (both \n, r = -0.99, P < 0.0001), i e more species per ha on small areas. Total density (the mean number of birds of all species per 100 ha) was related to area size and to number of species per ha (all|n, r = -0.80 and 0.85, P = 0.006 and 0.002). These relationships did not depend on the small 1ha area U and 2ha V, and still held when the analysis excluded both areas. Results Number and density of species summering in woodland or scrub, and size of habitat areas The number of species in woodland or scrub ranged from 12 on the small area V to 27 on the big A-C (Table 2). Big areas held more species but lower total densities (notes below Table 2). Passerine density in woodland and wader density on farmland Passerine density (the mean number of passerines per 100 ha) in Birch area A-C exceeded that in other woodland and scrub (Table 2). Area A-C had a higher density than D in all 13 years (Wilcoxon paired ranks T = 0, P < 0.003), and likewise A-C than K. Peak density was at U, a streamside hardwood belt with short grassland round it. Some species had high density in any woodland beside short grassland, eg Willow Warbler and Chaffinch at P, U and V. Greenfinches were at high density in Cypresses at P. Permanent pasture on Q held more Lapwings Vanellus vanellus than Oystercatchers Haematopus ostralegus, but arable land on M-O had similar numbers of both (Table 3). Arable land held more Oystercatchers per unit area than on Q’s pasture. Both species reared young in both habitats. Fluctuations in numbers All species fluctuated over the years (Table 1), Tree Pipits Anthus trivialis and Long- tailed Tits Aegithalos caudatus from 0 to 4, and Stonechats Saxicola torquata from 0 to Scottish Birds (1999) Birds in an Aberdeenshire glen in March-June 1987-99 87 Table 3 Annual number of occupied Oystercatcher and Lapwing sites per 100 ha (rounded to nearest whole number) on area Q’s permanent pasture and M-O’s arable land including leys. 87 88 89 90 9192 93 94 95 96 97 98 99 Median Pasture, 26ha Oystercatcher 8 8 8 4 8 4 8 78 4 4 4 8 8 8 Lapwing Zoom O23 oone27 “Sle Sil 27> 91S) SA IZ 2 Arable, 49ha Oysiercaichern 16°16 25 20 2016 20 16 18 14 8 8 1@ VS Lapwing IZ 1618 20 27 26 Wey iS aS OP Oe IZ 16... 16 Most occupied Lapwing sites held pairs, but Oystercatcher sites often held singletons due to temporary absence of a partner feeding on farmland down Feughside and in nearby glens outside the study area. On Q’s pasture, Lapwing density exceeded that of Oystercatchers (n = 13 years, Wilcoxon paired ranks test, P = 0.0002). Area M-O had similar density of both. Lapwing density on Q tended to exceed that on M-O, but not significantly so. Oystercatcher density on M-O exceeded that on Q (P = 0.0004). Area T’s 14 ha of pasture (in 2 years barley) held no breeding Oystercatchers in 1987, 1997 and 1999, 2 in 1993, 3 in 1990 and 1992, and one in other years, and no Lapwings. Pastures of 1 and 1.2 ha below K and L, both overlooked by trees, had no waders. 9. The amplitude (highest value divided by lowest) was 8 in Blue Tits, 4.7 in Treecreepers Certhia familiaris, 4.5 in Redstarts Phoenicurus phoenicurus, 3 in Blackbirds Turdus merula, 1.9 in Whinchats Saxicola rubetra and Chaffinches, and 1.8 in Meadow Pipits Anthus pratensis. We tested changes over the 13 year period by comparing the number singing with the year, treating 1987 as one, 1988 as 2 etc. They were related positively (increase) in Pied Wagtails Motacilla alba, Robins, Stonechats, Willow Warblers, Spotted Flycatchers Muscicapa Striata, Blue Tits, Great Tits Parus major and Treecreepers (e=O5omioOnO: Si, < 0:05 to < 0.001). Goldcrests and Coal Tits Parus ater also increased (r = 0.84 and 0.74, P < 0.001 and < 0.01, excltiding Goldcrests on P and Coal Tits on A-C because of no counts in some years). Wrens increased with year, though not significantly (r =0.52, P <0.1). Goosanders, Oystercatchers, Common Gulls Larus canus and Cuckoos Cuculus canorus declined with the year (r =-0.64 to -0.93, P< 0.05 to < 0.001), and °Curlews Numenius arquata and Yellowhammers Emberiza citrinella though not significantly (r = -0.54, P < 0.1). The gull colony increased to 55 nest sites in 1990 but declined to none in 1996-99. To find whether numbers changed in relation to winter weather, we used proportionate change in each species (one year’s number 88 D Jenkins & A Watson SB 20(2) divided by the previous year’s number). We compared this with the intervening winter’s cumulative snowfall, number of mornings with snow lying, longest run of Successive days with snow lying all day, and cumulative minimum air frost. Proportionate change in 5 species was strongly related to the intervening winter’s factors of snow or frost. Because the weather factors were inter related, we used multiple regression to determine for each species which weather factor accounted best for the variation in proportionate change. The longest run of days with snow lying all day accounted best for variation in Wrens, the number of mornings with snow lying in Robins and Stonechats, and cumulative frost in Oystercatchers and Mistle Thrushes Turdus viscivorus (Table 4, Fig 2). For other species, proportionate change showed only weak associations with the weather factors. Frost was negatively associated with proportionate change in Song Thrushes and Chaffinches (r= -0.43 and - 0.42, P>0.16), and the number of mornings of snow negatively with proportionate change in Great Tits (r = -0.44, P = 0.16). Most associations were negative, as expected from Greenwood & Baillie (1991), but some positive. All were far from significant. Even with species in Table 4, the variation accounted for by all the weather factors was only 27-48%, bar 84% in Oystercatcher. Jenkins (1995) saw fewer stream birds in the early 1990s than previously, and Grey Wagtails and Dippers bred poorly. After Table 4 Proportionate change in the number of singing birds on the study area from one summer to the next in relation to the intervening winter’s number of mornings with snow lying, longest run of days with snow lying all day, and frost (Cumulative number of day degrees with minimum air temperature below 0°C). Each value under a weather column shows the percentage of the variation in proportionate change that is accounted by that weather factor. Snow mornings Longest run Frost P Oystercatcher 84’ < 0.0001 Wren 21 0.049 Robin 41 0.015 Stonechat 48 0.011 Mistle Thrush 43 0.012 Data were transformed (In) for proportionate change except in Oystercatcher. From multiple regression, data for each species show the adjusted R? under whichever weather factor accounted best for the observed variation in proportionate change (i.e. the highest R?), and n = 12, bar 17 in Stonechat. Incorporating the other weather factors contributed O extra to R? in Oystercatchers, Robins and Mistle Thrushes. Incorporating the longest run contributed an extra 10% in Stonechats. A Scottish Oystercatchers winter south of Scotland (Goss-Custard 1996), but many inland breeders return in February to lower Deeside, frosty winters may coincide in Deeside and further south, and many died in lower Deeside during hard frost and snow in March 1980 (Watson 1980). The 4 passerines were not seen on the study area in mid December-mid February save in mild winters. Scottish Birds (1999) Birds in an Aberdeenshire glen in March-June 1987-99 89 Figure 2 Proportionate change in numbers from one spring/summer to the next, and weather factors in the intervening November to April inclusive, a) in Oystercatcher in relation to cumulative frost (total of day ° of minimum air temperature below 0°C), and b) in Robin in relation to number of mornings with snow lying. The fitted regression line is shown. Proportionate change in Oystercatcher (Seabee 2d ee oO Oo Oo / ER. | — “i a | 05 , af 50 90 430 170 Cumulative frost cite iP (b) | eo a 2 is! ow £ ny iz Cc 143] eed) & 144 = Sy 2 rs o ———— a @ = ie ss ~ cr on Pe = Se ro) —- Qa === o 10 rw oO 0 SS. SS 06+ = Ss 7 4 10 16 22 28 34 Number of momings with snow 90 D Jenkins & A Watson SB 20(2) Common Sandpipers Actitis hypoleucos increased from a pair in 1987 and 1988 to 3 in 1989 and 1990, he saw none for 2 years, but later 2 in 1993 and 1996, and one in 1997. Grey Wagtails increased from 3 singing in 1987 to 15 in 1990, declined to 3 in 1994, and increased to 14 in 1998 and 11 in 1999. Dipper numbers were high in 1989-91, low in 1992-95, and high again in 1998 and 1999. Proportionate change in Grey Wagtail and Dipper was unrelated to snow or frost, but stream acidity may explain their changes in numbers (see Discussion). Discussion Number of species breeding in woodland or scrub, their density, and habitat area The number of species was positively related to the size of the area of woodland or scrub, thus fitting the species/area relationship of more species on big areas (MacArthur & Wilson 1967). The total number of birds per 100 ha was negatively related to the size of woodland or scrub areas. This is a common result when comparing densities in different studies, possibly due to observers choosing small patches because these have high density, or to small patches involving big edge effects where birds use ground outside (Gaston et a/ 1999). The former does not apply in our study, but the latter is inevitable in any small area. Small woodland areas P, U, and V lay beside fields and short grassland where some birds foraged after flying from the woodland. All 3 were narrow, where edge effects are bound to exceed those in square or circular areas. This is not a criticism of narrow areas, for it is realistic to expect Square and circular areas to have higher local density at their edges than at their centres. In woods further up Deeside, densities of selected species varied greatly between areas and years (D Jenkins and J Conroy, species Tables in Buckland et a/ 1991). The data are not strictly comparable with ours because their years differed (1980-84) and their Glen Tanar Pine and Crathie Birch are ancient woods. Most densities of selected species in ancient woods exceeded those in our more recent woods, by up to 2-4 times for the more abundant species. Total densities in ancient birch exceeded those in ancient pine (French et al 1986), and in ancient woods exceeded those in our woods by 1.2-fold in Pine and 2.6-fold in predominant Birch. Proportionate change in numbers from one summer to the next Snow or frost accounted well for proportionate change in numbers in Wrens, Robins, Stonechats, Mistle Thrushes and Oystercatchers. In the BTO data for England during more than 2 decades up to 1987-88, proportionate change in Wrens and Robins, and in several other species not in Table 4, was related to the number of days with snow (Greenwood & Baillie 1991). These authors suggested that other weather factors including frost appeared less important than snow, but did not analyse bird numbers and frost, and their measure of frost was the number of days with a grass minimum temperature below 0°C (ie not frost severity). Like us, they found that proportionate change in some species showed insignificant associations with snow, including a few associations with a positive sign. Because hard winters in Deeside and England usually coincide, one might expect proportionate change in bird numbers in Deeside and England to be similar. The BTO sent us their CBC data for some species in Scottish Birds (1999) 1987-95 for Scotland and 1987-96 for England, thus covering most years of our study. Because our data are not exactly comparable with BTO data, for instance because some BIO observers and sites changed between years, our analyses are tentative. Correlations in Dunnock Prunella modularis and Robin, where the BTO gave us data as examples, were very weak (n= 9 years with BTO Scottish data, r = 0.18 and 0.04). In the case of migrant species which winter in Africa, Deeside and UK birds may share migration and wintering grounds, but correlations between our data and BTO data in Willow Warbler and Spotted Flycatcher were again very weak (r = 0.26 and -0.41 with BTO UK data, and - 0:02 and - 0.004 with BTO Scottish data, n= 10 years with UK data and 9 with Scottish). Presumably our area was atypical of the lowland habitats which predominate in CBC plots, and certainly its lack o7 intensive farm and forestry practices fits this (below). Decline and recovery of stream birds in relation to acidity Soulsby et a/ (1997) measured acidity and invertebrate abundance in the Water of Feugh at Ballochan in the study area. An initial rise in acidity followed by a fall may explain the decline of stream birds noted by Jenkins (1995), and their later partial or intermittent recovery. Acidification in the early and mid 1980s coincided with declines of aquatic invertebrates, but in 1983-94 the acidity decreased and Mayflies Ephemeroptera increased in the Feugh and other Deeside streams over granite. Birds in an Aberdeenshire glen in March-June 1987-99 91 Comparison with national longterm trends Blue Tits increased on the study area over the 13 years, as in the CBC in the UK over 25 years (Marchant et a/ 1999). However, Lapwing, Dunnock, Song Thrush Turdus philomelos, Willow Warbler, Spotted Flycatcher and Goldcrest showed no obvious decline as reported for the UK (Gibbons et a/ 1993, Marchant et a/ 1999) and the last 3 species increased. Counts in a wood near Edinburgh also show no Dunnock decline (Smith & Appleton 1998), and data from the BBS show no decline of Song Thrush and Willow Warbler in Scotland (Anonymous 1999). Our study area lacked the heavy use of insecticide and herbicide which typify the UK and which especially affected southern England where most CBC plots have been sited. Because many species not seen in winter on our study area are known to winter on farmland and woodland at lower altitudes in Feughside and the Dee valley (Jenkins, unpublished), it might be argued that they might be affected there by intensive agriculture. However, these lower areas lie on infertile soils of class 3 or poorer, where stock rearing dominates farming (Black 1985). Because of the lack of intensive arable cropping, there has been far less pesticide use and other intensification there than on fertile arable soils in east Scotland and England. The study area’s habitats remained broadly unchanged during the study, not Subject to intensive forestry practices, intensive farming, or transfer of moorland to woodland or farm grassland. Hence the study area’s birds form a useful ‘control’, compared with those on land more affected by the profound farm and forestry changes over much of the UK in recent decades. 92 D Jenkins & A Watson SB 20(2) Acknowledgements We thank E Cameron for help with maps, R J Fuller, | Newton and S R D da Prato for comments, F Sheridan for weather data, J Watson for Fig 1, A M Wilson for BTO data, and landowners, gamekeepers and farmers for cooperation. References Anonymous 1999. The BBS perspective on Scottish breeding birds 1994-1998. Scottish Bird News 55: 3-4. Black H L 1985. Agricultural land use beside the River Dee. The Biology and Management of the River Dee (ed D Jenkins) pp. 94-99. Institute of Terrestrial Ecology, Huntingdon. British Trust for Ornithology 1996. Breeding Bird Survey 1996 Instructions. Thetford, Norfolk. Buckland S T, Bell MV & Picozzi N 1991. The Birds of North-East Scotland. North-East Scotland Bird Club, Aberdeen. French D D, Jenkins D & Conroy J W H 1986. Guidelines for managing woods in Aberdeenshire for song birds. In Trees and Wildlife in the Scottish Uplands (ed D Jenkins) pp 129-143. Institute of Terrestrial Ecology, Huntingdon. Gaston K J, Blackburn T M & Gregory R D 1999. Does variation in census area confound density comparisons? Journal of Applied Ecology 36: 191- 204. Gibbons D W, Reid JB & Chapman RA 1993. The New Atlas of Breeding Birds in Britain and Ireland: 1988-1991. Poyser, London. Goss-Custard J D (ed) 1996. The Oystercatcher: from Individuals to Populations. Oxford University Press, Oxford. Greenwood J J D & Baillie S R 1991. Effects of density-dependence and weather on population changes of English passerines using a non- experimental paradigm. /bis 133, supplement 1: UZUENSS Jenkins D 1995. Changes in numbers of some common birds in the Forest of Birse, 1987-1994. North-east Scotland Bird Report 1994: 66-71. MacArthur R H & Wilson E O 1967. The Theory of Island Biogeography. Princeton University Press, Princeton. Marchant J H 1983. BTO Common Birds Census Instructions. British Trust for Ornithology, Tring. Marchant J, Sanderson F & Glue D 1999. Changes in breeding bird populations, 1997-98. BTO News 220: 10-14. Smith B & Appleton G 1998. Every woodland tells a story. Scottish Bird News 51: 5. Soulsby C, Turnbull D, Hirst D, Langan S J & Owen R 1997. Reversibility of stream acidification in the Cairngorm region of Scotland. Journal of Hydrology 195.29i- Sine Watson A 1980. Starving Oystercatchers in Deeside after severe snowstorm. Scottish Birds 11: 55-56. Watson A & Hinge M 1989. Natural Tree Regeneration on Open Upland in Deeside and Donside. Institute of Terrestrial Ecology, Banchory. Andrew Dowell Long-tailed Tit Scottish Birds (1999) Appendix. Description of areas A-C to V, with the number of tree species in parentheses. A-C Mostly mature Birch, understorey mainly Bracken (20) D __Birchand westwards mostly Pine, Gorse, Heather, Blaeberry and Bracken (7) E Mature Pine and westwards Heather, Blaeberry, Bracken, young Pine, Gorse, Broom (14) F-J Heather and Bracken with Gorse, Broom, Dog Rose Rosa canina and rushy flushes (8) K Mostly mature Pine with much Larch (3) L Mature Larch, understorey of Bracken and some Gorse (15) M-O Rotating cereal, leys/hay and Turnips, and some permanent pasture (0) P Planted mature Larch and Pine, and cypress Chamaecyperus (8) Birds in an Aberdeenshire glen in March-June 1987-99 93 Q_ Pasture since 1970 (0) S Mainly grazed Grass with Gorse, Broom and Bracken, and trees scattered or in groups (14) T Pasture (barley in 2 years) surrounded by A-C (0) U Streamside belt, mainly Birch, Alder and Willow, with Broom, Gorse and Dog Rose (10) V Mature planted Pine and Larch with some Broom and Gorse (9) Allareas bar P had streamside Alders, Willows and Birches, and all bar K, L and P had flushes with tall Rushes Juncus spp. All woodland had open glades, and A-C, K, L, P, S, U and V were near grass fields. The avenue of dense Cypresses in P was thinned after 1991. The linear R was the Water of Feugh and Burn of Corn. The study area included all the main tract of Birch in A-C and all the glen’s farmland. David Jenkins, c/o Department of Zoology, University of Aberdeen AB24 3TZ Adam Watson, c/o Institute of Terrestrial Ecology, Banchory AB31 4BY Revised manuscript accepted September 1999 Pied Wagtail David Mitchell 94 Scottish Birds (1999) 20: 94-97 SB 20(2) Breeding success of Red-throated Divers on Orkney Mainland, 1973-1998 C J BOOTH The breeding success of Red-throated Divers on 8 hill lochans in the West Mainland of Orkney was monitored over a period of 26 years (1973-1998). Of a total of 201 breeding attempts, 58% were successful with 0.73 young and 1.23 young being fledged per attempt and successful attempt respectively. Comparing data for the 2 periods 1973-1985 and 1986-1998, no significant difference in breeding success could be detected in terms of number of breeding attempts, proportion of successful breeding attempts or the numbers of young fledged. Introduction In Britain the Red-throated Diver Gavia stellata breeds only in Scotland and Northern Ireland but the numbers in the latter province are probably less than 10 pairs (Gibbons, Reid and Chapman 1993). In Scotland the majority of breeding pairs are found in the Northern and Western Isles with others in the north and west of the mainland (Gibbons et a/ 1997). The favoured nesting habitats are the banks of small moorland lochs but some pairs nest on the shores of larger lochs. The purpose of this paper is to document the breeding success of Red-throated Divers on 8 hill lochans in the West Mainland of Orkney during the 26 years between 1973 and 1998. Red-throated Diver nesting success in Britain is reported to have declined over the 15 year period 1980 to 1995 (Crick et al 1997) and in consequence the species is included on the ‘medium alert’ list prepared by the British Trust for Ornithology. Study area The lochans are situated in blanket bog on moorland in the West Mainland of Orkney, at heights of between 180 and200 metres above sea level. The furthest distance between 2 lochans is 7 km and the shortest 400 metres and they are sited from 3.5 to 5 km from the sea. The areas of the lochans range from 0.011ha to 0.042ha and the depth varies between 0.5 and 1.5 metres. There is very little plant growth. Methods Human disturbance at nesting lochs of Red- throated Divers can affect breeding success (Booth 1982). In order to minimise disturbance from monitoring, no attempts were made to find nests with eggs. In May, each lochan was observed from a distance to establish if pairs were present. Further visits were undertaken in early July and then in early August to assess, again from a distance, the presence Scottish Birds (1999) Breeding of Red-throated Divers ,Orkney Mainland, 1973-1998 95 Table 1 Annual breeding success 1973-1998. Year Noof pairs No of successful No of young No of young No of young attempting attempts fledged fledged per fledged per to breed attempt successful attempt 1973 8 7 Z 0.87 1 1974 8 6 7 0.87 1216 1975 8 6 g et2 15 1976 8 5 6 0.75 ers 1977 8 2 2 0.25 1 1978 8 - 5 0.63 e235 1979 8 6 6 Q:75 1 1980 8 6 9 Lal2 5 1981 8 5 6 0:75 1.2 1982 8 3 5 0.63 1.66 1983 8 ~ ~ 0.5 1 1984 8 5 iL 0.87 1.4 1985 8 5 A 0.87 1.4 1986 8 5 5 0.63 1 1987 8 - 7 0.87 175 1988 8 5 f 0.87 1.4 1989 7 - 5 0.71 1:25 1990 7 S. 7 1 1.4 1991 ih - - 0.57 1 1992 7 6 ih 1 1.16 1993 re 3 - 0:57 1.33 1994 8 2 3 0.37 Las 1995 8 2 4 0:5 2 1996 if 5 5 0.71 1 1997 Ts 3 3 0.42 1 1998 8 5 r 0.87 1.4 1973-98 201 PZ 148 0:73 i223 of young and adult birds. If the lochan appeared deserted a search of the bank was carried out for signs of a nesting scrape. When young were present, their age was estimated and another visit made a few days before it was thought that they would fledge. A full grown young bird, showing no obvious sign of down, was recorded as fledged and the pair as being successful. The presence of a definite nesting scrape was considered to be a nesting attempt, although it was just possible that no eggs were ever laid. To look for a possible decline in breeding success, data were grouped into the 2 time periods, 1973-1985 and 1986-1998. Tests for significant differences were then undertaken 96 C J Booth SB 20(2) using chi-square analysis and, if p > 0.5, differences were taken to be not significant (Fowler and Cohen, BTO Guide No 22). Results Only one pair of divers was found nesting on a lochan in any one year. At 6 of the lochans, nesting was attempted in each year of the study. Of the other 2 lochans, one was abandoned from 1989-93 and the other 1996- 97(Table 1). The clutch size of Red-throated Divers ranges from one to 3 eggs (Cramp 1977) but fledged broods of only one and 2 young were recorded on the monitored lochans. The total number of young reared to fledging annually varied from 2 to 9, whilst the number of young reared per breeding attempt ranged from 0.25 to 1.12 and for successful attempts from 1.0 to 2.0 (Table 1). Overall there were 31(26%) broods of 2 young and 86(74%) of 1 young. A comparison of the breeding success in the 2 periods of the study showed no significant differences (Table 2). Discussion It can be expected that there will be some annual variation in breeding success but Table 1 shows that there are large variations from year to year in chick output from the monitored lochans (with peak productivity x 4.5 lowest). Only 2 young were reared in 1977, a time when egg collectors were particularly active in Orkney. Several clutches were thought to have been taken from the study lochans and this probably accounted for the very poor breeding success in that year. Ina10 year study in Shetland, avariation in the annual breeding success of Red- throated Divers was also found (Okill and Wanless 1990). The results from the Shetland study suggested less variability from year to year but this could be partly explained by different monitoring methods and because some failures may have been missed. Overall there were fewer breeding attempts (97 compared to 104) during the second period of the present study and a lower proportion were successful (54.6% compared Table 2 Breeding success in the periods 1973-1985 and 1986-1998. Parameter No of breeding attempts No (%) of successful breeding attempts No of young fledged 80 No of young fledged per breeding attempt No of young fledged per successful breeding attempt Period 1973-1985 Period 1986-1998 104 64 (61.5%) 0.77 125 Significance of difference 97 ns 53 (54.6%) ns 68 ns 0.7 ns eZee ns Scottish Birds (1999) Breeding of Red-throated Divers , Orkney Mainland, 1973-1998 97 to 61.5%. See Table 2). However, the observed differences between the periods are not statistically significant and are within the range attributed to chance variation. Therefore, there is no evidence of a decline in the breeding success of this monitored population in the West Mainland of Orkney over the period 1973 - 1998. This isin contrast to the trend noted for Red-throated Divers in Britain (Crick ef a/ 1997). Acknowledgements lam very grateful to Dr Peter Reynolds for his help with the statistics and his comments on an earlier draft of this paper and to Dr Paul Heppleston for additional comments and advice. Dr S da Prato and an anonymous referee made further improvemenis. References Booth C J 1982 Fledging success of some Red- throated Divers in Orkney. Scottish Birds 12:33- 38. Cramp S (ed) 1977 The Birds of the Western Palearctic Vol 1. Oxford University Press, Oxford. Crick H Q P, Baillie S R, Balmer D E, Bashford R |, Beaven L P, Dudley C, Glue D E, Gregory R D. Marchant C H, Peach W J and Wilson A M 1997. Breeding Birds in the Wider Countryside: their conservation status (1972 - 1996) BTO Research Report No 198. BTO, Thetford. Fowler J & Cohen L Siatistics for Ornithologists. BTO Guide No. 22. Gibbons D W, Bainbridge | P, Thorne AP and Ellis P M 1997. The status and distribution of the Red- throated Diver Gavia stellata in Britain in 1994. Bird Study 44:194—203. Gibbons D W, Reid J B and Chapman R A 1993. The New Atlas of Breeding Birds in Britain and Ireland, 1988-1991. T & AD Poyser. Okill J D and Wanless S 1990. Breeding success and chick growth of Red - throated Divers Gavia stellata in Shetland 1979 - 1988. Ringing & Migration 11:65-—72: C J Booth, 34 High Street, Kirkwall, Orkney, KW15 1AZ. Revised manuscript accepted September, 1999 Andrew Dowell 98 Scottish Birds (1999) 20:98-110 SB 20(2) Persecution of birds of prey in north Scotland 1912-69 as evidenced by taxidermists’ stuffing books HENRY McGHIE Persecution of raptors and owls was studied using the records of an Inverness based firm of taxidermists covering the period 1912-1969. A total of 767 birds of 14 species were submitted from north Scotland; many constitute hitherto unknown records of rare species. The numbers of birds submitted increased until 1930 and declined thereafter. The numbers of birds submitted were analysed in relation to vice county, land use and season. Persecution was most intense in East Ness but was also high in East Ross and East Sutherland. Records of Hen Harrier were analysed to chart the recolonisation of mainland Scotland by this species in the 1940s. Introduction The sporting estate developed as a major land use in Scotland from the beginning of the 19th century, when land vacated by the Highland Clearances was developed with wealth generated by the Industrial Revolution and the Agricultural Improvements (Prebble 1974, Temperley 1951, Brown 1976). The persecution of birds of prey was a routine part of land management and the populations of all birds of prey were reduced to varying degrees, with the possible exception of Kestrel Falco tinnunculus (Newton 1972). This persecution contributed towards some of the most drastic changes in the Scottish avifauna, and culminated in the extinction of 5 species of raptor as breeding birds in Scotland; bird of prey populations reached their nadir around the turn of the present century (Ritchie 1920, Baxter & Rintoul 1953). In spite of the importance of these changes to the ecology of the Highlands there is little quantitative information § available. Information for the 19ths century is largely based on estate vermin lists, which some workers have considered to be subject to exaggeration (Pearsall 1952, Brown 1976, Nethersole- Thompson in Ratcliffe 1980) and other sources for this period relied on patchy or speculative information (eg Harvie-Brown and Buckley 1895). Declines in persecution in the present century are inferred from declines in gamekeeping which occurred after each of the World Wars, and the spread of many birds of prey into areas from which they were formerly extinguished (Watson 1977, Ratcliffe 1980, Tapper 1992), but few actual numbers of birds killed or rates of population increase are published (Newton 1972, 1979). There were 2 large firms which offered a taxidermy service in the Highlands in the first part of the 20th century: John Macpherson & Sons (1887-1976) and Macleay’s (c1850-1960s). Both of these firms were based in Inverness although a number of small firms operated for short periods both in Inverness and Dingwall. Macpherson’s, Scottish Birds (1999) Evidence for persecution from taxidermists’ stuffing books 99 the larger of the 2 main firms, employed a resident taxidermist from 1912-67 and a visiting taxidermist from 1967-70; the firm remained open for business during both World Wars and was the only firm in operation after 1954. The taxidermists’ records from Macpherson’s were donated to Inverness Museum and Art Gallery in 1976; these consist of large ledgers or stuffing books which contain the details of every bird and mammal submitted for work and cover the period 1912-69. Each record consisted of the species, the date of submission and the name and posial address of the submitter; some eniries were annotated as to the condition of the skin with regard to damage. Methods Records of raptors and owls (Falconiformes and Strigiformes) were extracted from the stuffing books for Caithness, Sutherland, Ross-shire, Inverness-shire, the Outer Hebrides and Skye on a Watsonian vice county basis. These were the counties which were naturally served by Inverness, and as very little taxidermy work was undertaken away from the Highland capital the bulk of work would have been likely to make its way there; Orkney and Shetland were excluded as there were firms providing a taxidermy service there in the earlier part of the 20th century. Whilst it was not implicit from the stuffing books that the birds submitted died as a result of persecution several lines of evidence indicated that this was indeed the case: the monthly pattern of submission of birds differed from the natural seasonal pattern of mortality (Newton 1979); the majority of birds were submitted by sporting estates or estate workers, and comments relating to damage generally referred to broken legs which would be consistent with birds having been trapped. Taxidermists’ identifications were taken at face value. lt can be demonstrated for carnivores (McGhie and Moran in prep) that the majority of specimens were procured from the same area (Same or adjacent 10km square of the National Grid) as that from which they were submitted and within one calendar month of the date of submission, with decomposition being an obvious consideration; the same Situation was assumed to apply to birds of prey. One bird, a Red Kite Milvus milvus, was submitted from the same named locality, outside the area under study, as that from which it had been procured and 4 days later (Scottish Naturalist 1929). Stuffed birds, particularly birds of prey, were formerly much more popular than they are at present and Macpherson’s often purchased birds from submitters to be sold as stock; it is likely that no ceiling was set by the firm which would limit the numbers of birds submitted and the stuffing books record all items which were submitted whether work was carried out or not. The localities from which birds were submitted were traced using standard atlases (listed in references). Each locality was classified as being typified by upland, lowland or mixed land management practices in the early part of this century. The ‘upland land use’ class was defined as land managed primarily as grouse moor and/or deer forest in alocality (‘locality’ defined as circle of 3km radius centred on the place from which birds were submitted) with negligible or no agricultural land other than sheepwalk; the ‘lowland land use’ class was defined as land used solely for agriculture; localities which contained a combination of the aforementioned 2 classes were Classified as being of ‘mixed land use’ class. Upland, mixed and lowland land use occupied approximately 60%, 30% and 10% respectively of the area under study. Areas 100 Henry McGhie SB 20(2) of lowland land use were restricted to the Firthlands; areas of mixed land use extended up straths and glens and around the coast, whilst the remainder consisted of land under predominantly upland land use. Estate forest plantations (State forests were largely developed after the Second World War) were most widespread in the ‘mixed land use’ class, but still common in the ‘lowland land use’ class. Birds submitted on the first of a calendar month were allocated to the previous month for analysis, as they would have been unlikely to be submitted for preservation on the same day as that on which they were procured, given the remoteness of the majority of localities from Inverness. Results Records relating to a total of 767 individual birds of prey of 14 species of diurnal raptors and 5 species of owl from the area under study were detailed within the stuffing books. Of these, 766 could be related to the locality of origin, 766 to the year and 760 the month of submission; records were well distributed throughout the area. Of these, 6 species of diurnal raptors were regular breeders within the area (Golden Eagle Aquila chrysaetos, Buzzard Buteo buteo, Sparrowhawk Accipiter nisus, Peregrine Falco peregrinus, Kestrel, Merlin Falco columbarius) as were 4 species of owl (Barn Owl Tyto alba, Tawny Owl Strix aluco, Long-eared Owl Asio otus, Short- eared Owl Asio flammeus). Three species were irruptive winter visitors (Rough-legged Buzzard Buteo lagopus, Gyrfalcon Falco rusticolus, Snowy Owl Nyctea scandiaca), 4 were rare migrants (Marsh Harrier Circus aeruginosus, Montagu’s Harrier Circus pygargus, Honey Buzzard Pernis apivorus, Hobby Falco subbuteo) and 2 were formerly widespread breeders which had been confined to small areas (Hen Harrier Circus cyaneus) or wholly extinguished (Osprey Pandion haliaetus) as a result of earlier persecution but again established themselves during the period under study (status from Baxter and Rintoul 1953). The Scottish Naturalist was checked for details of each of the rarer species which were found in the stuffing books, but the only bird which was detailed was the Red Kite mentioned above; the birds mentioned within the stuffing books therefore constitute new records. The numbers submitted of each of the regularly breeding species, with the exception of Barn Owl, increased between 1912-29 and declined thereafter (Table 1). More than 40% of the total numbers of each of these species were submitted between 1920-29; high numbers (>20% of totals) of six of the regularly breeding species continued to be submitted in the period 1930- 39, during which period the greatest numbers of Barn Owl were submitted (31% of total). The numbers of birds submitted dropped dramatically after 1930-39 and dwindled thereafter until the period covered by the Stuffing books drew to a close: less than 10% of the total numbers of each of the regularly breeding species were submitted between 1950-69. Only Montagu’s Harrier, Hen Harrier, Osprey and Gyrfalcon had significant numbers submitted after 1930- 39: The greatest numbers of 9 of the 10 regularly breeding species (all except Short-eared Owl) were submitted from East Ness (vice county 96). The proportion of birds being submitted from East Ness declined steadily however: 47% of birds were submitted from the vice county in 1913-19, falling to 36%, 28%, 23% and 23% in the ’20s, 30s, *40s Scottish Birds (1999) Evidence for persecution from taxidermists’ stuffing books 107 Table 1 Numbers of birds of prey submitted to Macpherson’s 1912-69. Species 1912-19 1920-29 1930-39 1940-49 1950-59 1960-69 Total Golden Eagle 45 73 25 6 9 3 161 Peregrine 31 63 26 9 2 ike Buzzard 18 51 23 11 4 1 108 Sparrowhawk 15 29 13 2 3 62 Kestrel 13 17 7 2 4 43 Merlin 7é 14 6 2 1 30 Hen Harrier 1 3 2 3 2 Ll Gyrfalcon 3 3 1 f Rough-legged Buzzard 2 6 8 Osprey 2 2 Honey Buzzard 1 1 Marsh Harrier 1 1 Montagu’s Harrier 1 1 Hobby 1 1 Tawny Owl 25 38 9 7 1 1 81 Barn Owl 15 14 16 4 1 2 52 Long-eared Owl LIZ: 23 9 49 Short-eared Owl 4 6 4 1 15 Snowy Owl 3 3 Totals 193 342 144 52 24 12 767 and ‘50s respectively. Comparatively small numbers of all species, with the exception of Barn Owl, were submitted from Skye and only Gyrfalcon was submitted in a high proportion from the Outer Hebrides (Table 2). The commoner farmland species (Buzzard, Sparrowhawk, Kestrel, Barn Owl, Tawny Owl and Long-eared Owl) were mainly submitted from East Ross and East Ness although significant numbers were also submitted from East Sutherland. Large numbers of Barn Owl were also submitted from Skye. Three species (Hen Harrier, Short-eared Owl and Snowy Owl) were submitted in the greatest numbers from Caithness. Three of the 4 rare migrant species were only submitted from East Ness (Honey Buzzard, Marsh Harrier, Hobby) with the other being submitted from East Ross (Montagu’s Harrier), whilst Rough-legged Buzzard was mainly submitted from East Ness and East Sutherland. Golden Eagle and Peregrine were submitted in the greatest numbers from East Ness (Table 2) and in greater numbers from each named locality within this vice county (Table 3), although large numbers of Golden Eagle were also submitted from West Sutherland and West Ross localities. A total of 12 peregrines were submitted from one named East Ness locality between April - June in 7 years between 1915-38, with 2 birds being submitted to Macpherson’s together in April 1924, 1928 and 1938; low numbers of Peregrine were 102. Henry McGhie SB 20(2) Table 2 Numbers of birds of prey submitted to Macpherson’s from each vice county. Species Caiths W Suth W Ross W Ness Skye OHeb E Suth E Ross E Ness Total G Eagle 6 23 18 6 3 1 21 23 60 161 Peregrine 19 3 15 9 5 3 11 20 46 eA Buzzard 1 4 9 4 7 3 U7 29 34 108 Sparrowhawk 3 9g 2 6 11 31 62 Kestrel 4 2 3 2 3 1 1 11 16 43 Merlin 6 1 4 3 3 3 3 7 30 Hen Harrier 6 1 1 3 11 Gyrfalcon 1 2 3 6 Rough-legged Buzzard 1 3 1 3 8 Osprey 1 1 2 Honey Buzzard 1 1 Marsh Harrier 1 1 Montagu’s Harrier 1 1 Hobby 1 1 Tawny Owl 3 2 U 5 8 5 16 35 81 Barn Owl 1 1 8 ii 11 1 7 16 52 Long-eared Owl 4 1 2 2 2 4 2 22 49 Short-eared Owl 7 1 1 3 2 1 15 Snowy Owl 3 3 Total 64 40 75 35 45 16 75 135 281 766 submitted from West Sutherland (Tables 2 and 3). Of 19 instances where 2 birds were submitted together between March - August, 5 instances were of Golden Eagle (one from East Ross and 4 from East Ness), 8 instances were of Peregrine (one each from West and East Ross and6 from East Ness), 3 instances were of Sparrowhawk (2 from East Ross and one from East Ness), one instance was of Kestrel (from East Ross) and 2 instances were of Merlin (from East Sutherland and East Ness). The greatest numbers of birds were submitted from localities typified by mixed and upland land use (Table 4 and 5). Very few birds were submitted from localities which did not agree with their published habitat requirements; for example, only 2% of Golden Eagle were submitted from areas of lowland land use, with which the species is not associated (Table 4). This further indicated that the majority of birds were submitted from the areas in which they had been procured. Two species, Golden Eagle and Peregrine, were mainly submitted from areas of upland and mixed land use (96% and 92% respectively); the majority (>50%) of Hen Harrier, Merlin, Kestrel, Barn Owl, Tawny Owl and Short-eared Owl were submitted from areas with mixed land use, as were high numbers (>40%) of Buzzard, Sparrowhawk and Long-eared Owl. Significant numbers of Sparrowhawk (37%), Scottish Birds (1999) Evidence for persecution from taxidermists’ stuffing books 103 Table 3 Number of years in which Golden Eagle and Peregrine were submitted to Macpherson’s from each named locality. Species and Number of Number of years in which birds were submitted from an vice county Localities individual locality 1 2 3 4 Te Golden Eagle Caithness 6 6 W Suth 16 11 5 W Ross 14 11 2 1 W Ness 6 6 Skye 2 2 O Hebrides 1 1 E Suth 15 12 2 1 E Ross 13 8 2 2 1 E Ness 42 30 10 Zs Total 115 87 21 6 1 Peregrine’ Caithness 10 if 1 2 W Suth 3 3 W Ross 11 9 2 W Ness 8 8 Skye 4 4 O Hebrides 3 3 E Suth 9 8 1 E Ross 10 9 1 E Ness 26 20 5 1 Total 84 71 10 2 1 Notes: 1 Records for areas typified by upland or mixed land use from Feb-Sept only to exclude non breeding birds. Kestrel (35%), Long-eared Owl (43%) and Tawny Owl (36%) were submitted from localities typified by lowland use given the restricted area of this land use type. The seasonal pattern of submission varied between species and in relation to land use (Tables 4 and 5). In areas dominated by grouse moor and deer forest, the majority of Golden Eagle (53%), Peregrine (87%), Kestrel (100%) and Merlin (76%) were submitted between March-August. The majority of Sparrowhawk (60%) were submitted between September - February, whilst Buzzard was submitted in roughly equal numbers in spring/summer (40%) and autumn/winter (60%), with a slight peak in March-April and a slight low in July-August. In areas of lowland land use the majority of Peregrine were submitted between July- 104 Henry McGhie August (60%) and January-February (20%) and Merlin were submitted between January- April (75%) and September-October (25%). Buzzard, Sparrowhawk and Kestrel were submitted more evenly throughout the year from agricultural land than in areas of grouse SB 20(2) moor/deer forest. Areas of mixed land use generally showed patterns of submission intermediate between those from land managed as grouse moor/deer forest and agricultural land. Allowls, except Short-eared Owl, were submitted in greater numbers Table 4 Numbers of birds of prey submitted to Macpherson’s in relation to land Mar-Aug Sept-Feb Mar-Aug Sept-Feb Mixed (30%) Upland (60%) Mar-Aug Sept-Feb Percentage of each species use. Species Total Lowland (10%)' Golden Eagle 161 2 2 Peregrine ent 4 4 Buzzard 108 9 18 Sparrowhawk 62°23 1S Kestrel 43 23 12 Merlin 30 ii Vi Tawny Owl St 10 26 Barn Owl 52 8 8 Long-eared Owl 49 16 2h Short-eared Owl 16) ik Hen Harrier 1913-397 8 1 Hen Harrier 1940-69° 8 Gyrfalcon 5 1 Rough-legged Buzzard 8 Osprey 2 Honey Buzzard 1 Marsh Harrier 1 Snowy Owl S 24 21 32 ie 29 18 39 6 14 30 iil 18 18 29 6 9 23 28 14 37 23 20 6 15 41 + 18 46 10 10 16 27 + 10 40 46 Number of each species A 1 2 4 2 2 2 2 oe 1 1 4 1 1 1 1 3 Notes: 1 Figures in brackets are approximate percentage of the total land area which each land use type occupied 2 Includes 2 birds from Morayshire (see Table 6) 3 Includes 2 birds from Morayshire and one from Aberdeenshire (see Table 6) Scottish Birds (1999) Evidence for persecution from taxidermists’ stuffing books 105 Table 5 Monthly pattern of submission of birds of prey to Macpherson’s in relation to land use. Species Land use Total Jan-Feb Mar-Apr May-JunJul-Aug Sept-Oct Nov-Dec type Percentage of total submitted Golden Eagle 160 is) 30 28) 6 12 17 Peregrine Lowland 10 20 10 10 30 30 Mixed 62 13 24 24 18 8 18 Upland 59 Si// 50 8 5 Buzzard Lowland 29 28 10 10 14 21 al Mixed 47 19 15 9 9 19 30 Upland 30 7 20 17 3 26 UZ SparrowhawkLowland 23 4 iS 22 26 is 22 Mixed 29 4 10 21 Z 41 17 Upland 10 10 30 30 30 Kestrel Lowland 1S i Pil 27 12 27 Mixed 22 9 9 18 18 14 32 Upland 6 34 16 50 Merlin Lowland 4 Zo 50 7A) Mixed 18 28 ile 22 28 I Upland 8 12 25 51 JZ Tawny Owl 81 24 10 6 V1 21 28 Barn Owl 52 35) 21 8 6 il 18 Long-eared Owl 49 22 12 12 12 22 18 Short-eared Owl 15 13 7 33 20 27 Number of birds submitted Hen Harrier 1912-39' Lowland 1 1 Mixed 4 1 2 1 Upland 3 1 1 1 Hen Harrier 1940-692 Lowland 0 Mixed 6 1 2 1 1 1 Upland 2 2 Gyrfalcon 6 1 5 Rough-legged Buzzard 8 3 3 2 Osprey 2 2 Honey Buzzard 1 1 Marsh Harrier 1 1 Snowy Owl 3 Notes: 1 Includes 2 birds submitted from Morayshire (see Table 4) 2 Includes 2 birds submitted from Morayshire and one from Aberdeenshire (see Table 4) 106 Henry McGhie SB 20(2) during September-February; the more migratory Short-eared Owl was also submitted in high numbers between July- August. The seasonal pattern of submission of Hen Harrier differed before and after 1946 (Table 6). Prior to 1946, of 8 records where the month of submission was known, 6 (75%) were submitted between Oct-Jan, one was submitted in Feb and one was submitted in March and none were submitted between April-August. After 1946, of 9 records, one was submitted between Oct-Jan, one was Submitted in Feb, and 7 (78%) were submitted between April-August. This difference was statistically significant when records for Sept- Feb and Mar-Aug were compared for the 2 periods (chi-squared, 1 degree of freedom, P=0:011)): Discussion Two factors affected the numbers of birds submitted: persecution pressure and the population size of each species. In addition to this, species varied in their desirability as trophies (and consequent financial value): it was not surprising that the 2 species submitted in the greatest numbers, namely Golden Eagle and Peregrine, were those generally considered to be the most ‘majestic’. Rare species such as Hobby, Montagu’s Harrier and Honey Buzzard, whilst not generally considered to be particularly ‘majestic’, would have been desirable as curiosities. Thus, different proportions of the numbers of each species which fell into human hands would have been likely to be submitted for preservation. As a result of Table 6 Details of all Hen Harriers submitted to Macpherson’s. Locality Vice county Date of submissionLand use type Sex/Age Feshie Bridge East Ness 30/1/15 Mixed Edderton East Ross 12/10/21 Lowland Dumphail Moray PX AAC Upland Stanstill Caithness 18/12/26 Mixed Pitcroy, Blacksboat Moray S/S Upland Kildonan E Sutherland 14/2/29 Upland Male (imm) Bettyhill Caithness 20/10/30 Mixed Leys East Ness 1934 Lowland Male Forss Caithness 16/12/40 Mixed Male Clashindarroch' Aberdeenshire 23/5/46 Upland Male Dallas Moray 8/6/46 Mixed Dallas Moray 25/4/47 Mixed Male Berriedale Caithness 26/5/47 Mixed Halkirk Caithness 20/8/48 Mixed Islay 11/2/49 Mixed Female Lochindorb East Ness 18/5/54 Upland Male Islay 22/7/56 Mixed Male Pentland Caithness 6/10/59 Mixed Female Notes: 1 Recorded as having “both legs broken” Scottish Birds (1999) this, the numbers submitted of each species were not directly comparable with each other, but intraspecific comparisons between different areas and different time periods were considered permissible. The high proportion of Hen Harrier, Merlin, Short- eared Owl and Snowy Owl submitted from Caithness, and of the more familiar farmland species from the Firthland vice counties may reflect the higher numbers of these species found there, whilst the low proportion of Peregrine which were submitted from West Sutherland possibly reflects the low density which occurs in that vice county (Ratcliffe 1980). The different seasonal patterns of submission of Peregrine and Merlin in areas of differing land use were probably related to the seasonal presence of these species in these areas (Table 5).The high numbers of Golden Eagle and Peregrine, both species which would have been valuable as mounts, which were submitted give some idea of the intensity of persecution which the populations of these and probably other species underwent, even allowing for the fact that not all birds of prey which fell into human hands would have been fit for preservation (See especially Table 3). Persecution of birds of prey was reduced during the years of the 2 World Wars as gamekeepers were called up for service, allowing some species to increase (eg Newton 1972, Watson 1977, Ratcliffe 1980). Changes in social structure following each of the World Wars resulted in declines in gamekeeping: there were 23,056 gamekeepers in Britain in 1911, 13,350 in 1920 and 4,391 in 1951 (Tapper 1992) and this would have presumably resulted in declines in persecution. The numbers of birds submitted were thus more closely related to overall levels of persecution and Evidence for persecution from taxidermists’ stuffing books 107 not related to the species’ population sizes for the period 1913-1950. The exception to this may have been the Barn Owl: the different pattern of submission of this species, peaking in the 1930s, may have been due to widespread declines of this species in the 1920s which Blaker (1933) and Shawyer (1998) attributed to climatic factors. Birds of prey did not receive complete legislative protection until 1954 with the Protection of Birds Act, although Sparrowhawk was not included until the Act was amended in 1968; the Protection of Birds Order for Inverness- shire (1951) was important enough to everyday life as to occupy 2 pages of the Inverness Courier, complete with the Gaelic name of each species. The 1954 Act was widely regarded as ineffective and unenforceable, particularly in remote upland areas where the bulk of persecution occurred and illegal persecution continues to the present time (eg Etheridge et a/ 1997, Scottish Raptor Study Groups 1997). Ineffective or not, the Act had the effect of driving persecution underground, thus putting an end to the usefulness of the stuffing books as a tool for examining persecution; declines in the numbers of birds submitted were already well underway by 1954, however persecution pressure was greatest in East Ness, particularly the grouse moor districts of Strath Errick, the Monadhliath and Strathspey. The greater numbers of birds submitted, the higher numbers of Golden Eagle and Peregrine submitted from each named locality and the high proportion of instances where 2 individuals of the same species were submitted at the same time all suggest greater persecution pressure in East Ness. Persecution pressure was also high in East Ross and East Sutherland where there are large amounts of agricultural land and forests adjacent to formerly important grouse moors, and where many large estates have 108 Henry McGhie their seats. The lower numbers of birds which were sent in from Skye, the Outer Hebrides and the western vice counties reflect the lower intensity of gamekeeping which existed in these areas and particularly the low amount of grouse moor; management of deer forest did not include the intensive persecution of birds of prey associated with grouse moor (see Brown 1976). Persecution of birds of prey would most likely have been by shooting and trapping, with pole traps and traps set at nests, and with poison baits. The seasonal patterns of submission of different species reflect the different methods of persecution used: in areas of grouse moor and deer forest many birds were submitted during the early breeding season, especially Golden Eagle and Peregrine, which suggests that they were either trapped or shot at nest sites; this method of persecution was possibly more widespread in these areas than elsewhere, and the use of relatively small numbers of traditional nest sites would have made these 2 species especially vulnerable. The higher incidence of 2 birds being submitted together from grouse moor and deer forest also suggests that breeding birds were being targetted in these areas. Owls were mainly submitted during the autumn and winter months, when long hours of winter darkenss may have made them more susceptible to persecution than during the breeding season. Two species detailed in the stuffing books, namely Osprey and Hen Harrier, underwent significant changes in status during the period under study. Osprey recolonised Scotland in the 1950s after an absence of almost 40 years although individuals had been seen with increasing frequency since the end of the Second World War (Brown and Waterson 1962). The records of Hen Harrier contained SB 20(2) within the stuffing books are of special interest: Hen Harrier was probably extinguished as a breeding species in mainland Scotland by the beginning of the 20th century and there were very few confirmed records of breeding up until the Second World War, after which time the species recolonised the mainland. Campbell (1957), Blake (1961) and Watson (1977) all speculated that the Hen Harrier was already established by the 1940s although they held different opinions as to the extent to which recolonisation had taken place by 1946. Three lines from the stuffing books suggest that there was indeed a great increase during or shortly after the Seond World War (Table 6). Firstly, Hen Harrier was the only species, excepting very rare migrants, which was submitted in significant numbers after 1940 and the numbers submitted of each of the regularly breeding species were already in sharp decline by this time. Secondly, the fact that Hen Harrier was not being submitted before 1946 when ail other birds of prey were undergoing heavy persecution suggests that they were not breeding in any numbers although Campbell (1957) considered that it bred intermittently in the northern Highlands but does not provide any evidence. Thirdly, the change in the species status around the time of the Second World War is most clearly illustrated by the difference in the seasonal pattern of submission before and after 1946: prior to 1946 Hen Harrier was only submitted during the winter of pre breeding period. (February- March) whilst after 1946 it was most frequently submitted during the breeding season (April-August). Watson (1977) speculated that the recolonisation of the mainland had advanced considerably by 1946 because of the widely scattered nature of the first breeding records after the Second World War, but Etheridge et a/ (1997) Scoitish Birds (1999) Evidence for persecution from taxidermists’ stuffing books 109 calculaied that Hen Harrier populations can rapidly recover following the cessation of persecution. The recovery of the Hen Harrier is normally aiiributed to declines in gamekeeping during and afier the war, and to poorer heaiher burning practices leading io an increase in suitable nesiing habitat. li is inieresiing to noie that Hen Harrier was noi submitied from the Outer Hebrides, where it was known io breed during ithe period under siudy (Waison 1977), although small numbers of other species were suomitied for preservaion ai this time (Table 3). The Hen Harrier persisted as a British species only in the Outer Hebrides. where ii was preserved. and in Orkney, and ithe evidence of the stuffing books corroboraies this. Nethersole- Thompson (in Waison 1977) knew of Hen Harriers nesting in Moray in 1945, and the 1946 and 1947 records in the siuffing books may refer to this locality. The stuffing books provide an insighi inio a time oi different values, when birds of prey were regarding as vermin and ihe conservation movemeni was in iis infancy. They are fascinating naiural and social history documenis and other workers are encouraged io make use of sources of historical data such as these while this is siill possible. Acknowledgedments | am very grateful to Inverness Museum and Art Gallery for allowing me to work on their collections and to Stephen Moran (Assistant Curator Natural Sciences) for his help and advice. References Anon 1996. Road Ailas of Great Britain Automobile etn Anon 1964 Road Ailas of Great Britiain eee mew and roan EACELIGh. Baxier E V & Rinioul L J 1953 Birds in Scotland Oliver and Boyd, Edinburgh. Blake E 1961 Return of the Hen Harrier. Scoftish Field, February issue. Blaker G B 1933 The Bam Owl in England. Bird Notes and News. 15:169-192. Brown L 1976 Birds of Prey. Collins New Naturalist. London. Brown P & Waiersion G 1962 The Reium of the sprey. Heinemann. London. Campbell L W 1957 The Rarer Birds of Prey: their preseni siaius in the British Isles. British Birds 50:4. Etheridge B, Summers R W & Green RE 1997 The efiecis of illegal killing and desiruciion of nesis by humans on the population dynamics of the Hen Harrier Circus cyaneus in Scoiland. Joumal of Applied Ecology 34:1081-1105. Harvie-Brown J A & Buckley T E 1895 A Vertebrate Fauna of the Moray Basin, David Douglas, Edinburgh. Newton | 1972 Birds of Prey in Scotland: conservaiion problems. Scoitish Birds 7:5 —- Newion 1 1979 9 Popul lation Ecology in Raptors. T & A D Poyser, encanta Pearsall J A 1952 Mountains and Moorlands. Collins New Naturalist, London. Prebble J 1963 The Highland Clearances. Penguin, Harmondsworih. Raiclifie D A 1980 The Peregrine Falcon. T & AD Poyser, Calton. Ritchie J 1920 The Influence of Man on Animal Life in Scotland. Cambridge University Press, Cambirdge. Scottish Raptor Study Groups 1997 The illegal persecution of raptors in Scotland. Scoftish Birds 19:65-85. Shawyer C 19 Chelmsford. 98 The Bam Owl. Ariequin Press, 110 Henry McGhie SB 20(2) Tapper S 1992 Game History: an ecological perspective. Game Conservancy Council, Fordingbridge Temperley G W 1951 The Natural History of Raasay. Scottish Naturalist 63 Watson D 1977 The Hen Harrier. T & AD Poyser, Berkhamsted. Henry A McGhie, Nurse’s House, West Road, Muir of Ord, Ross-shire lV6 7TD Revised manuscript accepted October 1999 re eid PR att: | EGS Golden Eagle chasing winter plumaged Ptarmigan David Mitchell Wi Scottish Birds (1999) Short Notes 20: 111-112 SHORT NOTES Recolonisation of the Isle of May by Sandwich Terns during 1999 The recolonisation of the island by Sandwich Terns Sterna sandvicensis was the most noteworthy event ofthe 1999 breeding season on the Isle of May. Sandwich Terns have not bred on the island in numbers since 1956 with only one record in the interim of a single pair raising one chick in 1990. In the past 2 seasons large numbers of Sandwich Terns have been present around the island throughout the season and, in 1998, 3 birds laid eggs very late in the season before abandoning them. In 1999 birds returned in good numbers with 291 individual birds noted in the colony on the evening of June 25th The first incubating birds were noted on 20 June with 3 birds sitting, and large numbers were present displaying. This is quite a late date for first laying and would tend to indicate that these birds were either inexperienced younger birds or had attempted to breed elsewhere earlier in the season and had failed. The maximum number of sitting birds was counted on 4 July with 110 birds on eggs. The birds formed a tightly packed sub-colony within the Arctic/Common Tern nesting area at the summit of the island around the Beacon. The first chick was noted on 21 July. On 11 August chicks began to move around the colony as part of a large creche with 87 large chicks counted. The group was extremely mobile, and moved east across the island despite rocky terrain and large nettle patches. They crossed Holyman’s Road into the Burrian area and moved on to the seaward rocks from 12th when chicks moved below the tideline. Chick numbers had reduced to 4 or 5 by the next day, with most chicks assumed washed away or predated by gulls, and by the 14th no chicks could be found. This is awelcome return of abreeding species which previously bred on the island in good numbers, with a population of 1400-1500 pairs in 1946 (Eggeling, W J 1985 The Isle of May. ), and despite the disastrous end to the season itis hoped thatthe colony will continue to grow and succeed in fledging chicks in future seasons. Darren Hemsley, 3 Carselea Road, Invergowrie, Dundee DD2 5AW Revised mansucript accepted August 1999 Sandwich Tern ~— a Steven Brown 112 Short Notes SB 20(2) Interactions between breeding Choughs and Fulmars on Islay The Isle of Islay holds a significant proportion of the Scottish Chough Pyrrhocorax phyrrhocorax population but recent surveys suggest that the numbers of Choughs on the island are declining (Madders et a/ 1998 Distribution and foraging habitat preferences of Choughs on the Oa peninsula, Islay. Scottish Birds 19:280-289). Several reasons for the decline have been put forward but it seems likely that there is no one cause, rather anumber of factors combine to reduce the breeding population. | have visited Islay every year for the last 20 years and have known certain Chough sites to be successful year after year. More recently, 2 particular sites have failed despite adults being present throughout the breeding season. These nests are on a stretch of coastline comprising low cliffs with numerous long and narrow gullies (up to 50m long and 5-10m wide). The nests themselves are situated towards the inland end of two of the larger gullies in inaccessible caves. While walking along this coastline in June 1998 | noticed that many of the gullies, including those containing the Chough nests, were inhabited by several (up to 15) pairs of nesting Fulmars Fulmaris glacialis. These, | know, were not present 12 years ago and were not recorded in the 1986 seabird survey on Islay (M Ogilvie pers comm). On the same walk | observed that both pairs of Choughs were present around their nesting gullies and at each site | witnessed interactions between the Choughs and Fulmars. The Choughs seemed keen to get into their nest sites but as they attempted to fly in, aggressive aerial attacks from the nesting Fulmars deterred them. Having seen these birds running the gauntlet to get to their nests, | was not surprised that they had failed to rear young in recent years. Similar interactions have been recorded in Wales where, in 1997, Fulmars trampled a Chough nest containing eggs and prevented the Chough from returning (A Cross pers comm). My observations suggest there has been an increase in the Fulmar population on Islay although no recent surveys have been carried out. There are many sites where the 2 species seem able to breed successfully in close proximity. However, it is possible that problems are associated with specific topographic features (eg gullies) where the approaches Choughs use to reach their nests are limited. lf Fulmars continue to spread and colonize more of the rocky coastline there Is a risk that they will have an impact on other Chough pairs and thus add to the factors already contributing to the decline of Choughs on Islay. Derek Hayward & Sian Davies, Logwood Cottage, Mealbank, Kendal, Cumbria LA& 9DJ Revised manuscript accepted August 1999 Advice to contributors Authors should bear in mind that only a small proportion of the Scottish Birds readership are scientists, and should aim to present their material concisely, interestingly and clearly. Unfamiliar technical terms and symbols should be avoided wherever possible and, if deemed essential, should be explained. Supporting statistics should be kept to a minimum. All papers and short notes are accepted on the understanding that they have not been offered for publication elsewhere and that they will be subject to editing. Papers will be acknowledged on receipt and are normally reviewed by atleast 2 members of the editorial panel and, in most cases, also by an independent referee. They will normally be published in order of acceptance of fully revised manuscripts. The editor will be happy to advise authors on the preparation of papers. Reference should be made to the most recent issues of Scottish Birds for guidance on Style of presentation, use of capitals, form of references, etc. Papers should be typed on one side of the paper only, double spaced and with wide margins and of good quality; 2 copies are required and the author should also retain one. We are happy to accept papers on computer discs; however, please state the type of word processing programme used. Contact Sylvia Laing on 0131 556 6042 if you wish further information on this. Headings should not be underlined, nor typed entirely in capitals. Scientific names in italics should normally follow the first text reference to each species unless all can be incorporated into a table. Names of birds should follow the official Scottish list (Scottish Birds 1994 Vol 17:146-159). Only single quotation marks should be used throughout. Numbers should be written as numerals except for one and the Start of sentences. Avoid hyphens except where essential eg in bird names. Dates should be written:...on 5 August 1991...... but not on the 5th (if the name of the month does notfollow). Please do not use headers, footers and page numbers. Please note that papers shorter than c700 words will be treated as short notes, where all references should be incorporated into the text, and not listed at the end, as in full papers. Tables, maps and diagrams should be designed to fit either a single column or the full page width. Tables should be self explanatory and headings should be kept as simple as possible, with footnotes used to provide extra details where necessary. Each table should be on a separate sheet. Maps and diagrams should be either good quality computer print outs in black and white (please do not use greyscale shading) or in black ink and be camera ready, but drawn so as to permit reduction from their original size. Scottish Birds Volume 20 Part 2 December 1999 Contents Main papers Raven population size and distribution on the Isle of Islay in winter. M Madders & F M Leckie 49 Barnacle Geese on the Solway: 1990-1996. J M Black, D Patterson, P Shimmings & E C Rees 63 Breeding population estimates for Lapwing,Oystercatcher, and Curlew in Scotland: results of the 1998 BTO Lapwing Survey. A M Wilson & S J Browne 3 Bird numbers in an Aberdeenshire glen in March-June 1987-99. D Jenkins & A Watson 81 Breeding success of Red-throated Divers on Orkney Mainland 1973-1998. — C J Booth 94 Persecution of birds of prey in north Scotland 1912-69 as evidenced by taxidermists’ stuffing books. Henry McGhie 98 Short notes _ : Recolonisation of the Isle of May by Sandwich Terns during 1999. | Darren Hemsley he Interaction between breeding Choughs and Fulmars on Islay. Derek Hayward & Sian Davies AZ Advice to contributors Inside back cover Published by the Scottish Ornithologists’ Club, 21 Regent Terrace, Edinburgh EH7 5BT. © 1999 THE JOURNAL OF THE SOC Vol 21 No 1 June 2000 Amendments to the Scottish List Breeding birds of the Isle of May Wintering seafowl in Scapa Flow Greylag Geese breeding in Shetland Rookeries in East Ross Scottish Birds The Journal of the Scottish Ornithologists’ Club Editor: Dr S da Prato Assisted by: Professor D Jenkins, Dr J B Nelson, Dr | Bainbridge, Dr M Marquiss and R Swann Business Editor: The Secretary SOC, 21 Regent Terrace Selinetealy ERi7 5Bih (tel 0131-556 6042, fax 0131 558 9947). Scottish Birds, the official journal of the Scottish Ornithologists’ Club, publishes original material relating to ornithology in Scotland. Papers and notes should be sent to The Editor, Scottish Birds, 21 Regent Terrace, Edinburgh EH7 5BT. Two issues of Scottish Birds are published each year, in June and in December. Scottish Birds is issued free to members of the Scottish Ornithologists’ Club, who also receive the quarterly newsletter Scottish Bird News, the annual Scottish Bird Reportand the annual Raptor round up. These are available to Institutions at a subscription rate (1997) of £36. The Scottish Ornithologists’ Club was formed in 1936 to encourage all aspects of ornithology in Scotland. It has local branches which meet in Aberdeen, Ayr, the Borders, Dumfries, Dundee, Edinburgh, Glasgow, Inverness, New Galloway, Orkney, St Andrews, Stirling, Stranraer and Thurso, each with its own programme of field meetings and winter lectures. The Waterston Library at the Club’s headquarters at 21 Regent Terrace, Edinburgh EH7 5BT is the most comprehensive ornithological library in Scotland, and is available for reference during office hours (Monday to Friday, 0930-1630 hrs). (But please phone beforehand). SOC annual membership subscription rates Direct Debit Other methods Adult £18.00 £20.00 Family (2 adults and any children under 18) living at one address £27.00 £30.00 Junior (under 18, or student under 25) £7.00 £8.00 Pensioner/unwaged £10.00 £11.00 Pensioner Family (2 adults living at one address) £14.50 £16.00 Life £360.00 £400.00 Life Family £540.00 £600.00 All subscriptions may be paid by Direct Debit and Covenanted. Subscriptions paid by Direct Debit greatly assist the Club. Please ask for a Direct Debit form by phoning the Secretary at the above address. Published by the Scottish Ornithologists’ Club, 21 Regent Terrace, Edinburgh EH7 5BT Printed by Meigle Print, Block 11, Units 1 & 2, Tweedbank Industrial Estate, Galashiels TD1 3RS Scottish Birds (2000) 21:1-5 1 Amendments to the Scottish List Incorporating an update to records of species recorded in Scotland on 5 or fewer RONALD W FORRESTER The Scettish-Birds Records Committee is responsible for maintaining the Scottish List first published in 1994 (Scottish Birds 17:146-159). The Committee also lists acceptable records for all species that have occurred in Scotland on 5 or fewer occasions. This is the third subsequent report of the Committee, the others being in Scottish Birds 18:129-143 and 19:259-261. For over 100 years the British Ornithologists’ Union (BOU) has maintained a list of birds that have been recorded in Britain and Ireland. This is undertaken by the BOU’s Records Committee (BOURC), which periodically publishes checklists and reports. The first principle that was established for the Scottish List was that it would use identical categories (A, B, C and D) to the BOU. During 1998, BOURC introduced a revised system of categorisation. In order to maintain the link it is therefore necessary to revise the definitions of the categories used for the Scottish List. The main changes are that the start date for Category A has been changed from 1 January 1958 to 1 January 1950, Categories C and D have been re- defined, with a new Category E being created to accommodate species occurring as escapes. The Scottish List — Categories A Species which have been recorded in an apparently natural state at least once since | January 1950. for the Scottish Birds Records Committee B_ Species which were recorded in an apparently natural state at least once up to 31 December 1949, but have not been recorded subsequently. C Species that although originally introduced by man, either deliberately or accidentally, have established breeding populations derived from introduced stock, that maintain themselves without necessary recourse to further introduction. Category C has been further subdivided to differentiate between various groups of naturalised species: Cl naturalised introductions, C2 naturalised establishments, C3 naturalised reestablishments, C4 naturalised feral species. C5 vagrant naturalised species. D Species that would otherwise appear in Category A or B except that there is reasonable doubt that they have ever occurred in a natural state. Category D species do not form any part of the species totals and are not regarded as members of the Scottish List. E Species that have been recorded as introductions, transportees or escapees from captivity and whose breeding populations, if any, are thought not to be self sustaining. Category E species form no part of the Scottish List. BOURC published on their web site during 1999 a provisional list of 276 species in Category E of the British List. A list of Category E species recorded in Scotland has not yet been produced. 2 SBRC SB 21(1) As a result of the date change for Category A, Magnificent Frigatebird, Black-billed Cuckoo, Crested Lark and Buff-bellied Pipit all move from Category B to Category A of the Scottish List. Egyptian Goose has appeared in Category D of the Scottish List but the species no longer falls within the new definition for that Category. The species does, however, remain on Category C of the British List due to the established breeding population in East Anglia. If it can be proved that at least one Scottish record refers to vagrants from the East Anglia population the species would meet C4 criteria. Until such time it requires to be removed from Category D, although it will probably be included in Category E. Northern Flicker, previously in Category D, has been omitted from the BOURC’s new list and as aresult should now be removed from our Category D. We expect that BOURC will included it in Category E. Red Kite previously Category A and D4 should now be Categories A and C, as a result of the re- introduction programme which has resulted in an increasing population since 1992, when pairs established from released birds bred for the first time. There now appears to be a self sustaining population which meets the definition for Category C3. Table 1 BOURC revised generic names. previous scientific name Great White Egret Sociable Plover Gull-billed Tern Savannah Sparrow Song Sparrow Egretta alba Chettusia gregaria Gelochelidon nilotica Ammodramus sandwichensis Zonotrichia melodia The genus Butorides was previously considered to be monotypic. BOURC has now followed most authorities in considering the form found in North America to be a separate species from the cosmopolitan Butorides striatus, now commonly called Striated Heron. The only Scottish record is of the North American species, now called Green Heron, and given the Latin name Butorides virescens. The BOURC has adopted new generic names for the following species. It is our practice to follow their use of scientific names and we have therefore amended the scientific names accordingly in Table IL. As aresult of recent decisions by the British Birds Records Committee and BOURC, the following additional records have been accepted, of species recorded in Scotland on 5 or fewer occasions. Pied-billed Grebe Podilymbus podiceps Airthrey Loch, Stirling, Forth Area 3-7 June 1998 (British Birds 92:557); Loch Peallach, Mishnish Lochs, near Tobermory, Mull 8-15 June 1998 (British Birds 92:557).5" and 6" Scottish records. Removed from the list of species recorded on 5 or fewer occasions. new scientific name Ardea alba Vanellus gregarius Sterna nilotica Passerculus sandwichensis Melospiza melodia Scottish Birds (2000) Amendments to Scottish List 3 Red-breasted Goose Branta ruficollis The 6* to 8* Scottish records occurred during 1998 and the species is now removed from the list of those recorded on 5 or fewer occasions. Black Duck Anas rubripes Female Alturlie. Highland 13 January — 1 April 1997 (British Birds 92:564).6" Scottish Record. Removed from the list of species recorded on 5 or fewer occasions. Lesser Scaup Aythya ajfinis Male, Martnaham Loch. Ayrshire 13-17 September 1997 (British Birds 91-467). 5* Scottish record. The male at St John’s Loch, Caithness from I February 1996, previously accepted as 4* Scottish record (Scottish Birds 19:259). remained until 20 January 1997 (British Birds 91:467). There were 3 further records during 1998 and the species is removed from the list recorded on 5 or fewer occasions. Semipalmated Sandpiper Calidris pusilla Juvenile Pool of Virkie and Scatness. Shetland 11-22 September 1996 (British Birds 90:471). 3™ Scottish record. Adult. Deerness, Orkney 25 July 1997 (British Birds 91-473). 4° Scottish record. Western Sandpiper Calidris mauri 1*S or adult, Musselburgh. Lothian, 9-25 August 1997, probably since 25 July (British Birds 91:474). 3“ Scottish record. Juvenile. Deemess. Orkney, 28 September — 3 October 1998 (British Birds 92:572). 4° Scottish record. Upland Sandpiper Bartramia longicauda Foula, Shetland 2 September 1996, presumed same 14°-15* (British Birds 90:478) 6° Scottish record. Removed from the list of species recorded on 5 or fewer occasions. Sooty Tern Sterna fuscata moved from Category B to Category A. Isle of May, Fife, 14 July 1989 (British Birds 91:495). 3° Scottish record. Little Swift Apus affinis etlar, Shetland. 29 May 1997 (British Birds 91:497). 3™ Scottish record. Blue-cheeked Bee-eater Merops superciliosus added to Category A. Bressay., Asta. Tingwall Valley and Lerwick area, Shetland 20 June — 3 July 1997 (British Birds 91-497). 1* Scottish record. Cedar Waxwing Bombycilla cedrorum transferred from Category D to Category A. The record of one at Noss, Shetland 25-26 June 1985, originally placed in Category D (Scottish Birds 18:143) is now accepted as referring to a genuine vagrant (British Birds 91:500). Itbecomes the Ist Scottish record. The record predates the Nottmgham bird 20 February -8 March 1996 as the first record for Britain & Ireland (Jbis 140:182). Isabelline Wheatear Oenanthe isabellina 1*W, Fair Isle, Shetland 20 —29 September 1998 (British Birds 92:591). 3™ Scottish record. Hermit Thrush Catharus guttatus etlar, Shetland 30 April — 1 May 1998 (British Birds 92:593). 3™ Scottish record. Dusky Thrush Turdus naumanni The 2 1993 records assigned to this species (Scottish Birds 18:140) should be deleted — the records should refer to Black-throated Thrushes. There are only 3 Scottish records of Dusky Thrush. Hume’s Warbler Phylloscopus humei added to Category A. BOURC has recently acknowledged Hume’s Warbler as a separate species. It wasformerly considered a race of Yellow-browed Warbler Phylloscopus inornatus. Auchmithie, Angus & Dundee 13 October 1991 (British Birds 91:508). South Ronaldsay, Orkney 4-6 November 1994 (British Birds 91:508). Isle of May, Fife 10 November 1994 (British Birds 91:508). 4 SBRC SB 21(1) Bullers of Buchan, NE Scotland 10-13 November 1994 (British Birds 91:508). 1 to 4" Scottish records. Eastern Bonelli’s Warbler Phylloscopus orientalis added to Category A. Sumburgh, Shetland 27-28 August 1998 (British Birds 92:599). 1* Scottish record. In 1997, BOURC split Bonelli’s Warbler into 2 species: Western Bonelli’s Warbler Phylloscopus bonelli and Eastern Bonelli’s Warbler Phylloscopus orientalis. Following are evaluation of all British records (British Birds 91:122-123), 6 Scottish records are now accepted as Western Bonelli’s Warblers and the remaining 14 as Western/Eastern Bonelli’s Warblers. Southern Grey Shrike Lanius meridionalis IstW, trapped Fair Isle, Shetland 21 September 1956 (British Birds 50:246-249 and 90:70). This pre dates the 1964 bird (Scottish Birds 19:260) as the Ist Scottish record. IstW male North Ronaldsay, Orkney 14 September — 16 October 1994, trapped (British Birds 90:70). Probably IstW, Boddam, Shetland 7-10 November 1994 (British Birds 90:70). Papa Westray, Orkney 11-26 November 1994 (British Birds 90:70). These records bring the Scottish total to 5, all of which are considered to be of the race pallidirostris. Blackpoll Warbler Dendroica striata Stornoway, Lewis, Outer Hebrides 26-29 October 1996 (British Birds 90:509). 4" Scottish record. Common Yellowthroat Geothlypis trichas 1“S female, Baltasound, Unst, Shetland 16-23 May 1997 (British Birds 91:513). 2™ Scottish record. Yellow-browed Bunting Emberiza chrysophrys Hoy, Orkney 4-5 May 1998 (British Birds 92:604). 3 Scottish record. Bobolink Dolichonyx oryzivorus Durigarth, Shetland 28 September — 5 October 1998 (British Birds 92:606). 4 Scottish record. Category D Falcated Duck Anas falcata add to Category D. Male, Merryton Haugh, Clyde 8 — 15 March 1998 and 1 November 1998-24 May 1999 (British Birds 92:607). 1“ Scottish record. As a result of the above changes, 9 species have been added to Category A and 5 have been removed from Category B. In addition 3 species have been removed from Category C with one added. The revised totals are now: Category A 467 Category B g Category C 6 CategoryD _10 SBRC wish to thank Angus Murray and Mike Pennington for drawing to our attention previous errors now corrected. 1999 records awaiting acceptance include Crag Martin in Orkney on 3 May, Short-billed Dowitcher at Rosehearty, Aberdeen 11—24 September and Royal Tern in Lothian on 9 August. All these species are potential additions to Category A. The ‘large white headed gull’ complex continues to cause taxonomic difficulties. Although BOURC has not separated any of the ‘yellow-legged gulls’ from Herring Gull Larus argentatus, it appears only a matter of time before this happens. Accepted Scottish records of the race michahellis are now well into double figures and there are currently descriptions under consideration by BBRC of birds at Doonfoot, Ayrshire during 1992, 1997 and 1999, which, if accepted, will become the first Scottish Birds (2000) Amendments to Scottish List B Scottish records of the race cachinnans. If BOURC follows the lead already taken in several European countries, these 2 forms could become separate species Yellow-legged Gull and Caspian Gull. Progress is now being made in producing a list of subspecies recorded in Scotland. The current Scottish List appears in full on the SOC’s web site www.the-soc.org.uk where there is also a full list of records of those species occurring in Scotland on 5 or fewer occasions. The Scottish Birds Records Committee, which is responsible for maintaining the Scottish List now. consists of Kevin Osborn, David Clugston, Bruce Forrester. Ian Andrews. Colin Crooke, Martin Gray. Dougie Dickson and Ron Forrester, Secretary. Ronald W Forrester. Secretary, Scottish Birds Records Committee, The Gables, Eastlands Road, Rothesay, Isle of Bute PA20 9JZ Red-breasted Goose with Barnacle Geese Sam Alexander 6 Scottish Birds (2000) 21:6-14 SB 21(1) Breeding birds of the Isle of May, Firth of Forth, 1972-99 M P HARRIS, S WANLESS, I DARLING & C GALLACHER During the last 28 years the numbers of auks and Eiders breeding on the Isle of May have increased dramatically, terns have recolonised and gull numbers have declined. With the exception of the feral pigeon, landbirds have done less well. Lapwing, Stock Dove, Swallow, Carrion Crow, Dunnock, Blackbird, Song Thrush and Meadow Pipit have not bred in recent years and numbers of Pied Wagtail and Starling are much reduced. Ringed Plover, Stock Dove and, probably, Black Redstart bred for the first time. The Isle of May is the largest island (57 ha) off the east coast of mainland Britain. The west side 1s bounded by sheer cliffs some 45m high whereas on the east the land slopes gently into the sea. The centre of the island 1s well vegetated with a mixture of maritime grasslands and now defunct gardens and there are small areas of semipermanent freshwater. Thus there is a variety of habitats for breeding birds (Eggeling 1960). The last review of breeding birds was that of Eggeling (1974) who documented nesting by 41 species. Since 1972 we have been involved in research on the island and at least one of us has been resident there each year from April to July. This paper summarizes information on the status and numbers of the breeding birds collected by ourselves and the organisations that we represent, over the last 28 years. We refer to Eggeling’s 2 papers for earlier records and changes of status. Methods Information comes from our own observations, those of visitors to the Isle of May Bird Observatory extracted from the observatory’s records and reports up to 1984 which were published in Scottish Birds and subsequently in the annual Isle of May Bird Observatory Reports, and the Warden’s reports and reports of Scottish Natural Heritage. Records up to, and including 1973, are detailed in Eggeling (1960,1974) and some later counts of seabirds can be found in Harris & Galbraith (1983). The seabird populations were counted in 1969 as part of Operation Seafarer and for some species we have calculated rates of increase (Yoper annum based on linear regression of log transformed counts) starting with these counts. In the figures we plot counts of seabirds back to 1960 to put recent changes into a longer context. For species not shown in the figures or where counts are not easily summarised in the text, annual estimates of the populations are given in an appendix. Counting units were occupied nests omitting known relays for Shag, Eider, gulls and terns, Oystercatcher, Lapwing, apparently occupied sites for Fulmar and burrows for Puffin, and pairs for most other species; Counts of individual Guillemots and Razorbills were converted to pairs using year specific correction factors. For Shelduck, Oystercatcher and Rock Pipit we assume that the numbers of occupied territories reflected the breeding population. Since the early 1980s seabirds have been counted in a systematic way in late May or early June. Species accounts Fulmar Fulamrus glacialis (Fig 1a) Breeding was first proved in 1930. In 1969, there were 53 occupied sites and the population has Scottish Birds (2000) since increased at an average rate of 6.8% pa. The highest count was 382 sites in 1997. Shag Phalacrocorax aristotelis (Fig 1b) Counts of nests can seriously underestimate the true population since in some years (eg 1973-76, 1993) many adults present in the colony did not nest (Aebischer 1986, Aebischer & Wanless 1992). Numbers increased during much of the 20" century and reached a peak count of 1919 nests in 1987. Numbers then fluctuated around c1,500 nests before declining. Figure 1b gives the impression that this decline occurred mainly between 1992 and 1993. This was not the case, however, since studies of colour ringed adults indicated clearly thatno unusual mortality occurred over this winter, rather many adults present on the island in 1993 failed to breed and later died during a major mortality of adults in February 1994. It was this wreck that was responsible for the population crash (Harris & Wanless 1996). The population recovered slightly between 1995 and 1997 but declined again between 1998 and 1999 when the count of 259 nests was the lowest since 1955. Eider Somateria mollissima (Fig 1c) Starting in the early 1970s, numbers increased greatly (Calladine et al 1995). The highest count was 1,200 nests in 1999, and the average rate of increase between 1978, when the first systematic count was made, and 1999 was 7.6% pa. Shelduck Tadorna tadorna This species was first proved to breed in 1936 and now nests annually. Numbers peaked at 10-12 pairs between 1984 and 1987 but decreased to 3- 6 pairs in the late 1990s. Moorhen Gallinula chloropus A pair bred at the pools at the north end of Rona in 1977 and 1978. Breeding birds of the Isle of May, Firth of Forth, 1972-99 7 Lapwing Vanellus vanellus The Lapwing colonised the island in 1976 and bred in each year until 1986. The peak number was 7 pairs in 1981. Ringed Plover Charadrius hiaticula A pair laid 2 eggs at Kirkhaven beach in 1977 but these were soon lost due to disturbance. This was the first recorded nesting on the island. Oystercatcher Haematopus ostralegus (Fig 1d) The population increased from 15 pairs in 1973 to 39 pairs in 1989 (Harris & Wanless 1997). Numbers then declined to 28 pairs by 1997. In 1999 there were 30 territorial pairs. Great Black-backed Gull Larus marinus The first pair bred in 1962 and numbers increased to 4 pairs between 1969 and 1972. The species then ceased to nest, possibly as a result of the culling of the next 2 species. One pair bred in 1982 and 1985 since when numbers have increased steadily up to the present with 19 pairs recorded in 1999. Herring Gull Larus argentatus (Fig |e) This species first bred in 1907 and the numbers increased at arate of 13% pa to about 15,000 pairs in 1972 (Eggeling 1974, Chabrzyk & Coulson 1976). It was considered that such large numbers of gulls were harming the island’s fauna and flora so, between 1972 and 1986, many adults of this and the next species were killed. Numbers were reduced to 1,711 nests in 1988 but have since increased. The 1999 count was 3,115 nests. Numbers, culling and population dynamics of this and the next species are well documented — see Wanless et al 1996 for the most recent review. S M P Harris, S Wanless, I Darling, C Gallacher SB 21(1) Figure 1 Changes in the numbers of Fulmar, Shag, Eider, Oystercatcher, Herring and Lesser Black-backed Gull and Kittiwake on the Isle of May, 1961-99. a) Fulmar Go jo) jo) L Occupied sites 8 i aamenrey Fa Natl eee oe ye Vee 1960)> 918705) 1960) 1990) 9 2000 c) Eider 1400. 1200 | 1000. a a TT a rar Nr, rr eames | 1960 1970 1980 1990 2000 e) Gulls 16000 . Pein | ied - @ - -Herrring | 12000, ¢@ | ! gl ae | —e—LBBGull | 38000, e/ Boe ees alas x = 4000, 0| ge 26 | 1960 1970 1980 1990 2000 Lesser Black-backed Gull Larus fuscus (Fig le) Breeding was first recorded in 1930 and numbers increased at 14% pato 2,100 pairs in 1972 (Duncan 1981). Control measures apparently had less b) Shag 2000 + 1500 | WV w 1000 - za 500 - 0 i ee Slice “aaa oe a ae oe, MiSs = ol 1960 1970 1980 1990 2000 d) Oystercatcher 40 39 | Seen w 20 + fot 10 (ot ——— = 1960 1970 1980 1990 2000 f) Kittiwake 9000 - yy 6000 : — 2] i) = 3000 + 8) i a oe el =F ial (gues 2 See wae! 1960 1970 1980 ISSO 2008 impact on this species than on the Herring Gull but even so numbers had declined to 534 pairs by 1987. Numbers subsequently increased to 1,635 nests in 1995. There were 1,519 nests in 1999. Scottish Birds (2000) Kittiwake Rissa tridactyla (Fig 1f) Numbers of this species increased throughout most of the 20" century and by 1969 there were 3,100 nests. The increase continued, averaging 4.8% pa 1969-90 to a peak count of 8,129 nests in 1990. Apart froma very low count in 1994, which was due to many adults not nesting, the population then fluctuated between 6,300 and 7,600 nests until 1997 after which numbers started to decline. In 1999 there were 4,196 nests. Common Tern Sterna hirundo (Fig 2a) In the 1940s this was the most numerous species on the island and some 5,000-6,000 pairs were estimated to have bred in 1946 and 1947. Numbers then declined rapidly and no breeding was recorded between 1958 and 1972. Isolated breeding attempts Breeding birds of the Isle of May, Firth of Forth, 1972-99 9 were recorded in 1973, 1979, 1980, and 1981, and a small colony of 15 nests was reestablished in 1982 (Wanless 1988). There were 37 nests the next year and the population increased at 13.8% pa between 1983 and 1999 when the count of 415 nests was the highest since 1947. Arctic Tern Sterna paradisaea (Fig 2a) The species was common during the 1930s, eg there were up to 800 pairs in 1936, but numbers declined and the species became extinct as a nesting species in 1957. A single pair bred in 1980 but it was not until 1984, when 19 pairs bred, that real recolonisation occurred (Wanless 1988). There were 87 pairs the following year and the population subsequently increased to 737 nests in 1999. Figure 2a suggests that the rate of increase slowed in the late 1990s. Figure 2 Changes in the numbers of Common and Arctic Terns, Guillemot, Razorbill and Puffin on the Isle of May, 1968-99. a) Terns ] | —®—CommontTern | ° | | ’ 600 | | +> @--Arctic Tern Gor 4 | Paes ® 400 1 = | 200 ; j 0 a a T TN 1960 1970 1980 1990 2000 c) Razoprbill 4000 - 3000 - ad : 2000 - De | 1000 - ; = ws : 1960 1970 1980 1990 2000 b) Guillemot 18000 - 45000 12000 + 9000 - 6000 | 3000 - Pairs ae a= 1960 1970 1980 1990 2000 d) Puffin 50000 - 40000 - 30000 | 20000 | 10000 - 1960 1970 1980 1990 Burrows 2000 10. M P Harris, S Wanless, I Darling, C Gallacher SB 21(1) Roseate Tern Sterna dougallii During the 1930s and 1940s up to 20 pairs bred and a single pair nested in 1956. There was no further nesting until a single pair bred in 1995 (Harris & Harding 1995) and 1996. Sandwich Tern Sterna sandvicensis Up to 1,400 pairs bred during the 1940s. The population became extinct on the Isle of May in 1957. A single pair laid in 1990, there were 2 pairs in 1998 and at least 110 pairs in 1999 (Hemsley 1999). Guillemot Uria aalge (Fig 2b) In 1969, the population was estimated at 9,000 birds which suggests a breeding population of about 6,000 pairs. Annual counts of incubating and brooding individuals between 1972 and 1976 ranged from 3,360 to 3,920. Such methods seriously underestimate the true breeding population and counts of 9,730 and 11,300 individual birds in 1974 and 1975 respectively indicate that 6,500-7,500 pairs were present. By 1981 the population had increased to 11,250 pairs and there was a further increase to 14,750 pairs by 1983. The population declined somewhat during the next 8 years but between 1991 and 1999 numbers increased again at an average rate of 5.8%pa. Razorbill Alca torda (Fig 2c) Between 1969 (200-300pairs) and 1999 (3,429 pairs) the population increased at an average rate of 8.1%pa. Puffin Fratercula arctica (Fig 2d) Since the 1960s, when there were only a few pairs on the Isle of May, the population has increased dramatically with whole island counts of 2,000 pairs in 1970, 12,200 burrows in 1984, 20,100 burrows in 1992 and 42,000 burrows in 1998, an average rate of increase of 11%pa. Feral Pigeon Columba livia A common breeder whose number increased from less that 20 pairs in the 1970s and early 1980s to ‘hundreds’ by 1994. This increase has been associated with birds starting to nest down Puffin and Rabbit burrows. Stock Dove Columba oenas One to 3 pairs bred each year between 1979 and 1989. These are the only records of breeding on the island. Swallow Hirundo rustica The species bred each year between 1972 and 1994 with a maximum of 6 pairs in 1990. Carrion Crow Corvus corone One pair has been present each year since 1921. In some years a nest is built but only rarely does laying occur. In our period eggs were recorded only in 1984, 1990, 1991 and 1994. Wren Troglodytes troglodytes Wrens bred in 1996, 1997 and probably also in 1975 and 1998, when 2-3 males were recorded singing. Song Thrush Turdus philomelos Two pairs bred in 1972, following a colonisation of the island by a single pair in 1971, 3 pairs in 1973 and one pair in 1974. Blackbird Turdus merula Two or 3 pairs bred in 1972 and 1973. Scottish Birds (2000) Breeding birds of the Isle of May, Firth of Forth, 1972-99 11 Black Redstart Pheonicurus ochruros A pair probably bred in 1994 when a male regularly sang during May and 3 recently fledged, but proficiently flying, young accompanied by 2 adults were present from July to October. This would be the first recorded nesting on the island. However, the possibility that this was immigration of a family group after successful breeding on the mainland cannot be ruled out. Such immigration by recently fledged juveniles, with and without adults, has also been recorded for Song Thrush, Blackbird, Wheatear, Starling and Carrion Crow. The appearance of 4 birds including 2 recently fledged young in mid August 1997 was thought less likely than the former record to indicate successful breeding on the island. Wheatear Oenanthe oenanthe One pair bred in 1991 and 2 pairs were present the following year. A male sang but probably did not breed in 1993. Up to 4 pairs bred between 1994 and 1999. Dunnock Prunella modularis Two or 3 pairs bred in 1972 and one pair in 1973. Meadow Pipit Anthus pratensis This species nested in most, if not all, years between 1972 and 1987. Usually there was just a single pair but an additional pair may have bred in up to 6 years and there were 4 pairs in 1980 (Appendix). Rock Pipit Anthus petrosus Few thorough censuses have been made (Appendix) but it is obvious that the population varies considerably. The highest estimate was 40- 50 pairs (1975-76, 1992-93) and the lowest 6-10 pairs (1977-79). Pied Wagtail Motacilla alba The species did not breed in 1972 or 1973. The status in 1974-76 was uncertain but 2 pairs nested in 1977. Numbers increased to 5 pairs in 1982, declined to a single pair in 1986, increased to 7 pairs in 1990-92 and then declined gradually to 2 pairs in 1999. Most adults were M a yarrellii although there were several records of M a alba breeding. Starling Sturnus vulgaris The population was assessed at 10 pairs during the 1970s and guessed at around 30 pairs in 1989, 1990 and 1991. There were 9 pairs in 1992, less than 10 in 1993, 7 in 1994, 3 in 1995 and 1998 and one in 1999. Linnet Carduelis cannabina At least 2 pairs bred in 1973. The species was not recorded breeding for the next 25 years until one pair nested in 1998. Previous and possible breeding species Eggeling reported the following definite or probable breeding species in earlier years: Gannet Morus bassanus present in the 19" century and a pair started to build a nest in 1922, Cormorant Phalacrocorax carbo apair carried nesting material to a cliff in 1938, Mallard Anus platyrhynchos 2 pairs in 1965, Teal Anas crecca one nest in 1960, Peregrine Falco peregrinus bred until 1929 and possibly in 1941, Redshank Tringa totanus bred in 1912, Black Guillemot Cepphus grylle present in the 19" century, Willow Warbler Phylloscopus trochilus in 1922, House Sparrow Passer domesticus breeding until 1947 and Tree Sparrow P montanus breeding until 1922. The occasional Peregrine still displays in the winter and spring and there were afew records of single adult Black Guillemots near the island in several springs, and 12 M P Harris, S Wanless, I Darling, C Gallacher one bird ashore with Puffins, during the 1970s and, early 1980s. In several years, eg 1991-92 and 1996-98, Manx Shearwaters Puffinus puffinus have been heard calling in flight and from burrows in late summer but there was no evidence of nesting. In most recent years Black-headed Gulls Larus ridibundus have shown a passing interest in the tern colony and, since this species breeds on other tern islands such as the Farne Islands and Coquet, this seems a likely candidate to colonize the May. Discussion There have been dramatic changes in the numbers of seabirds on the Isle of May during the last 100 years (Fig 3). First, the very large numbers of terns which bred during the 1940s and 1950s disappeared and only in the last 10 years can the island again be considered to have a thriving ternery. Second, during the 1970s the very large numbers of large gulls then nesting were considered to pose a threat to the fauna and flora of the island and their populations were substantially reduced. Whether or not this contributed to the return of the terns is unclear. Third, the numbers of auks have increased dramatically. The Isle of May now has one of the largest concentrations of Puffins in Britain and auks are now the dominant component of the avifauna. Finally, Eider have increased so that the island now holds nationally important numbers. As long as ground predators such as rats and feral Mink, both of which are abundant at the harbours on the nearby Fife coast, do not find their way to the island the Isle of May should remain a safe haven for breeding seabirds. Nest sites do not appear to be limiting for any species. We speculate that there is potentially room on the island for maybe a quarter of a million pairs of Puffins and double the numbers of species which breed on the cliffs. The landbirds have done less well. In 1959. there were possible 100 pairs of 8 species nesting SB 21(1) Figure 3 The numbers of seabirds breeding on the Isle of May between the 1880s and 1998. Gulls include the Kittiwake. Data 1880-69 from Eggeling (1960, 1974). 70000). a= ae | @ Auks 60000 | | O Gulls | Oferns |; 50000. __NOthers — aN wn 40000 - : = | wo O 30000 = Ny) (>) (=) (S) (ep) eS (ep) oO oO oO 1880 Re 1946 BCL 1959 HE 1969 Ges 1975 1984 1989 1992 1998 (Eggeling 1960); in 1999. there were less than half this number of 4 ‘native’ species, although there were several hundreds of pairs of feral pigeons. During the period covered by our records Stock Dove, Lapwing, Wren and Wheatear colonised the island; all but the latter soon became extinct. Meadow Pipit, Swallow, Blackbird, Dunnock and Song Thrush no longer breed and it remains to be seen whether the long established Starling and Wheatear will persist. Only the Rock Pipit appears to be secure. The island has changed considerably during the time that we have knownit. In 1972. the lighthouse keepers and their families were present and several gardens were cultivated. Although the Sheep and Ferrets had already gone, Dogs were still present, Rabbits and gull eggs were harvested and sent ashore to market. Gulls were nesting over much of the island, except for the top and near the lighthouse buildings, and some areas such as Rona had little vegetation (see photographs in Eggeling 1974). Now the island has no permanent human residents, dogs are not allowed ashore, Scottish Birds (2000) Breeding birds of the Isle of May, Firth of Forth, 1972-99 1B} birds have complete protection, there are terns nesting on the top of the island and there are Puffin burrows virtually everywhere. It is, therefore, not surprising that the land avifauna has changed though the processes involved are far from clear. More unexpected is the lack of breeding corvids and birds of prey. If the Buzzard Buteo buteo continues its spread in Fife it might well consider the Isle of May with its abundant Rabbits to be attractive. Acknowledgements We are extremely grateful to all the many people who, while staying at the Low Light or undertakening wardening duties or their own research, made and documented the observations and counts on which this paper is based. References Aebischer N J 1986. Retrospective investigation of an ecological disaster in the Shag, Phalacrocorax aristotelis: a general method based on long-term marking. Journal of Animal Ecology 55:613-619. Aebischer N J and Wanless S 1992. Relationship between colony size, adult non-breeding and environmental conditions for Shags Phalacrocorax aristotelis on the Isle of May, Scotland. Bird Study 39:43-52. Calladine J, Harris M P, Taylor S and Wanless S. 1995. The status of the Eider on the Isle of May and other Forth Islands. Scottish Birds 18:1-10. Chabrzyk G and Coulson J C 1976. Survival and recruitment in the Herring Gull Larus argentatus. Journal of Animal Ecology 45: 187-203. Duncan N 1981. The Lesser Black-backed Gull on the Isle of May. Scottish Birds 11:180-188. Eggeling W J 1960. The Isle of May. Oliver and Boyd, Edinburgh. Eggeling W J 1974. The birds of the Isle of May - a revised assessment of status. Scottish Birds 8:93-148. Harris M P and Galbraith H 1983. Seabird populations of the Isle of May. Scottish Birds 12:174-180. Harris M P and Harding N 1995. Roseate Terns return to the Isle of May. Scottish Birds 18:11. Harris M P and Wanless S 1996. Differential responses of Guillemot Uria aalge and Shag Phalacrocorax aristotelis to a late winter wreck. Bird Study 43:220- 230. Harris M P and Wanless S 1997. The effect of removing large numbers of gulls Larus spp on an island population of Oystercatchers Haematopus ostralegus: implications for management. Biological Conservation 82:167- JJ Hemsley D 1999 Recolonisation of the Isle of May by Sandwich Terns during 1999. Scottish Birds 20:11. Wanless S 1988. The recolonisation of the Isle of May by Common and Arctic Terns. Scottish Birds 15:1-8. Wanless S, Harris M P, Calladine J and Rothery P 1996. Modelling responses of herring gull and lesser black- backed gull populations to reduction of reproductive output: implications for control measures. Journal of Applied Ecology 33:1420-1432. M P Harris and S Wanless, Institute of Terrestrial Ecology, Hill of Brathens, Banchory, Kincardineshire, AB31 4BY I Darling, Isle of May Bird Observatory, 579 Lanark Road West, Balerno, Edinburgh, EH14 7BL C Gallacher, Scottish Natural Heritage, 46 Crossgate, Cupar, Fife, KY15 5HS Manuscript accepted January 2000 14 M P Harris, S Wanless, I Darling, C Gallacher SB 21(1) Appendix: Counts of pairs of some breeding birds on Isle of May, 1972-1999. nc indicates no count, ? indicates uncertainty whether the species bred Year Shelduck Lapwing Great Stock Swallow Wheatear Meadow Rock Pied Black- Dove Pipit Pipit Wagtail backed Gull 1972 1 0 4 0 3) 0 i ne 0 L973 1 0 1 0 3 0 D ne 0 1974 0 0 0) 0 1 0 ? 43 P NSS 0 0 0 0 l 0 2 45-50 ? 1976 0 2 0 0 pe 0) 2? 40-50 ? 1977 0 6 0 0 i 0 2 6-8 2 1978 | 4 0 0 | 0 | ne 3 1979 3 5 0 1 3 0 l 8 3 1980 2 4 0 | 4 0 3-4 13 3 1981 3 7 0 2 6 0 1 Wi 4 1982 3 5 | 2 4 0 2 25 5 1983 6 4 0 2 3 0 1-2 23 3 1984 9-11 4 0 2 3 0 | 31 3 1985 10 2-3 | | 3 0 1-2 20+ 2 1986 fh | 3 | 3 0 l 25+ 1 1987 10 0 c. 2 3 0 1 20-30 Z 1988 5 0 3 l 3 0 0 31 4 1989 4 0 3 | 5 0 Oy S738 6 1990 i 0 3 0 6 0) 0 ne 7 199] 7 0 4 0 3 | 0 39+ i 1992 3 0 8 0 + 3 0 44 6-7 Nees /; 0 7 0 2 0 0 49 3 1994 8 0 6 0 4 Ps 0 ne 4 1995 6 0 7 0 0 | 0 30 4 1996 6 0 ii 0 0 23 0 25+ 4 Ego 3 0 9 0 0 4-5 0 25+ 5 1998 4 0 14 0) 0 2-3 0 ne 3 1999 3 0 19 0 0 + 0 ne 2 Scottish Birds (2000) 21:15-26 1D) Wintering seafowl in Scapa Flow, Orkney, October 1998 to March 1999 E J WILLIAMS Monthly counts of all seafowl were carried out in Scapa Flow for the 6 months from October 1998 to March 1999, utilising both shore and boat based observations. Thirty species were recorded and the maximum number of individuals of all species recorded in any month was 14,886. Two species, Great Northern Diver and Slavonian Grebe, were present at levels of international importance, and a further 11 species were present at levels of national importance. The maximum of 781 Great Northern Divers recorded was an unprecedented figure for this species in any single locality in the UK and represents 26% of the GB and 15.6% of the European winter population respectively. Slavonian Grebes peaked at 124, representing 31% of the GB and 2.5% of the European winter population. The other species of national importance were Red- throated Diver, 1.2% of GB winter population, Black-throated Diver (8.1%), Red- necked Grebe (14%), Shag (9%), Wigeon (1.4%), Teal (1%), Eider (3%), Long-tailed Duck (6.7%), Goldeneye (1.7%), Red-breasted Merganser (6.2%), and Black Guillemot (2.6%). Surface feeding duck numbered c6,300 and diving duck c4,200. Wigeon was the most abundant species with a maximum of 3,895. Overall numbers of most species exceeded those recorded during previous surveys in 1974-78 and in 1988-89. Introduction Scapa Flow is asea area almost entirely surrounded by the southern islands of the Orkney archipelago and forms one of the largest areas of sheltered seawater in the UK. It covers c230km/”, and is approximately 20kms x 20kms wide with a shoreline of c170kms including the smaller islands within the Flow (Fig 1). It reaches a maximum depth of 60m although the majority of birds are found in the shallower bays where water depth is less than c20m. The combination of along shoreline and sheltered water makes the area attractive for wintering seafowl. Its importance has long been recognised, and the first land based survey in 1974-78 (Lea 1980) quantified this. A repeat survey in 1988-89 (Christer 1989) incorporated additional counts of the central Scapa Flow area which was not visible from shore. The present survey updates the situation. Methods Counts were made from 132 points on the shore each month, covering waters up to cl.S5kms from the shore, while waters in Central Scapa Flow not visible from shore were counted from an 8m boat, also on a monthly basis. Boat transects were run east to west and spaced 2kms apart, the counts being completed ina single day. Days when winds were Force | or less were selected for these counts, when the sea was mirror smooth or nearly so, and birds could be spotted easily. Up to 3 observers were used for this part of the survey. Complete surveys were possible in all months except February, when bad weather prevented counting in the Flotta/Graemsay/South Walls area and no boat count was possible. Counting was not carried out when winds were in excess of Force 4, and as the winter weather was not so severe as 16 EJ Williams SB 21(1) Figure 1 Map of Scapa Flow, Orkney, showing extent of survey area and count points. ' ' -—<—<<- eet 1 ' 1 i} 4 caaaee an aa : Bay | usual, it proved possible to count all sectors on good days at least once during the winter. Results Thirty species were recorded during the 6 months. Many of the winter visitors were arriving during October, joining the locally bred birds already present. Numbers continued to build up through November, and peaked in January. By February, birds were beginning to leave, although the incomplete coverage that month masked the effect. By March, numbers had returned to a similar level ab---~— 2 eee Seer eer Phases ' ' ' ' ' ‘ ' ' ' ‘ ' ' ' ' ‘ ' ' ' 1 : ‘ t ' ' ' ' ' ‘ ' ' ' t ' ' ' ' ' t ' ' t ' ' ' ' ' ' ' ' ‘ ' ' 1 ' ' ' ' ‘ ' ' ' ' ' ' ' ' ' 5 ‘ 7 t poSeSe . -=---- | ------ ip ee oF —---- Knee -_0 ‘ 1 1 1 1 1 1 1 Orphir | ‘ ' ' Re ' 1 ' ‘ ' ' ' ' ' ' ee “ ----- 4 ~-----' L~~---4-----1---~ 7 ft ry ! Waulkmill Bay ' u ' Siimees iia i ~~. --- a5 aTauma se Switha to October. The winter maximum of 14,886 birds was present in January. The figures for November and December were similar, at 14,385 and 13,833 respectively and the average for the 6 months October to March was 11,424 (Table 1). These figures indicate that the birds had settled into their winter quarters by November, and that consistency was being achieved with the counting methods employed. In October, the bulk of the birds were thought to be largely of local origin, with high numbers of Shags and Eider, these 2 species accounting for i PJ QUOIS Wos Sans] UoNLyndod sz9jUIM FH, ‘ae;dutooutL juno Aaenaga, 66 S I He] IM 6) | I 6661 JoA# » Orkney October 1998 to March 1999 17 g seafowl in Scapa Flow nterin I V; bor 8SL8 CIBL ORE ELBET CREF 6988 [PIO] é £1 L L 0) iam IE ia £ Und é FIC L6 CLI £ FIC cl 9S Ol yay apy] LPG OL6 OOS9E L89 OL6 8SE bol cS9 BLL 89S JOW][INE) YOR | i 19] by 19] OF It tl 6 L [[Iq410zP yy é FOV LL@ OEE fe 99 SBE FOr OF JOU] | INE) | 0> £ I | | £ 0 | 0) JEpUBsOOL) c9 8c9 00001 8hr PCL OFC LLY LLY 89 EES JOSUBDIATA, Paisvaiq-pay L'| C8C OOOLI LI L&C OLE C8C 8VC IL C aXauaplon 90 ol OOOE 6 L 8 6 6 fl | IBJON§ JAAA é | 0 0 0 10) 0 0) 0 | INO9OG JANG | (> £ Cc C C (¢ 0 £ I I9}O9§ UOWILUO,) L9 8S I OOSEC LSII 9rCI LOL CBS ter rCCl 689 yond payin-suo7 Oe BOEC OOSLL Tes BEOC 10e1 O89] COLI 80EC 6981 Japly 10> £ | 10) 0) 0 £ 10) 0 Ja[aAOYS 10> c | 0 | 10) (6 0 10) [erate CO LOW QOOOOS CSL PSI £09 LCII O1OI 8EOT cBS PAPTTLIN 0'| cc OO8SE | ORS tcl TOE CCE OLL OEL tL) [POL 10> te | 0) 10) 0 0 IE 0 []UMpeAy v1 SO8L OOBLLE PL9C ILIV 8SLC SORE SILE CORT I8cl WOasI A 0) ls OOSEL OF cL LL (Gu OC Cc 0 YONPLEYS 10> | 0 0 0 0 0 | 0 asoory dviAain 10> bl OSLSC 8 tI ial 8 b 4 S UBMS an} 06 LOEL OOSLE LLO\ beh 89 PSIC OLTEC LOEL S99C ous LO (ao) O0CE I pS ck bE cs OS CO 09 JURIOULIOT) OTE bol OOt 68 FL SOI tol 16 col SE AQesL) UBIUOARTS OVI tC OSI 6 £C S 1c b £ 0) AQalL) payoaul-pay £0 Ol O6CE g Cc V4 8 Ol L 0 aqalyy apy] 09C [8L OOOE BLE T&L Stl 16S bro OFS 99| JOATC] UISYION We) [8 LS OOL tr ch 6b Lt LS oF Co JIATC] PHVOAYI-YOU] ET cl 6S OS8P GE BE 9 60 9 6S OS JIATC] PHWOIY)-Pory vonepndod xwu yuonendod uno IVA HOM une 0d AON po UOT sopadgy JOM PHY, — Apypuou OVUM HO61/866T MOV HdBOS ri i) SpHIOD APYPUO TA h Birds (2000) V ttis Sco MOjy Pdbog Wt uonDpndod FD Jo % =» DUD Mop Ddbos ‘Saipuisa HoNPpndod aM FO [PIOL SeuytE ‘MOjy PADIS UL | MO[PAS AULATIM JO STUNOD APYIMOW B6/R66T 1 GPL, 18 EJ Williams SB 21(1) 51% of the birds present. Red-breasted Mergansers and Black Guillemot accounted for a further 12%. At the end of the month, numbers had been augmented by winter visitors, the bulk of the Great Northern and Black-throated Divers, Slavonian Grebes, surface feeding ducks and Long-tailed Ducks arriving at this ttme. Numbers remained on a plateau during December and January and into early February, after which there was a marked drop in numbers, especially among surface feeding duck. Bad weather at this time prevented complete coverage, and the numbers of Great Northern Divers recorded in particular reflect this, as no count of their main wintering area in the central Flow was possible. By March, overall numbers had returned to a similar level to October, although the proportion of many species present was very different. Great Northern Divers and Long-tailed Ducks were still numerous, with 23% of the March total of seafowl compared to just 10% in October, reflecting the late departure for the breeding grounds of these 2 species. Shag numbers were down to 744 from 2,665 in October, much of this drop being accounted for in presumed mortality of first year birds. Of all the species recorded, Black Guillemot was the most difficult to census accurately due to its small size and overall grey colouration. This species was easily overlooked. The high March figure therefore is thought to represent a more accurate winter figure than those obtained earlier as by this time, birds were returning to their breeding areas, and easier to count. Species of national and international importance Red-throated Diver Gavia stellata Monthly maximum 59 Present throughout the period in small numbers. Numbers were highest in late autumn, with a peak of 59 recorded in November (Fig 2a). This represents |.2% of the GB wintering population. For the rest of the winter, numbers were much fewer, the lowest total being 6 in February, which preceded a gradual build up of birds to 38 in March. Two colour ringed individuals were seen which had been ringed in Shetland in 1988 and 1990. A third individual, metal ringed on the left leg, was almost certainly an Orkney ringed bird as Shetland birds are ringed on the right leg and Orkney birds on the left. Thus, at least some of the Shetland breeding population winter in Orkney, and a few Orkney birds may also remain in Scapa Flow close to their breeding grounds throughout the winter. Black-throated Diver Gavia arctica Monthly maximum 57 Flocks of up to 15 birds were noted at widely varying locations following the first bird on 12 October. Nearly all the early birds were adults, still in breeding plumage. A flock of 15 in Swanbister Bay on 13 October contained only 4 juveniles. After the main arrivals were complete at the end of October, the month on month figures were similar (Fig 2a), varying from 39 to the maximum 57 which represents 8.1% of the GB wintering population. Great Northern Diver Gavia immer Monthly maximum 781 Arguably the most important species in Scapa Flow, the area holding 26% of the GB wintering population, or 15.6% of the European winter population. These birds are almost certainly mostly of Nearctic origin. Small numbers are present in the Flow throughout the summer, these being mostly first year birds. The first returning adults were noted on 19 October, with 8 adults off Fara followed by a flock of 24 adults off Graemsay on the 20", and from then on numbers built rapidly, continuing into mid November, by which time the bulk of the wintering population had arrived (Fig 2a). At this time it was estimated that over 95% of the birds were adults in breeding plumage —it was not until later that first year birds began to arrive. Scottish Birds (2000) Wintering seafowl in Scapa Flow, Orkney October 1998 to March 1999 19 Le aR A ett i eS EE RR Figure 2a Monthly maximum counts of wintering divers, grebes and Cormorant in Scapa Flow, Orkney, 1988-89 and 1998-99. Se OE TINOWIDEC PANT FEB|MAR reseee} 384 50 | of taf 12) a | [| JOCT NOV [DEC [JAN |FEB [MAR | ramps] co ra 0-| 195 ss [1 1988/89 1988/89 | ried eee 1988/89 | | 1998/99 1998/99 400 1998/99 | nctaaene} : } 200 -— are 4 0 H RN 6G S © ov NS) x Na e w & Ro) S = & Ro Great Northem Diver op ee Tee ES | __—|OCT |NOV |DEC |JAN [FEB [MAR | 5 [| [OCT NOV [DEC [JAN |FEB [MAR] 1988/89 Ps vassal ool sa| 2 1988/89 | 150 = = ee 1998/99 | Red-necked Grebe ce) SS SO Si w@ we Slavonian Grebe 1988/89 100 1998/99 50 ~ | 1988/89 | _ 1998/99 | is Oo NS > SS 6 fou RS) x Ne w \s Cormorant Note: counts for October 1988, December 1988, February 1989 and February 1999 were incomplete During November, in calm weather, some large flocks were sighted. On the 3" there were 2 flocks of 25 and 55 birds, and the largest flock recorded during the winter, 66 on 17 November, was off Hunda. The November boat survey revealed 167 birds in the central Flow, with 373 observed in the shore sectors. By December, birds moved further offshore, and during this month there were 391 in the central Flow, compared to 149 in the shore sectors, followed by 282 and 309 respectively in January. Nocomplete count was made in February, and the figures for this month reflect this, but in March, when the highest monthly figure was recorded, there were 438 in the central Flow, and 342 in the shore sectors, giving a grand total and winter maximum of 781. It is evident, therefore, that counts from the shore alone can give a very unrepresentative picture of the total birds present. The numbers recorded are thought thought to be the highest totals for a single area in Europe. During most of the winter, birds were distributed throughout the Flow, and they were recorded in almostevery kilometre square. There were marked concentrations in the first half of the winter in the southern half of the Flow, particularly off Flotta and across to Hunda. Red-necked Grebe Podiceps grisegena Monthly maximum 21 Recorded in every month except October (Fig 2a). The first 3 birds were recorded on 17 November. The maximum for any locality was 13 on6 March, 20 EJ Williams SB 21(1) at Bay of Sandoyne. Otherwise, there was a scattering of records from as far west as Houton Head, along the north and east coast of the Flow as far as Echnaloch Bay. No birds were encountered on the west side of the Flow. The maximum figure represents 14% of the GB wintering population. Slavonian Grebe Podiceps auritus Monthly maximum 124 Although present in much lower numbers than Great Northern Divers, Slavonian Grebes are the most important species present in the Flow in terms of national importance, holding 31% of the UK population. Already present in the Flow by 1 October, when 8 were noted at Echnaloch Bay, the bulk of the wintering population had arrived by the beginning of November (Fig 2a). The localities holding most birds were the traditional sites of Echnaloch Bay/Bay of Sandoyne area in the east of the Flow and Swanbister Bay/Waulkmill Bay. The peak for any one locality was 38 in Echnaloch Bay on 31 January. Swanbister Bay generally held between 20 and 30 birds. The maximum day count covering Echnaloch — St. Mary’s Bay, but excluding Bay of Sandoyne, was 50 on 31 January. Shag Phalocrocorax aristotelis Monthly maximum 3,393 9.0% of the GB wintering population was recorded in Scapa Flow The highest numbers occurred in the early part of the winter (Fig 2b), with peak counts of 1,011 birds in Burra Sound on 20 October and 929 in Switha Sound on 18 November. In the early part of the winter, at some localities up to 90% were first year birds, this proportion dropping rapidly as winter progressed. By the end of January into February, overall totals dropped drastically as birds departed for their breeding grounds, leaving behind what was presumed to be the local breeding population. Wigeon Anas penelope Monthly maximum 3,895 Numbers built rapidly through October, and reached a maximum 3,895 in January representing 1.4% of the GB wintering population (Fig 2b). A sharp decline in numbers began in mid February, continuing into March, as birds left their winter quarters. Five areas held almost 90% of the wintering birds with the Water Sound area from Hunda and Burray to South Ronaldsay holding a maximum of 1,227 in January. Teal Anas crecca Monthly maximum 1,322 Small numbers present in October built up through the following months to peak at 1322 in January, 1.0% of the GB wintering population (Fig 2b). As with Wigeon, there was a rapid drop in numbers after mid February. Hunda held the largest single gathering, with 322 in November. Several other localities supported up to 100 birds. Eider Somateria mollissima Monthly maximum 2,308 Present throughout the period with 3.0% of the GB wintering population present. Numbers built from the October figure of 1,869 to a peak in November of 2,308 (Fig 2b). Thereafter, numbers declined gradually to a February low of 1,301, then in March there was arise to 2,038. The prime area for this species 1s the northwest corner of the Flow, where around 600-700 birds were located. This includes the area off the south shore of Graemsay, the Bay of Ireland, and the coastline southeast to Houton Head. Off Graemsay, a maximum count of 466 was made in November. Birds were still numerous here until March with around 300 birds present while the Orphir coast from Clestrain to Petertown held a maximum of 200. Scottish Birds (2000) Wintering seafowl in Scapa Flow, Orkney October 1998 to March 1999 2] Figure 2b Monthly maximum counts of wintering Shag, Scapa Flow, Orkney, 1988-99. a OCTINOW IDES NAN IFES MAR 1988/89 |1942] 2802] 778] 951] 792| 610) | _|OCT |NOV |DEC [JAN [FEB [MAR | 1988/89 ee ae surface feeding duck and Eider in | |OCT |NOV [DEC [JAN _|FEB |MAR | 1988/89 1988/89 | ‘missais9 misses 1988/89 | wis9s/s9 @is9s/S9 aS a NOM IDEC JAN _IFER IMAR SS eas CREINOY [DEC AN FEB [MARY SS ee ERG] cares ev es 70] 7 1988/89 wisesiss ‘migasise 1998/99 1989/89 _ misgass° Mallard Note: counts for October 1988, December 1 988, F ebruary 1989 and F ebruary 1999 were incomplete Long-tailed Duck Clangula hyemalis Monthly maximum 1,582 Birds arrived from early October with the first record from Scapa Bay on 12", and then built rapidly during November, peaking at 1,582 in January, 6.7% of the GB winter population (Fig 2c). The single most important area for this species is off the south side of Graemsay, where the January count of 610 was the maximum for any locality. Several sectors held up to 100 birds, in particular, Gutter Sound, Scapa Bay, Holm Sound east of No | and 2 Barriers, east of No 4 Barrier, Echnaloch Bay and Bay of Ireland. They were almost absent from the central Scapa Flow area. Goldeneye Bucephala clangula Monthly maximum 282 The first bird was not recorded until 16 October, and the species was still uncommon at the end of November. Numbers built rapidly then to a maximum of 282 in January, 1.7% of the GB wintering population, and remained at much the same level till the end of the period (Fig 2c). Highest numbers were found in Widewall Bay, peaking at 42 in January. The Bay of Ireland area also held over 40 birds but they were spread over a much wider area. North Bay (Hoy) maintained a population of around 20 birds as did St. Mary’s Bay and Weddell Sound (Burray). Red-breasted Merganser Mergus serrator Monthly maximum 628 Numbers peaked at 628 in November, 6.2% of the GB wintering population (Fig 2c). Bay of Ireland held over 40 birds throughout the period, with a peak of 96 in December. Echnaloch Bay and Water Sound between them held about 210 birds 22 EJ Williams SB 21(1) in November, but numbers dropped rapidly later. The bulk of these were moulting birds which traditionally gather at these 2 locations. Pegal and Lyrawa Bays held 94 birds between them in October, but again numbers dropped away later. On occasion, large numbers were seen off the south side of Gramesay feeding with the Eider and Long-tailed Ducks. Otherwise, birds could generally be found in ones and twos around the bulk of the coastline, with larger numbers in the shallow bays. Black Guillemot Cepphus grylle Monthly maximum 970 The maximum of 970, 2.65% of the GB wintering population, occurred in March (Fig 2c) The count in the central Flow area for this month was also the maximum, at 238, around 25% of the birds present. Apart from the high counts in the central Flow, numbers were regularly seen in Bay of Ireland, Scapa Bay, St. Mary’s Bay, Outer Water Sound, and off Houton Head with up to 64 being recorded. Otherwise smaller numbers were distributed widely. Many birds roost communally on marker buoys scattered throughout Scapa Flow, particularly in early to mid winter, withamaximum of 120 at the Mill Bay buoy, Hoy. Other species Cormorant Phalocrocorax carbo Monthly maximum 92 A build up occurred from October into November peaking at 92, with a steady decline thereafter, as Figure 2c Monthly maximum counts of wintering diving ducks and auks in Scapa Flow, Orkney, 1988-89 and 1998-99. ee eee ee ee ee epee Ss a 1988/89 | 1000 L rm ™ 1998/99 al | oe S & eS & a Long-tailed Duck Le OC TINGW BEE WANE IEEE MAR! 1988/89 1988/89 1998/99 | 1988/89 400 + mi998/99, ees eZ | | om. 1200 ;——_—__—- 1988/89 | 1998/99 | | aggre 1998/99 | | >\ SRS OD KS A fo to) OES | SOK Fe ¥ Ses Se | vw ¥ Black Guillemot Sipsee J | BS BO cS x Na x“ w Little Auk Note: counts for October 1988, December 1988, February 1989 and February 1999 were incomplete. Scottish Birds (2000) Wintering seafowl in Scapa Flow, Orkney October 1998 to March 1999 23 birds began to come into breeding plumage (Fig 2a). Shelduck Tadorna tadorna Monthly maximum 77 Small numbers in October and November began building in December as birds returned from their annual moult (Fig 2b). The peak of 77 occurred in February. Mallard Anas platyrhynchos Monthly maximum 1,127 Numbers built up through October, and reached a plateau from November to January, following which there was a sharp decline in numbers from mid February continuing into March (Fig 2b). The winter peak in January of 1127 was only slightly above the figures for November and December. By far the largest concentrations were in the Bay of Ireland area, where the maximum count was 326. Widewall Bay with 117in November was the next highest, while several bays held up to 100 birds. Velvet Scoter Melanitta fusca Monthly maximum 19 Small numbers of this species built up steadily from October to the peak of 19 in January, declining thereafter(Fig 2c). Little Auk Alle alle Monthly maximum 214 Recorded regularly from November onwards with amaximium of 214 in January (Fig 2c). While not forming an important proportion of the winter population occurring in British waters, it was nevertheless present in numbers not normally found regularly so close inshore. The first birds were found in the central Flow on 31 Oct. Thereafter it was seen regularly from various shore locations, especially off the north coast of Flotta where flocks of up to 25 were not unusual. However, the bulk of the birds were recorded from the boat transects. All birds appeared to be in good health, and as numbers were maintained through the winter, it must be assumed that the birds were not storm driven casualties, but found Scapa Flow a suitable wintering area. Maxima of 10 Little Grebe Tachybaptus ruficollis, 14 Mute Swan Cygnus olor, 1 Greylag Goose Anser anser, 3 Gadwall Anas strepera, 2 Pintail Anas acuta, 3 Shoveler Anas clypeata, 3 Common Scoter Melanitta nigra, 1 Surf Scoter Melanitta perspicillata, 3 Goosander Mergus merganser, 404 Guillemot Uria aalge, 161 Razorbill Alca torda, and 13 Puffin Fratercula arctica were also recorded during the survey. Discussion The peak number of seafowl in the Flow was 14,886 in January 1999, and the average for the 6 months October to March was 11,424. These numbers are around twice those recorded during the 1988-89 survey (Table 2). A similar distribution of birds was found in all 3 surveys, with most species being concentrated in the shallower inshore waters, but with Great Northern Divers distributed over the full expanse of the Flow. The survey in 1974-78 covered only selected areas of Scapa Flow and included counts from the adjacent Loch of Stenness. The loch was excluded from both subsequent surveys. The figures obtained from the 1974-78 survey are therefore not directly comparable with those of the later 2, and are included for information. All but 2 species, Cormorant and Velvet Scoter, were more plentiful in 1998-99 than 1988-89, some markedly so. The decline in Velvet Scoter numbers continues a trend evident since the 1974-78 survey and this Species is now no longer of national importance. Great Northern Divers have seen a more than threefold increase in the past 10 years, having been apparently static since 1974-78. The October 24 EJ Williams SB 21(1) Table 2 Maximum counts of wintering seafowl in Scapa Flow, Orkney, in 1998-99 and 1988- 89 and estimates for 1974-78. Species Max Max Max 1974-78 1998-99 1988-89 (estimate) Christer 1989 Lea 1980 Red-throated Diver 59 50 Black-throated Diver ay 22) Great Northern Diver 781 2 200 Little Grebe 10 5 Red-necked Grebe D3 5 Slavonian Grebe 124 28 60 Cormorant 92 123 Shag 3393 2802 4000 Mute Swan 14 Greylag Goose l Shelduck TY 63 Wigeon 3895 1356 Gadwall g) Teal L322 337 Mallard Le) 579 Pintail 2 2 Shoveler 3 Eider 2308 893 2000 Long-tailed Duck 1582 1005 2400 Common Scoter 3 l Surf Scoter | Velvet Scoter 19 4] 100 Goldeneye 282 118 270 Red-breasted Merganser 628 424 350 Goosander 3 2 Guillemot 404 188 Razorbill 161 124 Black Guillemot 970 410 1000 Little Auk 214 220 Puffin 13 10 Total 14886 7973 Scottish Birds (2000) Wintering seafowl in Scapa Flow, Orkney October 1998 to March 1999 25 1998 figure alone approached the maximum recorded in either of the previous surveys. For many of the species of major importance the figures for 1998-99 are much closer to the estimates in 1974-78 than those of 1988-89. The estimate for Black Guillemot in 1974-78 was c1000, which is very close to the present survey's maximum of 970, whereas the 1988-89 figure was less than half of this. A similar situation exists for Eider and Goldeneye. All3 abundant surface feeding ducks, Wigeon, Teal and Mallard, were present in significantly higher levels in 1998-99 than 1988- 89, and the winter peak for all these species was c6,300. In the past 10 years there has been a threefold increase in Wigeon numbers. Teal and Mallard show between two and threefold increases. During the present survey, the 2 most abundant diving ducks were Eider and Long-tailed Duck, which averaged cl,925 and c1,400 birds respectively through the mid winter period. Along with c550 Red-breasted Merganser and c250 Goldeneye, the total of diving ducks in the Flow was c4,200 birds. This compares with c2,500 in 1988-89, and c4,700 in 1974-78. The present figures thus come close to those of the earliest survey. However, the proportion of each species is different. In particular, Long-tailed Duck, an Arctic breeding species, while more abundant in 1998- 99 than 1988-89, has not returned to the 1974-78 levels, and is presently at about 62% of the population at that time. Eider on the other hand, which are thought to be largely of local origin, are now at a very similar level to 1974-78, having been very much scarcer in 1988-89, dropping at that time by 60% from the 1974-78 figure. During the last few years, Long-tailed Ducks have become less numerous over much of their GB range, and the Scapa Flow numbers are likely to be areflection of this national trend. Ata local level, the population trends between the 1988-89 and 1998-99 surveys of various species do however reveal some similarities. Thus, in both surveys Red-throated Divers were commonest in October and November, with minimal numbers remaining in February, before the return of local breeding birds boosted the figures in March. The peak month for Shag in both surveys was November, with a steady decline afterwards, and a similar March total for both. Cormorants exhibited a very similar trend to Shag over the 2 periods, although numbers were slightly lower in the 1998-99 survey. Red-breasted Mergansers had an early peak, with high numbers in October peaking in November, then dropping during the final part of the winter. Anumber of species did not exhibit this similarity. Black-throated Divers had been fairly uncommon in 1988-89, but they were encountered frequently in 1998-99, with numbers peaking in December, and then tailing off. Another species that has increased significantly in the intervening 10 years is Red-necked Grebe, with totals peaking at 23 birds, compared to 5 in 1988-89. The maximum of 220 Little Auks recorded in 1988-89 was a very similar figure to that of the present survey . The bulk of the birds were found in the central Flow area and it would appear that Scapa Flow is a regular wintering area for this species, supporting good numbers of this species in an inshore area. Conclusions There is no doubt that Scapa Flow continues to be of national importance as a wintering area for seafowl. Almost all species are at the highest levels ever recorded and birds were found widely distributed over the whole of the Flow. The Great Northern Diver population in Scapa Flow has been recognised as of considerable importance for a long time. The present hugely increased winter numbers, peaking at 78 1 in March, make them of prime conservation concern, with 26% of the GB and 15.6% of the European 26 EJ Williams SB 21(1) population present in the area. With Slavonian Grebe, the other species present in internationally important numbers, and a further 11 species of national importance, Scapa Flow rates as one of the most important areas for wintering seafow] in Great Britain. Acknowledgements This survey was undertaken on a contract jointly funded by Elf Enterprise, Orkney Islands Council Dept of Harbours, Scottish Natural Heritage and RSPB, and administered by RSPB. I would particularly like to thank the help received from B Ribbands, S J Williams, C J Booth, K Fairclough and A Skene who accompanied me on the boat transects. It would not have been possible to carry out this part of the survey without their assistance. T Todd made his boat available at the inevitably short notice that I was able to give him. T Dean and staff at SNH gave assistance with preparation and administration. ER Meek, RSPB Orkney Officer, supervised the work. Comments on drafts were made by E R Meek, I Bainbridge and S RD da Prato. Iam grateful for their help and comments. References Christer W G 1989 Winter Concentrations of Seaducks, Divers, Grebes and Auks in Scapa Flow, Orkney 1988- 59. RSPB Research Report. Cramp S & Simmons K 1977. Birds of the Western Palaearctic.Oxford University Press. Lea D 1980. Seafowl in Scapa Flow, Orkney 1974-78. RSPB report to NCC. Richardson M G, Heubeck M, Lyster I, & McGowan R 1979. Great Northern Divers in Shetland. Joint report of NCC, SOTEAG, Royal Scottish Museum, Edinburgh. Stone B H, Sears J, Cranswick P A, Gregory R D, Gibbons D W, Rehrisch M M, Aebischer N J, & Reid J B 1997. Population estimates of birds 1n Britain and in the United Kingdom. British Birds 90:1-22. Waters R J et al 1998. The Wetland Bird Survey 1996- 98 BTO, Wildfowl & Wetlands Trust, RSPB, JNCC Williams E J 1999. Wintering Seafowl in Scapa Flow 1998/99. Unpublished report to SNH, RSPB, Elf Enterprise, Orkney Islands Council, Harbours Dept. E J Williams, Fairholm, Finstown, Orkney. KW17 2EQ Revised manuscript accepted January 2000. Great Northern Divers in winter plumage - . Oe aeat%e . . : oes . P.G.1. David Mitchell Scottish Birds (2000) 21:27-35 Zi, Greylag Geese breeding in Shetland M G PENNINGTON Greylag Geese from the Icelandic breeding population have migrated through Shetland for many years. As the Icelandic population has increased, Greylags have become more common in Shetland as migrants and wintering birds, while more recently an increase in reseeded hillsides in the islands has apparently tempted some to remain to breed. Unst was colonised in 1985, possibly earlier, and birds have also been seen in suitable breeding habitat on Mainland since the late 1980s. The current breeding population in 1998-99 is estimated at between 80 to 100 pairs. About half are on Unst with most of the others on Mainland. The Shetland population is of particular interest as it forms a third distinct British breeding population, presumably of Icelandic origin, in addition to the native Scottish population and feral birds. Introduction The Greylag Goose Anser anseris the only species of goose that breeds naturally in temperate areas of Europe, although its distribution has been greatly influenced by man, especially in recent years (Harrison 1982). Formerly, the species was widespread as a native breeding bird in Britain, south to the East Anglian fens, but by the beginning of the twentieth century native stock were restricted to north west Scotland (Owen ef al 1986, Thom 1986). Despite a continuous breeding presence in Scotland, there is no evidence that Greylag Geese bred in Shetland in the past. George Low (1879), who visited Shetland in the summer of 1774, does not mention the species, while ‘wild geese’ are only mentioned as visitors to Shetland in the Old Statistical Account of Scotland, published in 1791- 99. Dunn (1837) states that “considerable numbers of this bird visit ... but do not breed there.” However, Greylag Geese appear to have been scarce in Shetland at the end of the nineteenth century as neither Saxby (1874) nor Evans and Buckley (1899) could give many dated records, although they were told that the bird was common on migration. Venables and Venables (1955) also found Greylags to be scarce in the middle of the twentieth century. All their records were between October and May and their largest count was a flock of 37 at the Loch of Spiggie in October. Bobby Tulloch recorded a small flock at Spiggie in the mid 1960s, this being the first regular wintering flock for several decades. The Greylag Goose is now a regular migrant in Shetland, with flocks of several hundred frequent in autumn. There is also an increasing wintering population, centred in South Mainland around Spiggie where up to 350 have overwintered in recent years. Smaller numbers, usually less than 30, winter elsewhere, especially around Sullom Voe and in the north isles. Dymond (1991) also recorded a post war increase in records on Fair Isle, where it is still principally an autumn migrant. Summering Greylag Geese were recorded with increasing frequency in Shetland during the 1970s, especially in Unst, where there have been almost annual summer records since 1973 according to the Shetland Bird Report. In retrospect, it can be seen that these birds were the vanguard for the species’ colonisation of the islands. 28 M G Pennington SB 21(1) SS aS In the most recent review of Western Palearctic geese populations Mitchell (1999) stated “Greylag Geese also breed in Shetland and Orkney although there is little evidence that these birds form part of the indigenous [ie Scottish] population.” This paper is intended to present a review of breeding records in Shetland as well as presenting the evidence for the origins of the population, which is presumed to be Iceland. Breeding in Unst Colonisation and population Breeding was first confirmed on Unst in 1985 when Mike Peacock discovered a pair with goslings during moorland bird survey work. However, breeding almost certainly took place before this. The late Ian Spence of Uyeasound, Unst recalled a group of 4 geese on the island of Haaf Gruney on 7 June 1981 that were behaving suspiciously in the light of later observations. He also had a brood of goslings reported to him in 1983. This report was never confirmed but 4 birds summered on Unst that year and, rather suspiciously, the number increased to 8 in August (Shetland Bird Report 1983). Breeding numbers on Unst have been monitored principally from counts of the number of broods sighted (Fig 1). In the first 4 years after breeding was first confirmed, only one to 3 pairs were known to breed, but since then the population has slowly increased. In 1998, most of the suitable breeding areas on the island were visited at least once in April or May and this confirmed that most pairs were breeding close to areas where broods were being sighted. However, one group of about half a dozen pairs was found in area where it was unlikely that broods would be located, and it is suspected this group has not been included in previous estimates of the Unst population. From counts of broods and visits to breeding areas the Unst breeding population was estimated at about 30 pairs in 1998. In 1999, however, a record 27 broods were sighted, including the first confirmation of breeding in the southwest and in the north of the island. Allowing for pairs that failed with eggs and birds in unvisited areas, the breeding population on Unst in 1999 was probably at least 40 pairs. Figure 1 Number of broods of Greylag Geese sighted on Unst, 1985-1999. 85 87 89 91 93,95 eec7aeod Nest sites All nest sites found so far have been in reasonably tall Heather Calluna vulgaris, up to 30-40 cm high, usually on permeable bedrock in an area of heathland between Uyeasound and Baltasound. Occasionally, nests are adjacent to small pools or lochans, but others are up to a 0.5 km from the nearest body of freshwater. All nests have been in fairly prominent positions, on a slope or knoll with a good view of the surrounding land in most directions. Other pairs which, judging by their behaviour, probably had active nests have been encountered in similar sites. However, a few pairs are now breeding in the north of the island, either in lowland marsh or heather moorland on impermeable bedrock. Eggs are laid in late April or early May judging by sightings of the first broods; young goslings are usually first sighted Scottish Birds (2000) Greylag Geese breeding in Shetland 29 between 21-25 May. Not surprisingly, this puts the timing of the breeding season in Shetland between the dates for England and Iceland given by Cramp and Simmons (1977). Brood sites Once hatched, broods are usually taken to reseeds on the edge of the heath. One particular reseed is especially favoured, and was the site for most of the initial sightings of broods despite being 2-3 km from some nest sites. Itis situated on serpentine bedrock, with the native vegetation reestablishing itself in several areas. The surrounding heathland is used as a refuge in case of disturbance; water is rarely used and the site has only a few shallow pools. In recent years, other reseeded areas in the vicinity of the heathland breeding areas have been used with increasing frequency. Brood sites are in use from the hatching of goslings in late May through to late June or early July. Although goslings have not fledged at this time, adults have already begun to moult and many will be flightless. Moult sites Moulting birds are very cautious and even harder to locate than broods. Although there are several large freshwater lochs on Unst that would appear to be suitable moult sites, these have been searched frequently in July without success. In July 1997, a flock of up to 100 adults and goslings was discovered on the sea, east of the main breeding area. These birds were also present in July 1998 and 1999, using a small island and the adjoining coast for feeding and resting, but swimming out to sea when disturbed, to a distance sometimes in excess of akilometre from land (pers obs and D & J Sandison pers comm). Occasional sightings of small flocks on the Unst hills, and the regular location of fresh droppings on reseeds in July, suggests that the moulting geese come further inland to feed on occasion. Non breeding status After adults regain the power of flight in late July or early August, geese wander around Unst and they use a wider variety of habitats than usual, with bogs and mires often favoured. Migrants may join the local flocks in autumn and this confuses subsequent movements of local birds. In midwinter geese are usually found on reseeds. It is rare to find more than about 50 birds on Unst in midwinter and in most years there is just a small flock of about 20, although about 70 wintered in 1999- 2000. It is not clear whether the birds that remain on Unst are locally bred, but as the numbers present are at their lowest at this time of the year, then at least some birds must leave the island in winter. It is possible that birds disperse elsewhere in Shetland, especially to Fetlar where small flocks overwinter regularly, but they could be moving further afield. In spring, in late March or early April, birds return to Unst, often gathering in one or 2 large flocks on reseeds around Uyeasound and Baltasound. These flocks usually consist of obvious pairs, which feed on fresh grass before moving on to territory on the heaths. In summer a single, discrete non breeding flock of about 15-25 birds was formerly present, but in recent years the number of non breeding birds has increased, with several small flocks present, often close to breeding areas. Breeding elsewhere in Shetland Foula The first probable breeding record in Shetland was on Foula in 1970. Two adults were seen throughout the summer, and 2 adults with 3 flying young were recorded on 24 August (Shetland Bird Rreport 1970). The significance was apparently 30 MG Pennington SB 21(1) -———oC_—eesSsSsS$9sSsSsSSsSsSsSSSS SS SS -SS not realised at the time, but in retrospect this would appear to suggest a successful breeding attempt. Single birds summered on Foula in 1971 and 1972 and there were 5 in July 1975, but there have been no further indications of breeding on the island. Fetlar Itis rather surprising that geese have taken so long to colonise Fetlar given the similarity in habitat with the breeding areas in Unst. Small numbers have wintered regularly since the late 1980s, with up to 50 in the early 1990s while birds summered for the first time in 1995, with 2 or 3 in that year increasing to 14 in 1997 (D Suddaby and B Thomason pers comm). Breeding was first confirmed in 1998 when a nest was found in heather on Vord Hill and 2 broods were seen in this area later in the summer (M Smith pers comm). Birds summered in 1999 but breeding was not confirmed. Yell Occasional pairs have been seen in late spring since the late 1980s but although they do not usually summer there have been a few breeding attempts. In June 1996 a brood of 4 goslings was seen on a small reseed at Colvister (K Osborn pers comm). In 1998 a pair appeared to be attempting to nest beside the Loch of Littlester, Burravoe but they departed in May (M Odie pers comm). A creche of goslings, comprising one or 2 broods, was found on hills northwest of Otterswick in July 1999 (R Riddington pers comm). North Mainland Pairs on territory have been seen regularly in spring since 1989 but no attempt has been made to survey this area. There is alarge amount of suitable breeding habitat: heather moorland dotted with small lochans and interspersed with reseeded hillsides, often far from metalled roads. Figure 2. Location of Greylag Goose breeding areas in Shetland used since 1990. The main breeding area on Unst is shown by a large closed circle. Areas with smaller populations are shown by small closed circles. Open circles denote areas with summering birds but no confirmation of breeding or single breeding records. The most accessible breeding site is around Yamna Field and Eela Water where birds have been present since at least the early 1990s (P Ellis pers comm). In June 1997, a crofter reported a creche of goslings on reseeds at Yamna Field, with | Scottish Birds (2000) goslings seen in the area for at least 3 years previously (J Swale pers comm). Post breeding flocks of up to 30 birds have been seen on Eela Water in late summer (P Ellis pers comm), suggesting a minimum of 5 pairs in the area. An even more remote breeding area is on the plateau north of between Ronas Hill with birds seen on territory here in several areas since 1989. Although broods are only seldom reported from this remote area there are several reports from late summer. A moulting flock of 30 was found on a loch in the area in 1999 (D Okill pers comm). Post breeding flocks of about 40 birds were found on Collafirth Hill in late summer 1997 and 1998 (T Rogers pers comm), while there were 63 at Ollaberry on 10th August 1999. In 1997 anest and at least one brood were found at Mangaster (H Towll pers comm) and a few pairs have been seen there in each subsequent summer. Although rarely visited, large amounts of goose droppings are often seen around lochs at Tingon in summer (D Okill pers comm), suggesting birds may breed there as well. With breeding taking place in 3 or 4 areas in North Mainland, the total breeding population is probably at least 15 pairs, but given the large amounts of suitable habitat and the post breeding counts from North Roe, it could easily be double this figure. West Mainland Pairs have been seen on territory since the early 1990s, mainly around Moor Field at the head of Brindister Voe and in the vicinity of West Burrafirth (P Ellis pers comm). This is another area with a large amount of suitable habitat, similar to areas used in North Mainland, with a mixture of heather moorland, small lochans and reseeded hillsides. Although, no census has been attempted and breeding is seldom confirmed, regular sightings of pairs in spring suggest that small numbers breed regularly, possibly as many as 10 pairs. Greylag Geese breeding in Shetland Syl Central Mainland Pairs on territory have been seen since the early 1990s on the Heather hills north and east of Scalloway and, although no census has been attempted, the area is relatively accessible and broods or nests have been seen in most years (P Ellis, D Okill and K Osborn pers comm). From the various sites known to be used it is likely that a small number breed here regularly, possibly as many as 10 pairs. South Mainland A pair summered on the island of South Havra in 1989 (P Ellis and D Okill pers comm). There are no indications that geese have used this island since. In 1992, a pair with 2 goslings was seen at the Loch of Clumlie. The following year 2 broods were seen in this area and breeding has been confirmed in each subsequent year. In 1998 nests were found in 3 areas around Clumlie, in damp rough grassland at Scousburgh (per P Harvey) and between Virdi Field and the loch (A Fitchett pers comm), and in heather on Virdi Field (M Mouatt per P Harvey). In 2 of these localities the observers had found nests for about the last 3 years. Broods were also seen to the north of this area at Levenwick in 1998 and 1999 (N Dymond and P Harvey pers comm). Although birds are surprisingly elusive in this area, figures over recent years suggest a current total of at least 10 breeding pairs. Total Shetland population The total number of pairs of Greylag Geese estimated to breed in Shetland is given in Table 1. The late summer population has been estimated by multiplying the number of pairs by a factor of 5. This is assuming both of the pair are still alive in late summer, each pair raise 2 goslings to fledging and the non breeding population is equivalent to the number of pairs. This last 32 MG Pennington SB 21(1) assumption 1s based on observations on the size of the non breeding flock on Unst, but is likely to be a slight overestimate as the population grows. However, Thom (1986) estimated the number of pairs in the native Scottish population from the late summer population using similar calculations. The figures broadly agree with observations of flocks of birds in Shetland in late summer. The late summer population is estimated at about 4-500 birds and the breeding population at 80-100 pairs. Table 1 Estimated number of breeding pairs and estimated late summer populations of Greylag Geese in Shetland in 1998-99. Area Estimated Estimated number of late summer pairs population Unst 40 200 Fetlar 1-2 5-10 Yell 1-2 5-10 North Mainland 15-30 75-150 West Mainland 5-10 25-50 Central Mainland 5-10 25-50 South Mainland 10 50 Total 77-104 385-520 Discussion Origins of the Shetland population _ Four separate populations of Greylag Geese occur in Britain - a small native breeding population, a larger feral breeding population, wintering Icelandic birds and Scandinavian breeding birds, the last population wintering in Iberia and only occurring occasionally (BOU 1971). The native British breeding population has bred only in Northern Scotland and the Hebrides since the turn of the century (Sharrock 1976). The population is essentially sedentary with most birds wintering within 100km of their breeding areas (Mitchell 1999). A considerable increase in Scottish breeding birds has occurred recently, with just 500-700 pairs in the early 1980s (Owen et al 1986) but most recent estimates put the population at about 9000 birds (Mitchell 1999). Although Scottish birds may have spread as far as Orkney, as discussed later, there is no evidence of Scottish breeding birds moving to Shetland. A population of feral Greylag Geese has colonised many lowland areas of Britain over the last few decades (Delaney 1993). These introduced birds, many of which were obtained from the Hebrides, are largely sedentary, undertaking mainly short local movements. There is no evidence that Greylag Geese breeding in Shetland originate from anything other than wild sources despite comments in recent Shetland Bird Reports, although a male Greylag Goose bred with a domestic goose at Benston Loch, Nesting, in Mainland in 1997. There have been no recorded releases of Greylag Geese in Shetland and the closest established feral flocks are in Sutherland (Thom 1986, Brown and Dick 1992), although there have been releases in Orkney. Icelandic Greylag Geese winter entirely in Britain and mainly within Scotland (Cramp and Simmons 1977). This population is monitored by counts of wintering birds in Britain with an almost fourfold increase between 1960 and 1984 to a total in excess of 100,000 birds (Owen 1986). Numbers have declined since, with the population in 1997 standing at just under 80,000 birds (Cranswick et al 1999). There is little doubt that migrant and wintering Greylag Geese in Shetland originate from Iceland. A few ringing recoveries help confirm this. An Icelandic bird ringed in 1988 was found on an oilrig in the Brent Oilfield in April 1993, while another ringed in Iceland in 1997 was Scottish Birds (2000) shot in Shetland in November of the same year. A bird ringed on Fair Isle in October 1979 was recovered in Iceland in April 1980 and several others ringed on passage on the island have been recovered in winter in Mainland Scotland. Several birds colour ringed on wintering grounds in Scotland have been seen in the Sumburgh area in early spring since 1995, presumably on their way back to Iceland. The recent increase in reseeded hillsides in Shetland has probably encouraged Icelandic wintering birds to remain and breed. The amount of grass, excluding rough grazing, in Shetland increased from 7000 ha in 1981 to almost 20,000 ha in 1996 (Shetland Islands Council 1997). Rough grazing decreased by almost 7000ha during the same period and most of this lost rough grazing will have been reseeded. The sudden appearance of lush grass on the edge of moorland will have been appealing to Greylag Geese, creating the mosaic of grassland, moorland and freshwater which appears to be their favoured breeding habitat in Shetland. Moorland nests in heather are the preferred sites amongst the Shetland population, as is the case in the Outer Hebrides (Paterson et al 1990), although this habit is largely restricted to Scotland, Icelandic birds usually nesting in rank vegetation (Cramp and Simmons 1977). Comparison with Orkney The following account is based on information supplied by Eric Meek. Greylag Geese started breeding in Orkney in the 1980s. Releases of feral birds on Shapinsay, beginning in 1984, have confused the situation, although birds of presumed wild origin started nesting at about the same time. These wild birds may have come from the Scottish breeding population, as marked birds were seen in Orkney in the winter of 1996-97. However, it is suspected that some of the large population of wintering Icelandic birds may have remained to breed as well. It is estimated that between 50-100 Greylag Geese breeding in Shetland 3} pairs currently nest in Orkney, with the true figure rapidly heading towards the upper end of this range. Most Orcadian birds breed on small islands, especially in Wide Firth near Shapinsay, although some nest around large lochs on Orkney Mainland and others, believed to be birds from wild populations, nest on moorland. Comparison with Faroe The following account is based on information supplied by Dorete Bloch of the Faroese Natural History Museum and Jens-Kjeld Jensen. Greylags formerly bred in Faroe in relatively large numbers until the middle of the nineteenth century, their extirpation being largely due to hunting, particularly by Danish officials. There were only a few sporadic breeding records for most of the twentieth century (Williamson 1970). However, a feral population was established in the capital Torshavn in the 1960s, originating from 2 injured wild birds released by a local resident and encouraged to breed. In 1976 there were 15-20 pairs in this area. Breeding by Icelandic migrants was not regular until the 1970s, and by 1981 there were still fewer than 10 wild pairs. Surveys of Faroese birds in 1987 and 1989 located just 18-24 pairs, presumably mainly feral birds. Since then the population has increased, and in 1993 there were 61-78 pairs, breeding in about a dozen areas, and a post breeding population of about 500 birds (Olsen 1994). The population has increased slightly since to an estimated 80-100 pairs. Faroese birds nest in a variety of habitats, including mountains and islets, but the semi tame Torshavn population, which now accounts for about a third of the total, nests beside lakes in the town. Wild and tame birds mix, partly because birds from Torshavn have been translocated to reduce theirimpact on gardens. Some birds remain in the islands to winter. About 200 were estimated to be present in 1993 (Olsen 1994) but others emigrate and there are single recoveries of Faroese ringed birds in Scotland and Denmark. 34 MG Pennington SB 21(1) The evidence from Faroe, with a rapid increase in the wild breeding population at the same time as the colonisation of Shetland, supports the view that the Shetland population originates from Icelandic migrants. Conservation implications With an expanding native Scottish population, the c400 birds in Shetland do not yet forma substantial part of the native stock. The numbers also pale into insignificance compared with a total British population of over 30,000 birds including the feral population, which numbered almost 20,000 in a census in summer 1991 (Delaney 1993), although only the wild population is considered to have conservation value (Batten et al 1990). Shetland birds have the added conservation interest of being the only population in Britain presumed to originate entirely from Icelandic stock. Although native breeding Greylag Geese are protected by special penalties during the close season as they are on Schedule | of the Wildlife and Countryside Act (1981), this protection only applies to specific areas and does not extend to Shetland breeding birds. Despite or often because of their protected status, geese are infamous for their involvement in conflicts between conservationists and agriculture. Conflicts involving native Scottish Greylag Goose already exist. On Uist a goose management committee is attempting to control geese through organised shooting and scaring, and on Tiree crofters have organised 2 autumn culls by shooting and have applied, unsuccessfully, for licences to take or prick eggs (Mitchell 1999, Malcolm Ogilvie pers comm). The situation in Shetland has the potential to be added to the list of conflicts. As the favoured breeding habitat of Greylag Geese in Shetland has been created by the reseeding of hills, this has inevitably led to conflict with the crofters who have stock on these hills. Indeed it is already known that crofters in one area have been shooting birds on their land in summer in an attempt to control numbers. One further consideration is the question of why a population of birds should suddenly start breeding to the south of their normal breeding range. While it would appear that the stimulus for breeding in Shetland was reseeding providing suitable breeding habitat, it is possible that some other factor is involved. Barnacle Geese have also begun to colonise areas along their normal migration route. Birds from the Russian breeding population birds have been breeding in the Baltic since the 1970s (Ganter etal 1999). More recently, small numbers of the Greenland population have been breeding in Iceland (Ogilvie et al 1999). The large numbers of feral geese breeding in Britain have already established that geese are capable of thriving in a far wider range of habitats than those found in their normal breeding range. It will be interesting to see if the colonisation of Shetland by Icelandic Greylag Geese is a further part of a continuing trend. Acknowledgements Additional information on Unst breeding birds was supplied by Jim Burgess, Wendy Dickson, Harry Edwards, Pete Ellis, Mike Peacock, Duncan and Jan Sandison, the late Ian Spence and the late Bobby Tulloch. Most of the information on birds breeding in the rest of Shetland was provided by Pete Ellis and Paul Harvey but other information came from Nick Dymond, Andy Fitchett, Andy Gear, Martin Heubeck, Ruth Mackenzie, Magny Mouatt, Mary Ellen Odie, Dave Okill, Kevin Osborn, Malcy Smith, Roger Riddington, Terry Rogers, Jonathan Swale, Dave Suddaby, Brydon Thomason and Howard Towll. Useful information and comments were provided by Dorete Bloch, Pete Ellis, Paul Harvey, Jens-Kjeld Jensen, Eric Meek, Carl Mitchell and Malcolm Ogilvie. Scottish Birds (2000) Greylag Geese breeding in Shetland Se) References Batten L A, Bibby CJ, Clement P, Elliott G D and Porter R F 1990. Red Data Birds in Britain. T & A D Poyser, London. British Ornithologists’ Union 1971. The Status of Birds in Britain and Ireland. Blackwell, Oxford. Brown A W and Dick G 1992. Distribution and numbers of feral Greylag Geese in Scotland. Scottish Birds 16:184-191. Cramp S and Simmons K E L (eds) 1977. The Birds of the Western Palearctic, Volume I. Oxford University Press, Oxford. Cranswick P A, Pollitt MS , Musgrove A J and Hughes RC 1999. The Wetland Bird Survey 1997-98: Wildfowl and Wader Counts. BTO/WWT/RSPB/ JNCC, Slimbridge. Delaney S 1993. Introduced and escaped geese in Britain in summer 1991. British Birds 86:591-99. Dunn R 1837. The Ornithologists’ Guide to the Islands of Orkney and Shetland. Privately published. Dymond J N 1991. The Birds of Fair Isle. Privately published. Evans A H and Buckley T E 1899. A Vertebrate Fauna of the Shetland Islands. Douglas, Edinburgh. Ganter B, Larsson K, Syroechkovsky E V, Litvin K E, Leito A and Madsen J 1999. Barnacle Goose Branta leucopsis: Russia/Baltic. pp 270-83 in Madsen J, Cracknell G and Fox T (eds) Goose Populations of the Western Palearctic. A review of status and distribution. Wetlands International Publ. No.48, Wageningen, Netherlands. Harrison C 1982. An Atlas of the Birds of the Western Palearctic. Collins, London. Low G 1879. A Tour through the Islands of Orkney and Schetland. Melven Press, Inverness. Mitchell C 1999. Greylag Goose Anser anser: Scotland. pp 172-7 in Madsen J, Cracknell G and Fox T (eds). Goose Populations of the Western Palearctic. A review of status and distribution. Wetlands International Publ. No.48, Wageningen, Netherlands. Ogilvie M A, Boertmann D, Cabot D, Merne O, Percival S M and Sigfusson A 1999. Barnacle Goose Branta leucopsis: Greenland. pp 246-57 in Madsen J, Cracknell G and Fox T (eds) Goose Populations of the Western Palearctic. A review of status and distribution. Wetlands International Publ. No.48, Wageningen, Netherlands. Olsen, B. 1994. Er gragdsastovnurin vordin ov st6rur? Faroese Bird Report Owen M 1986. Greylag Goose. Lack P (ed) The Atlas of Wintering Birds in Britain and Ireland. T & A D Poyser, Calton. Owen M, Atkinson-Willes GL and Salmon D G 1986. Wildfowl in Great Britain. Cambridge University Press, Cambridge. PatersonI W, Boyer PR and MassenD 1990. Variations in clutch size and breeding success of Greylag Geese Anser anser in the Uists, Scotland. Wildfowl 14:18-22. Prater T 1993. Greylag Goose. in: Gibbons D W, Reid J B and Chapman D A (eds) The New Atlas of Breeding Birds in Britain and Ireland: 1988-1991. T & A D Poyser, London. Saxby H L 1874. The Birds of Shetland. Maclaren and Stewart, Edinburgh. SIC 1997. Shetland in Statistics 1997. Shetland Islands Council, Lerwick. Thom V M 1986. Birds in Scotland. T & A D Poyser, Calton. Venables LS V and Venables U M 1955. The Birds and Mammals of Shetland. Oliver & Boyd, Edinburgh and London. Williamson K 1970. The Atlantic Islands. Oliver & Boyd, London. M G Pennington, 9 Daisy Park, Baltasound, Unst, Shetland ZE2 9EA Revised manuscript accepted March 2000 36 Scottish Birds (2000) 21:36-42 SB 21(1) East Ross Rookeries in 1998-99 HENRY McGHIE East Ross rookeries were surveyed in 1998-9 and the results compared with those of the 1975 survey organised by the BTO. Overall, there was a 33% decline in the number of nests and a 20% decrease in the number of rookeries between 1975 and 1998. Declines were pronounced on the Black Isle and in Easter Ross, but there was a 13% increase in the number of nests in Mid Ross and a 65% increase in the number of rookeries. The median rookery size declined in all 3 areas. Rookeries in farm woodland and forestry plantations showed greater declines than those in the vicinity of villages and gardens and the majority of new rookeries were established in villages and gardens. Introduction Ross-shire, rookeries were last surveyed in 1975, as part of the national survey organised by the BTO (Castle 1977, Sage and Vernon 1978). A comparison of the results of this survey with the 1946-47 national survey (Anon 1947-48) demonstrated that there had been widespread declines in the numbers of Rook Corvus frugilegus nests and that rookery sizes had declined in many areas including Ross-shire. Further declines have been reported from Sutherland since the 1975 survey (Bremner and Macdonald 1996). I conducted a full survey of East Ross rookeries in 1998 and 1999 and compared the results with the 1975 survey data to establish whether or not there had been any further change in the status of rookeries. Study area and methods The study area consisted of East Ross east of Garve to the district boundaries in the vicinities of Beauly (south boundary) and Kincardine (north boundary); this area encompassed the entire East Ross range of rookeries found by the 1975 survey. Searches were restricted to land below 250m above sea level (asl) so as to include all agricultural land in the area excepting sheepwalk. The centres of large conifer plantations (>0.75km from edge) were not searched as these were not known to be used as a breeding habitat from extensive previous experience and from published accounts (eg Sage and Vernon 1978). The study area was divided into 3 geographical areas on the basis of differing land use. The Black Isle (c230km/) was defined as that area lying east of the main Conon Bridge to Inverness road (A835 and AQ). Easter Ross (c159km7”) was defined as that area lying east of NH76 of the National Grid which runs from near Invergordon to near Tain. Both of these areas consist of rich agricultural land developed over Old Red Sandstone (ORS). Whilst Easter Ross is low lying, mainly <50m asl, good quality arable land extends to 150 m asl on the Black Isle. Mid Ross (c300km7?) was largely developed within Moinian schists to the west and ORS to the east. Farmland was restricted to the floors of alluviated valleys and drift covered slopes in the west but was more widely developed towards the east. Nests or groups of nests separated by more than 100 metres were regarded as separate rookeries, as is usual practice (Anon 1947-8, Sage and Vernon 1978). Rookeries were counted in 1998 and 1999 Scottish Birds (2000) East Ross Rookeries in 1998-99 as) ™ with most counts being carried out between late March and the end of April 1998. Most areas were rechecked several times in 1998 and in 1999 when several previously overlooked rookeries were located. Similarly, 2 rookeries were located when sites that had been occupied in 1975 were checked but had been missed in 1998. Coverage was good in terms of the amount of time spent searching each area and in discussion with land users and naturalists. No rookeries were considered to have been missed. Four rookeries were counted in both 1998 and 1999 and showed little change (<10%): the direction of change was not the same for these rookeries and it was deemed permissible to combine data from the 2 years. A degree of subjectivity was required when counting large amalgamated nests in long established rookeries: simultaneous independent counts by several observers at several rookeries produced generally consistent numbers (<5% difference) and this demonstrated that the count method was sound. One rookery on an island on Loch Ussie was not counted directly but the maximum number of pairs was estimated from the size of the feeding flock. Rookeries were classified into 4 types: village rookeries, defined as those within 200m of settlements of more than 20 houses, garden rookeries, defined as those being situated within 100m of more isolated habitations, usually in gardens themselves but also in farm woodlands within 100m of isolated habitations. farm woodland rookeries included those in shelterbelts, coverts, policy woodlands and one on an island in a loch; all woodlands were under 5ha and most under 2ha. Forestry plantation rookeries consisted of those in large (>Sha) stands of conifers. Rookeries which showed a change in numbers in excess of 10% of the 1975 count were considered to be increasing or decreasing. Statistics follow Fowler and Cohen (BTO Guide 22) throughout. Results The 1975 survey showed that rookeries were scattered across the whole of the Black Isle and Easter Ross, and in the lowlands of Mid Ross as far west as Marybank. In 1998-9, rookeries were found slightly (3km) further west to Contin, the western limit of cereal cultivation with 74% of rookeries being below 50m as] and 15% 50-100m asl. Most (75%) of the rookeries which were above 100m asl were on the Black Isle. where good quality agricultural land was developed at higher altitudes due to differences in underlying geology. These figures were remarkably similar to those found in the 1975 survey (76% and 90% below 50 and 100m asl respectively), and there Was no significant difference in the proportions nesting at different altitudes in 1975 and 1998-99 (not significant: chi-squared, 2 degrees of freedom). The highest and most isolated rookery which was located in 1975, at 180m as] in Strath Rusdale, was not in use in 1998. Of 139 rookeries used in 1975 and/or 1998-9, 56% were in stands of old pines (Scots Pinus sylvestris in 77 rookeries and Corsican P nigra in 3): 7% were in stands of other conifers, mainly Norway Spruce Picea abies; 25% of rookeries were in broadleaved woodland, with Beech Fagus sylvatica constituting 9%: 4% of rookeries were in mixed conifer/broadleaved woodland. A variety of tree species was thus involved, with more unusual ones including Chilean Pine Araucaria araucana in 2 rookeries, Western Hemlock Tsuga heterophylla in one rookery and birch Betula sp in 2 rookeries. In 3 rookeries in young conifers, rook nests were constructed one above the other in the tiers of branches of young conifers. Woodlands which were apparently suitable for nesting were well distributed throughout Mid Ross and the Black Isle and, whilst large woods were more scattered in Easter Ross, small farm woodlands and plantings along field boundaries were still 38 Henry McGhie SB 21(1) abundant there. Breeding habitat was not considered to limit the distribution of rookeries. On the Black Isle, the 1975 survey found 3,376 nests divided between 23 rookeries (Tables | and 2). Four rookeries, all on the south side of the crest of the Black Isle, contained over 300 nests whilst a quarter of rookeries contained less than 50 nests. The 1998-9 survey found a 54% decline to 1539 nests in 17 rookeries and only one rookery exceeded 300 nests. Two thirds of 1975 rookeries were not inuse in 1998-9 and median rookery size decreased significantly (P<0.01, Mann-Whitney U test). There was variation in the pattern of change within the area: on the north side of the Black Isle, there was a 19% reduction in the number of nests and half of the 1975 rookeries were extant in 1998-9. On the south side of the Black Isle there was a 62% reduction in the number of nests and only 21% of 14 rookeries were used in both years. In Easter Ross, the 1975 survey located 2,361 nests divided between 43 rookeries (Tables | and 2). Only 2 rookeries exceeded 200 nests; median rookery size was significantly lower than in the other 2 areas in both 1975 and 1998-9 (P<0.01, Mann-Whitney U tests). The 1998-9 survey found 1209 nests, a decline of 49%, which were divided between 20 rookeries. Of the 43 rookeries found in 1975, 26% were used in both years. There were no rookeries containing over 200 nests in 1998-9 and the median rookery size had declined significantly (P<0.01, Mann-Whitney U test). In Mid Ross in 1975, a total of 2,284 nests were located in 20 rookeries (Tables | and 2). Three rookeries exceeded 200 nests. The 1998-9 survey found 2570 nests in 32 rookeries, a 13% increase on the number found in 1975; 2 of these exceeded 200 nests. Mid Ross was the only area in which the numbers of nests and rookeries increased between 1975 and 1998-9. Of the 20 rookeries found in 1975, 60% were used in both years, a higher proportion than in the other two areas. The median rookery size was higher than in the other 2 areas in both 1975 and 1998-9 although the difference between Mid Ross and the Black Isle was not significant (Mann-Whitney U tests). The median rookery size declined significantly between 1975 and 1998-9 (P<0.01, Mann-Whitney U test). Changes in the numbers of nests varied within the area: numbers around Muir of Ord and Marybank were stable (2% increase), but there were increases west of Dingwall and along the north side of the Cromarty Firth (13% increase). Numbers decreased along the north side of the Beauly Firth (48% decrease). The density of Rook nests was greater in the Black Isle and Easter Ross with 14.7 nests per km? and 14.8 nests per km? respectively than in Mid Ross with 7.6 nests per km? in 1975. This situation was reversed in 1998-9, when Rook nests were found at greatest density in Mid Ross at 8.6 nests per km’; density had halved in the Black Isle and Easter Ross to 6.7 nests per km? and 7.6 nests per km? respectively. When the 1975 and 1998-9 counts were compared, large declines (>40%) were found at 93% of rookeries in woodland habitats and 77% were completely unused in 1998. Large declines were fewer in gardens (46% of rookeries) and villages (25% of rookeries). Large increases (>40%) were found at 3%, 31% and 25% of rookeries in woodland, gardens and villages respectively. There was some variation in this between the 3 areas: in woodland habitats, large decreases were highest in the Black Isle (96% of rookeries) and lowest in Mid Ross (82% of rookeries). Declines were noted at a higher proportion of garden rookeries in Mid Ross (72%) than in the Black Isle (20%) or Easter Ross (34%) although it was difficult to draw conclusions from these comparisons as the number of rookeries in each category was small. In 1975, rookeries in farm woodland and forestry represented 83%, 88% and 60% of colonies in the Scottish Birds (2000) East Ross Rookeries in 1998-99 39 Table 1 Change in numbers of Rook nests 1975-98/9. No of nests Median rookery size Maximum rookery size 1975 1998-99 % change 1975 1998-99 1975 1998-99 Black Isle 3376 1539 -54% 62 55 897 334 Easter Ross 2361 1209 -49% 35 28 Pa 199 Mid Ross 2284 2570 +13% 100 74 524 pss) Total 7962 5318 -33% 50 52 897 334 Table 2 Change in status of rookeries 1975-98/9. Number of rookeries No of rookeries used in abandoned decreasing increasing new in 1975 1998/99 1975 & 98/99! 1975-98-99 1975-98/997 1975-98/99? 1998/99 Black Isle des, 119 a 16 - 3 10 (0.30) (0.70) (0.17) (0.13) Easter Ross 43 20 ist 32 7 - 9 (0.26) (0.74) (0.16) (0.09) Mid Ross 20 32 Ww 8 6 - 20 (0.60) (0.40) (0.30) (0.25) Total 86 69 30 56 17 1] 39 (0:35) (0.65) (0.20) (0.13) 1 __ Figures in brackets represent the proportion constituted by each class of the total number of 1975 rookeries. 2 Rookeries showing >10% change in numbers of nests between 1975 and 1998. Black Isle, Easter Ross and Mid Ross respectively (Table 3). In 1998-9, these figures had dropped to 53%, 50% and 44% respectively. This was due to significant increases in the numbers of nests in gardens and villages (P<0.01; chi-squared tests, | degree of freedom) and decreases in the numbers of nests in woodland habitats (P<0.01; chi-squared tests, | degree of freedom). In the Black Isle and Easter Ross, gardens and villages were the only habitats in which the numbers of nests androokeries increased in real terms, whilst in Mid Ross the proportional increase was fivefold that in woodland habitats. Half of 10 rookeries established on the Black Isle between 1975 and 1998-9 were in gardens or villages, a large proportional increase considering the small extent of these habitats when compared to woodland. In Easter Ross, two thirds of 9 rookeries established between 1975 40 Henry McGhie SB 21(1) Table 3 Changes in nesting habitat 1975-98. Farm woodland Forestry Gardens Villages 1975 1998-99 1975 1998-99 1975 1998-99 1975 1998-99 Black Isle Nests ATS 799 466 5/5) Bis) AGI 97 424 (0.82) (0.52) (0.14) (0.04) (0.01) (0.17) (0.03) 27) Rookeries V7 8 2 1 | 4 3 4 (0.74) (0.47) (0.09) (0.06) (0.04) (0.24) (0.13) (0.23) Easter RossNests 27 832 0 0 234 314 0) 63 (0.90) (0.69) (0.10) (0.26) (0.05) Rookeries 38 10 0 0 5 6 0 4 (0.88) (0.50) (0:12), (30) (0.20) Mid Ross Nests 931 IZ 524 0 708 665 | 653 (0.41) (0.49) (0.23) (ORS ID) (O28) (0.05) (0.25) Rookeries 1] 14 1 0 7 a l oT (0.55) (0.44) (0.05) (0.35) (0.34) (0.05) (O22) Total Nests 3833 2883 990 35) 980 1240 218 1140 (0.73) (0.54) (0.12) (0.01) (O12) O23) (0.03) (0.21) Rookeries 66 32 3 l 13 21 4 15 (0.77) (0.46) (0.03) (0.02) (0.15) (0.30) (0.05) (0.22) I Numbers in brackets represent the proportions constituted by each class of the total for each year in each area. and 1998-9 were in gardens and villages. In Mid Ross, a third of 20 rookeries set up between 1975 and 1998 were in villages and a quarter were in gardens. Discussion The numbers of nests in rookeries is known to vary through the breeding season as late breeders settle to breed and as the nests of failed breeders are dismantled by other rooks securing sticks. Dunnet and Patterson (1968) recommend that repeat counts of rookeries are made and the maximum count is used. Counts should ideally be made in April but time constraints meant that this was not possible in the present study; however, the majority of counts were made between late March and April 1998. This was undoubtedly the single largest source of error but this has been a problem of all previous surveys (eg Sage and Vernon 1978). There were general correlations between declines in the number of nests, which is related to survey date, and the proportions of rookeries which were completely abandoned, which is not related to survey date, for each area and for each habitat. This provides circumstantial evidence that the differences between the 1975 and 1998-9 surveys were real and were not artefacts of the survey method. The study area was searched on a more or less random basis, with small sections being surveyed on each day, so that the large differences between the 3 sections of the study area were not Scottish Birds (2000) East Ross Rookeries in 1998-99 4] attributable to the date on which they were surveyed. The numbers of Rook nests declined in the Highlands between the 1880s and 1945-46 (Anon 1947-48), followed by large increases between 1945-46 and 1975 in Sutherland (41%) and Caithness (14%) but no figure is given for Ross- shire (Sage and Vernon 1978). A national sample survey of rookeries suggested that there was a 5- 9% increase between 1975 and 1980 (Sage and Whittington 1985) and some of the changes in Mid Ross were known to have occurred around this time. The timings of establishment and change in rookeries were imperfectly known. Much of the change on the Black Isle was thought to have occurred in recent years (post 1990), as nests and nest debris could be seen in some of the disused rookeries; this was supported by discussions with locals. The growth of the village rookeries on the Black Isle was a recent phenomenon, providing circumstantial evidence that at least some of the birds had moved from nearby farm woodland. Most of the new rookeries in Mid Ross, on the other hand, were known to be well established in 1982 although there were also more recent increases in Muir of Ord: one rookery which was established c1983 by c15 pairs contained 122 nests in 1998. A rookery which was established in nearby Beauly (Inverness-shire) by 4 pairs in 1997 contained 10 nests in 1998 and 12 nests in 1999. A further rookery of c10 pairs which was established in Muir of Ord in 1999 was not included in the present survey as these birds could have relocated from a nearby rookery which had been counted in 1998. Inan area of cSkm? near Dingwall, there was a 43% net increase in numbers between 1975 and 1998, with a 27% increase between 1976 and 1977 (R Graham pers comm). Persecution was heaviest on the south side of the Black Isle and in Easter Ross, where syndicates of farmers targetted rooks, but occurred at lower levels throughout the study area. Many rookeries suffered low levels of shooting with air rifles, mainly by children, but this method did not destroy many birds. More extreme persecution was in the form of blasting nests from below with shotguns during the breeding season, or by disturbing incubating birds and preventing them from resettling by discharging guns into the air. The 2 main persecution methods thus left the adult birds alive. The greatest declines between 1975 and 1998 were on the south side of the Black Isle and this area also demonstrated the greatest shift away from nesting in woodland habitats to nesting in villages and gardens. In an area of c82km° of prime agricultural land, the proportion of rookeries in woodland habitats fell from 69% in 1975 to 25% in 1998-9. Declines on the north side of the Black Isle were much smaller than on the south side. These 2 areas have similar land use and apparently differ only in the extent of persecution, with much more severe persecution on the south side. This suggested that persecution did at least contribute towards the decline of rookeries on the south side and to the shift from woodland habitats into villages and gardens. The contribution made by persecution towards declines in other areas is unknown although there is heavy shooting pressure in Easter Ross. Dunnet and Patterson (1968) and Sage and Vernon (1978) had found persecution to be of negligible importance in controlling the numbers of rooks; however Feijen (1976) found that persecution was responsible for large declines in the Netherlands. Itis unknown whether different methods of control or some other factor associated with persecution were responsible for the difference in efficacy of persecution between the present study and those of Dunnet and Patterson (1968) and Sage and Vernon (1978). The widespread declines on the Black Isle and in Easter Ross may also be related to changes in the 42 Henry McGhie SB 21(1) food supply. There have been large increases in the acreage of cereals throughout the study area in recent years (pers obs); these have been most marked in the Black Isle and Easter Ross and may have reduced the area of pasture. In Mid Ross, where agriculture is less intensive, increases in cereals have not reduced the amount of grazing land to the same extent as the steep sides of the straths are unsuitable for cereal cultivation. Brenchley (1984) found that Rooks favoured agricultural regimes in which the acreages of pastoral and arable land were approximately equal. Increases in the acreage of cereals may have approached a balance more favourable to Rooks in Mid Ross but may have reduced the proportion of pastoral land to a less favourable regime dominated be cereals in the Black Isle and Easter Ross. Rookeries have shown persistent long term declines in size (eg Sage and Vernon 1978); in the present study, rookeries declined in size in Mid Ross although there was an overall increase in numbers. The reduction of colony sizes may be a response to changes in food supply, possibly due to changes in the availability of invertebrates during the summer when Rooks have most difficulty in obtaining food (Dunnet and Patterson 1968, Feare etal 1974). The trend towards breeding inthe vicinity of gardens and villages may indicate that Rooks are obtaining a greater proportion of their diet from human refuse or from food deliberately put out for birds or it may be in avoidance of persecution. These are offered as possible causes of the changes in numbers of Rooks but are speculative. Acknowledgements I am grateful to Brian Etheridge for making the 1975 survey data available to me and to Ron Graham for providing some 1970s counts and for useful discussion on Rooks. Iamalso very grateful to Dr Ian Patterson and an anonymous referee for commenting upon drafts of this paper and making significant improvements. References Anon 1947-48. Rook Investigation. Report received by the Ministry from the Agricultural Research Council. Journal of the Ministry of Agriculture 55:20-23. Brenchley A 1984. The use of birds as indicators of changes in agriculture, ed E N Wright, IR Inglis & C J Feare pp20-27. British Crop Protection Council, Croydon. Bremner D M & Macdonald D 1996. Decline of Sutherland rookeries. Scottish Birds 18:248. Castle M E 1977. Rookeries in Scotland 1975. Scottish Birds 9:327-334. Dunnet G M & Patterson I J 1968. The Rook Problem In North-East Scotland. In The Problems of Birds as Pests pp119-139. Academic Press, London. Feare CJ, Dunnet GM & PattersonI J 1974. Ecological studies of the Rook Corvus frugilegus in north-east Scotland: food intake and feeding behaviour. Journal of Applied Ecology 11:867-896. Feijen H R 1976. Food, occurrence and decline of the Rook, Corvus frugilegus, in the Netherlands. Limosa 49:28-66. Fowler J & Cohen L no date. Statistics for Ornithologists. BTO Guide 22. Sage BL & Vernon J D R 1978. The 1975 National Survey of Rookeries. Bird Study 25:64-86. Sage B L & Whittington P A 1985. The 1980 sample survey of rookeries. Bird Study 32:77-81. Henry A McGhie, The Manchester Museum, University of Manchester, Oxford Road, Manchester M13 9 PL Revised manuscript accepted March 2000 Scottish Birds (2000) Short Notes 21:43-55 43 SHORT NOTES Table 1 Breeding success of Merlins monitored in Galloway 1965-98. Breeding success of Merlins in Galloway Year Nest areas No No nests No nests checked occupied eggs laid young Galloway, in south west Scotland, is the most fledged heavilycommercially afforestedareainScotland === 2 © and upland Kirkcudbrightshire is the only 1965 7 7 Zs 1 region with over 30% forest cover. Conifer 1966 8 8 3 1 planting commenced in 1921 and by the 1970s 1967 7 7 2 1 over 80% of land below 300m had been planted; 1968 2 9 - 1 the district still has amongst the highest levels of 1969 11 11 3 1 current planting in Scotland (Avery & Leslie 1970 10 10 ] | 1990, Birds and Forestry, London). This note 1971 9 9 6 5 summarises the breeding success of Merlins Falco 1972 6 6 + 3 columbarius in Galloway mainly in afforested 1973 13 13 5 5 areas between the years 1965-98 for which few 19/4 17 17 9 7 data have been published before. 1975 13 9 3 2 1976 14 10 4 3 Fieldwork commenced in 1965andcoveredabout 1977 14 8 4 3 70-80% of the 310km? where most Merlin breeding 1978 12 7 5 3 pairs were concentrated. The results (Table 1) nono I 8 2 1 show that of the 337 nest areas monitored, 234 1980 13 9 3 y (69%) were occupied. Only 89 (38%) of those 1981 13 6 l 1 pairs reached the clutch stage and of those, 61 1982 12 5 Z 2 (68.5%) were successful in fledging young: the 1983 1] q 2 Z average number of fledged young per breeding 1984 10 5 1 1 attempt was 2.5 (ground nests) and 2.3 (tree 1985 6 3 1 ] nests). Merlin nests were widely dispersed but 1986 7 5 1 1 were found to be separated by an average distance 1987 1 5 2: 2. of 2.8km(1-6 pairs/100km’). 1988 8 6 1 1 1989 6 4 1 1 Although the first tree nest was found in an 1990 2 5 5 - isolated tree in 1955 (A D Watson in Jitt) it was 1991 IA 5 Ze 1 not until 1982 that the first nest was found ina 1992 8 3 2 2 Sitka Spruce Picea sitchensis plantation. In1993— 1993 17 15 7 4 95 aforest edge survey of tree nests was made. Of 1994 1] 10 5 5 the 37 breeding areas checked, 13 (35%) were 1995 9 3 ] l situated in trees, 100% of those reached the 1996 af I = clutch stage and 84.6% were successful. The 1997 : ; - number of fledged young per breeding attempt 1998 was 2.2, about the same as at tree nests overall. There was 28 (31.5%) breeding attempts which Totals 337 234 89 61 failed completely. The causes include Fox Vulpes 44 Short Notes SB 21(1) vulpes (3) and Adder Vipera berus (2) predation; egg desertion (1); eggs burnt (1) eggs or young disappeared, cause unknown (17); young died (1); female found dead at nest (1); young fell through tree nest (1); nest robbed (1). Clutch sizes ranged from 2—5 eggs with a mean of 4.1140.93 SD. Of the 89 nests, 51 (57%) were situated on the ground; 17 (19%) were tree nests; 13 (15%) on rocky shelves amongst Heather Calluna vulgaris; 5 (6%) on sheer cliff faces on Heather ledges; 2 (2%) on top of boulders and | (1%) at the base of a Rowan Sorbus aucuparia. The altitudinal range of the nests ranged from 76m (250’) to 364m (1200’) with 52.6% of nests situated within the 152m (500’) to 242m (800’) contours. The results obtained in this study show that the overall success rate (68%) was exactly the same as that obtained in north east Scotland (where the habitat is different) in 1980—89 (Rebecca etal 1992, Scottish Birds 16:165—183), but 8% lower than for Scotland as a whole in 1988—90 (Table 2). The occupancy rate overall was 69% comparable to the data from north east Scotland but lower than the 78% for the Scottish national average in 1988-90. There are 54 known Merlin breeding sites in Galloway (42 in Kirkcudbrightshire, 12 in Wigtownshire) but not all are now occupied annually or regularly. At least 12 (40%) of the 30 traditional ground nests in the core areas in Kirkcudbright which were regularly used in the 1960s-1970s have now been completely abandoned with gaps in use of 20-30 years. The reasons are apparent with the sites being displaced by large blocks of dense mature conifer plantations. This in turn has had a direct effect on Merlin foraging activities by reducing open ground for hunting and acomplementary decrease in the numbers of their most common prey species, Meadow Pipits Anthus pratensis (unpublished data; see also Watson, 1979 Bird Study 26:253-— 258). With a decrease in the number of ground nests in core areas there may be a reciprocal increase in the use of tree nests at moor edges as has happened at Kielder Forest (Little et al 1995, Forest Ecology and Management 79:147-—152) and in Wales (Parr 1991, Bird Study 38:103-— 111). The results show that tree nests, when these are available, have a success rate of 76%, exactly the same as the overall success rate of all nests in Scotland (Table 2). I thank all those who provided data including E L Roberts, R Roxburgh, G Shaw and A D Watson amongst others. Table 2 Pooled data extracted from Raptor Round Up Reports of breeding success of Merlins in Scotland 1988-90. Data from Scottish Bird News 14:6-7, 18:8-9, 22:8-9. Sites checked 909 Sites occupied 707 ~—s (8%) Sites eggs laid 548 (78%) Sites young fledged 419 (76%) No. young per success 1264 3) R C Dickson, Lismore, New Luce, Newton Stewart, Wigtownshire DG8 0AJ Revised manuscript accepted September 1999 Scottish Birds (2000) Short Notes 45 Mute Swans rearing Greylag Geese Glendevon Pond is a small farmyard pond in rural West Lothian. For many years a small island has been used by nesting Mute Swans Cygnus olor. On 5 April 1998 a pair of Mute Swans along with several Greylag Geese Anser anser were present but as yet there were no signs of breeding. However, a subsequent visit on 20 April found the pair of Mute Swan building a nest on the island. On 8 and 10 May the swans were on their nest apparently incubating eggs while several Greylag were loafing around. On 20 May the unusual news that the swans were tending a brood of 4 young Greylag goslings was reported to me. A visit on 23 May confirmed this. The pair of swans had abandoned their nest and eggs (visible in nest) and were caring for 4 young Greylag goslings. Given the vigilance and aggression of nesting Must Swans it seems unlikely that the Greylag Geese would be able to lay their eggs in the swan’s nest. Therefore, it would appear the Mute Swans built their nest on top of an already existing Greylag nest and eggs. Presumably the Greylag eggs having been laid earlier and with a shorter incubation period than the swan eggs hatched first and the goslings were adopted by the swans at the expense of their own clutch. One of the 4 goslings vanished shortly after hatching. However, the remaining 3 developed well and dispersed the following spring. Unusually for Greylags, the geese fed for a short while by upending, a habit seemingly copied from their foster parents. However, they did not appear very adept and soon abandoned this method and fed by grazing. Gavin Paterson, 9 Howieson Green, Uphall, West Lothian EH52 6BW Manuscript accepted September 1999 A new record of successful tree nesting Peregrines In Europe, Australia and North America Peregrines Falco peregrinus have often been recorded nesting in old stick nests in trees (Ratcliffe D A 1993, The Peregrine Falcon. TY & A D Poyser, London; Emison W B, White C M, Hurley V Gand Brimm DJ 1997, Factors influencing breeding distribution of the Peregrine Falcon in Victoria Australia. Wildlife Research 24:433-444; Campbell R W, Paul M A and Rodway M S 1978, Tree-nesting Peregrine Falcons in British Columbia. Condor 80:500-1). In Britain only 3 reports of tree nesting have been recorded: 2 unsuccessful attempts in one nest (1983 & 1985) in a Scots Pine Pinus silvestris tree (Ratcliffe D A 1984, Tree nesting by Peregrines in Britain and Ireland. Bird Study 31:232-233; Ratcliffe D A 1993, The Peregrine Falcon. T and AD Poyser, London) and one successful attempt in a Scots Pine tree in Wales (I Williams in Ratcliffe 1993). Both of these nests were in old Raven Corvus corax sites and the sites appeared to be alternatives to local cliff sites. Ratcliffe (1984, Tree-nesting by Peregrines in Britain and Ireland. Bird Study 31:232-233) suggested that tree nesting may have started in the 1980s because of the increase in the Peregrine population and thus a decrease in the number of available cliff sites. However, since these reports from the mid 1980s we know of no other documented tree nesting Peregrines. This is despite the population increasing from 768 notional pairs in 1981 to 1283 in 1991 (Ratcliffe D A 1984, The Peregrine breeding population of the United Kingdom in 1981. Bird Study 31:1-18; Crick HQ P and Ratcliffe D A 1995, The Peregrine Falco peregrinus breeding population of the United Kingdom in 1991. Bird Study 42:1-19). Here we 46 Short Notes SB 21(1) report the successful breeding by Peregrines in a tree nest in Scotland in 1999. The site has been checked since 1992 and birds have been present on territory each year since 1994. Small ledges are available in the area along the course of some moorland streams and birds have attempted to breed in these sites on anumber of occasions, though not successfully. The territory is on the edge of an area of Heather moorland and rough sheep grazing. There were no other Peregrines nesting within a 5km radius of this pair. During the spring of 1999 a pair of territorial Peregrines were again noted in the area. This year the usual sites were not used, and the birds settled further down the valley, laying their eggs in an old Buzzard Buteo buteo tree nest (Photo). The large stick nest was situated 7-8m up in a fork of an Ash Fraxinus excelsior tree growing from the bottom of a steep sided gully. When found (21 April 1999), the female was sitting on 4 eggs. Three small chicks were present on 14 May 1999 and these 3 chicks successfully fledged. In other parts of the world, tree nesting is relatively localised and the reasons for it not spreading remain unclear (Newton I 1979, Population Ecology of Raptors. T & A D Poyser, London). It will be interesting to see if these Peregrines and their offspring utilise stick nests in future years. Good cliff sites are scarce in the general area, whilst populations of Ravens are increasing and Buzzards Buteo buteo are at high density. Stick nests are therefore relatively abundant and aswitch from cliffs to trees could lead to an increase in the local Peregrine population. Fiona Leckie, Centre for Ecology and Hydrology, Hill of Brathens, Banchory AB31I 4BY Steve Campbell, Game Conservancy Trust, Institute of Cell Animal and Population Biology, University of Edinburgh, West Mains Road, Edinburgh EH9 3JT Tree nesting Peregrine Manuscript accepted October 1999. = - ‘ Steve Campbell Scottish Birds (2000) Post breeding congregations of adult Black-throated Divers in Wester Ross Knowledge of the distribution and movements of Scottish Black-throated Divers Gavia arctica away from breeding territories is poorly known. This note describes the post breeding congregations of adult Black-throated Divers that occur in Loch Gairloch and around the Applecross peninsula on the southwest coast of Wester Ross (Fig 1). From July to December in 1997 and 1998, I noted all sightings of Black-throated Divers on the sea. Counts were attempted once per week butregarded as accurate only when optimum weather conditions and calm seas prevailed allowing good visibility. Table 1 gives the peak monthly count for each location. The counts, especially at Loch Gairloch during July to September, may have included adults still attending chicks on breeding lochs. However, I have assumed that the majority had finished breeding. Numbers built up during the late summer, peaking in September and October, and then declining during late autumn and winter. Only 2 juveniles were seen both at Applecross in 1998. These are not included in Table 1. The divers completed their post breeding body moult during the period of the counts, and the majority attaining winter plumage by late November. In calm conditions, birds tended to gather in flocks while in rougher conditions single Short Notes 47 Figure 1 Location of Loch Gairloch and the Applecross Peninsula on the coast of Wester Ross ae ~~ ij Loch Gairloch ey, » Applecross Peninsula Table 1 Peak monthly counts of adult Black-throated Divers. July August September October November December Applecross Peninsula 1997 11 15 AT 49 2]. 9 1998 - 5 44 58 11 7 Loch Gairloch 1997 28 17 25 33 0 0 1998 13 17 27 17 21 0 4§ Short Notes SB) birds were more often encountered. The peak counts at Applecross in September 1997 and 1998, and October 1997 and 1998 consisted of single flocks of 47, 44, 49 and 58 respectively. When conditions were calm and the flocks were close to land, birds could often be heard calling with a soft ‘honk’. The Black-throated Diver is arare breeding bird in Britain with an estimated population size of 151 summering pairs (Campbell L H & Talbot T R. 1987 Breeding Status of Black-throated Divers in Scotland, British Birds 80:1-8). Itis an amber list in ‘Species of Conservation Concern in the UK, Channel Islands and Isle of Man’ (Gibbons et al. 1996 RSPB Conservation Review), and is classed as a species of European Conservation Concern category 3 — vulnerable (Tucker G M & Heath M F. 1994, Birds in Europe: Their Conservation Status, Birdlife International). The stretch of sea encompassing Loch Gairloch and the Applecross peninsula is a regular thoroughfare for commercial, military and leisure craft. As these congregations occur soon after the breeding season it would seem that a high proportion of the total Scottish breeding population may be present at these 2 locations and therefore vulnerable to any oiling incident. I would like to thank Dr Ron Summers for commenting on an earlier draft of this note. Chris Donald, Primary Schoolhouse, Ruthven Road, Kingussie, Inverness-shire PH2ZI1EN Manuscript accepted November 1999. Swifts in Speyside Tree nesting Swifts Apus apus were referred to in Scottish Birds 20:27-30. In the latter half of the 1980s a triple storeyed nest box based on a successful Swedish pattern was attached to a Birch Betula sp on the edge of Abernethy Forest. This box is believed to have been continuously in use since 1990. In 1999 it was altered to allow access. In 1990 a similar box was fastened to a Scots Pine Pinus sylvestris some 250 metres away from the first site. Both boxes are fixed to the north side of the tree, with holes facing north, east and west. The Birch box has roughly circular holes, the Pine horizontally segment shaped. The latter box was not used by Swifts until 1997. Also fastened to the Pine is a large Robin/Tit type box with removable front. This has been in situ for a number of years and has often been occupied by Starlings Sturnus vulgaris. Swifts had, on occasion, prospected, so to encourage them the circular holed removable panel was replaced with one of segment shape. This failed to deter the Starlings but, in 1998 and 1999, Swifts moved in within days of the young Starlings fledging. This box faces east. Swift arrival dates at Scots Pine site were 10 June 1997, 6 June 1998 and 27 May 1999. The only known fledging date is for the 1997 pine site when the single young fledged between 8 and 11 September. Starlings have used 4 of the 6 nest box compartments over the years and also the smaller box. There have been 6 occupied Swfit nests in the past 3 years. Entrance holes facing all 3 compass points have been used. None appear to have laid Scottish Birds (2000) Short Notes 49 more than the normal clutch of 2eggs. Three pairs have produced 2 young and 3 pairs a single chick. All the young have been ringed, plus 7 adults. Two of the latter have been retrapped at the nest. Until 1999 the contents of the Birch tree site remained unknown. My thanks are due to RSPB staff for information and agreeing to nest box alterations and ringing; and also to generous friends who for many years have kindly allowed me to festoon their garden trees with a variety of nest boxes. Harvey Burton, Braenedin, Nethybridge, Inverness-shire PH25 3EF Revised manuscript accepted December 1999 Mallard eating Eel On 9 November 1999 on the Water of Leith, in Edinburgh, I came across a female Mallard Anas platyrhynchos with a small Eel Anguilla anguilla inits bill. The Eel was wriggling actively and was about 15cm long, with a body diameter of about half the width of the duck’s upper mandible — it was difficult to be more accurate about the dimensions due to the Eel’s coiling. The duck was swimming with 2 other females and a male in shallow water at the river’s edge. The duck was manoeuvring the Eel in its bill and shaking it in the water. Eventually, after a minute or so, the duck managed to swallow the Eel head first, despite close attention from the accompanying Mallards which had made token efforts to steal the Eel. The Eel stayed down, but the Mallard had to make several neck stretches and gaping motions, presumably to aid swallowing. The duck then drank some water and swam off. Although the Mallard is described as an opportunistic feeder, with “fish” listed among its food items (BWP Vol 1, p 509), this note may be of some interest given the specific identification and size of the prey item involved. Tom Dougall, 62 Leamington Terrace, Edinburgh EH10 4JL Revised manuscript accepted January 2000 Jackdaw killing and carrying newly fledged Blackbird in its feet On 30 May 1999 we watched a newly fledged Blackbird Turdus merula begging for food on the lawn of a cottage garden in Wigtownshire. A Jackdaw Corvus monedula, which had young ina nest nearby, landed beside the fledgling as did the male Blackbird. The Jackdaw struck the juvenile a couple of blows to its head and body and then a near fatal blow to its head, picked it up in its feet, and flew up towards its nest pursued by the adult Blackbird. The Jackdaw soon dropped the dying fledgling, still pursued by the Blackbird. On inspecting the fledgling there was no outward sign of injuries, its skin was unbroken but it died soon after. Birds of the Western Palearctic Volume 8 states that food is usually carried in the Jackdaw’s bill but gave one record of a Jackdaw transporting a nestling thrush in its foot, flying some distance pursued by 4 adult thrushes. RC &A P Dickson, Lismore, New Luce, Newton Stewart, DG8 0AJ Manuscript accepted January 2000 50 Short Notes SB 21(1) Merlins selectively preying on Chaffinches In Galloway the most common prey species of Merlins Falco columbarius in the breeding season is the Meadow Pipit Anthus pratensis (unpublished data, see Watson D 1979, Bird Study 26:253- 258). Because most of the uplands below 300m in Kirkcudbright have been planted with exotic conifer plantations (Avery M & Leslie R 1990, Birds and Forestry) there has been a complementary decrease in the number of Meadow Pipits available as prey. The spread of these plantations has undoubtedly influenced the increase in the number of scrub and woodland species. At one Merlin breeding site in 1995, which produced one young in a tree nest situated at the edge of open moorland, there was some evidence to indicate that the adults were selectively preying on one woodland species, Chaffinches Fringilla coelebs. On the 8 and 16 July 1995 I collected the prey remains from all plucking posts in and around the nest and they consisted of the usual varied remains of 11 Meadow Pipits, one Chaffinch, one Siskin Carduelis spinus, one Snipe Gallinago gallinago and one Northern Eggar Moth Lapsiocampa quercus. [returned on 25 July and again collected all the remains, but in the intervening period they were composed entirely of 9 Chaffinches. This suggests that the adult Merlins had been selectively preying on this readily available food source rather than search in the surrounding moors for pipits. On 30 July prey remains were of 2 Meadow Pipits and 2 Chaffinches. R C Dickson, Lismore, New Luce, Newton Stewart, Wigtownshire DG8 0AJ Revised manuscript accepted January 2000 Merlin prey Scottish Birds (2000) Ringed Plover using a covered nest site On 12 July 1999, during a visit to the island of Auskerry, Orkney, R G Adam found a Ringed Plover’s Charadrius hiaticula nest with 3 eggs, hidden in a tunnel under a sloping concrete slab situated close to the shore. The nest was about 40cm inside the tunnel and the birds could only gain access to the nest through one end of the tunnel, as the other end was blocked by a stone. Fairly regular visits have been made to the island over the last 30 years and a number of Ringed Plover nests found, but all of these have been in the open. Short Notes ul It is unusual for a Ringed Plover to use a covered nest site but this has been recorded before in Orkney. Laidlaw in 1913 (Scottish Naturalist 1913:212) found 4 pairs on Auskerry with nests under stones. Pennington in 1987 (British Birds 85:498-499) noted that, on the island of North Ronaldsay,7 out of 18 Ringed Plover nests located were under cover. These nests appeared to fare better than those in the open and he suggested that this type of site would give protection from predation by birds as well as from North Ronaldsay sheep, which feed around the shore. These sheep trample on nests and have also been recorded eating Ringed Plover eggs. Itis perhaps of interest that, in the last few years, North Ronaldsay sheep have been introduced to Auskerry and now form the major part of the sheep flock on the island. C J Booth, 34 High Street, Kirkwall, Orkney KWI15 [AZ Manuscript accepted January 2000 Interaction between Greylag Goose, Great Northern Diver and Bottle-nosed Dolphins The waters around the Isle of Tiree have been the home of a mother and calf Bottle-nosed Dolphin Tursiops truncatus since 1994. The mother, nicknamed Sparkle by the Hebridean Whale and Dolphin Trust, and her calf Starlight are regularly seen throughout the summer months, though less so during the winter, in Gunna Sound, between Tiree and Coll. Here on4 July 1999, Carl Mitchell, Gregg Corbett and I observed a Greylag Goose Anser anser behaving oddly 150m away from the boat. The bird was diving then surfacing after a short distance before frantically flapping across the surface of the water. This continued for about a minute until an adult Bottle-nosed Dolphin breached in front of the Greylag. The Greylag headed in the opposite direction, flapping across the surface of the water. The dolphin kept appearing just behind the bird or breaching ahead of it. We did not see the dolphin attempting to bite or catch the Greylag; it seemd to be enjoying the chase. After 4 or 5 minutes the dolphin appeared to lose interest and drifted off with her calf, which was nearby during the incident. The Greylag, apparently exhausted but unharmed, swam for the shores of Gunna. Whilst scanning offshore at Crossapol, Isle of Tiree, on 19 December 1999, I noticed a winter plumaged’ Great Northern Diver Gavia immer flapping across the surface of the water then looking under the surface. Two Bottle-nosed Dolphins surfaced alongside the bird. The next 2 to 3 minutes were similar to what I had observed with the Greylag. The diver tried to escape, and I can only assume it had become too waterlogged to actually take off. Unlike the Greylag, the diver would lower its head into the water to look for the 52 Short Notes SB 21(1) dolphins and just before the dolphins got close the diver would flap across the surface of the water. Whilst the dolphins were nearby the diver did not attempt to dive. After a short period of time both dolphins continued south along the coast. The diver headed north, swimming hard then diving, surfacing 30-40 metres further up the coast. Although not seen to fly the diver did appear unharmed but exhausted by the experience. I have heard of seals catching birds from the surface of the water but not cetaceans playing with birds apparently just for fun. A J Leitch, Shepherd’s Cottage, Heylipol, Isle of Tiree, Argyll PA77 6TY Revised manuscript accepted February 2000 Common Crossbill eating adelgids On 25 July 1997 Icame across a group of Common Crossbills Loxia curvirostra (identified from a sonogram) feeding in a mature stand of Norway Spruce Picea abies trees at Farr Forest, Inverness- shire. While watching the birds for around 10 minutes, the only object which fell from the tree was not a cone but a fresh Pineapple Gall. This is a growth deformity found in spruce shoots caused by an adelgid (a type of bug, Hemiptera). The gall, which was still green had been freshly broken into and all the adelgids removed. Because there were no insects left on the specimen it was not possible to determine with confidence the species of adelgid. Possible species involved are Adelges abietus, A Remains of a Pineapple Gall on Norway Spruce eaten by Common Crossbill, Farr Forest, Inverness-shire. July 1997 Ron Summers Scottish Birds (2000) Short Notes oy) viridis or A laricis, with A viridis the most likely. It appeared that both the adelgids and the surrounding green plant tissue had been eaten by the Crossbill (Photo) as no parts of the discarded gall were found below the tree. Pulliainen (1972. Summer nutrition of crossbills Loxia pytyopsittacus, L curvirostra and L leucoptera in north eastern Lapland in 1971. Ann Zool Fennici9:28-31), working from crop contents of dead birds, records aphids, which are closely related to adelgids, as acomponent of the summer diet of both Common and Two-barred Crossbills L leucoptera in north eastern Lapland. Storer 1921 (American Crossbill eating Elm aphis, Condor 23:98) and Traverner (1922. Crossbills eating aphis Condor 24:96) both reported crossbills eating aphids in North America, and there are many other references to invertebrates being eaten (Cramp & Perrins 1994 Birds of the Western Palearctic Vol 8, Oxford). In many, however, it was not possible for the observers to differentiate the species involved because the birds consumed the evidence. In this instance the remains of the gall allowed identification with some confidence. Insect food at this time of year is likely to provide ~ animal proteins valuable for the production of new feathers during the moult. David C Jardine, 49 Bellfield Road, North Kessock, Inverness IVI 3XX Revised manuscript accepted February 2000 Oystercatchers usurping Lapwings’ nests Dougall (Scottish Birds 18:184 and 19:181) reported cases in both 1995 and 1996 where Oystercatchers Haematopus ostralegus had apparently taken over nests of Lapwings Vanellus vanellus at Colquhar, Borders. In the 1995 case the Oystercatcher hatched 3 Lapwing eggs and at least one survived to 12 days. During a study of Lapwings breeding around Newtonmore, Strathspey carried out by Pat French between 1996 and 1998, 627 nests were monitored. While none of the nests checked in 1996 or 1997 were affected, in 1998 3 of the 185 nests studied were taken over by Oystercatchers, suggesting that while unusual this is not necessarily uncommon. In each of the 3 cases found in 1998 the outcome was different. The first nest was found at Pitmain Farm (NN736999) on 22 April when an Oystercatcher was incubating aclutch of 3 Lapwing eggs at 1200 GMT. An hour later a Lapwing was seen sitting on the same nest and the following day at 1200 an Oystercatcher was again seen on the nest which then contained 4 eggs, with the agitated Lapwings standing at times within 5m of the sitting bird. Between the 24 and 26 April both of the pair of Oystercatchers and both Lapwings were seen on or near the nest and on one occasion the male Lapwing was seen to dive at the Oystercatcher which responded with a threat display towards one of the Lapwing pair before settling on the eggs. On 1 May the clutch was found to have been trampled and the eggs crushed by cattle. The second example was found on uncultivated ground beside Newtonmore Golf Course (NN718987). This nest was first found on 24 April when it contained 3 warm Lapwing eggs. When the nest was rechecked on 30 April in addition to the 3 Lapwing eggs it contained an 54 Short Notes SB 21(1) SSS a SS Oystercatcher egg; all 4 were cold. On 6 May the nest contained the 3 Lapwing eggs and 2 Oystercatcher eggs all of which were warm. One of the Lapwing eggs was missing on 15 May and later when the nest was observed at 0900 on 19 May all 4 of the remaining eggs were being incubated by an Oystercatcher. At 0900 on 27 May one Oystercatcher egg was chipping and the other had hatched. On 29 May 2 cold Lapwing eggs remained in the nest together with fragments of the hatched Oystercatcher eggshell. It is not known at what stage the third Lapwing nest was taken over by Oystercatchers. When found at 1700 on 19 May this nest at Nuide Farm (NN726984) contained a newly hatched Lapwing chick together with 2 eggs, one chipping and the other apparently unviable. At 1700 GMT on 20 May the nest contained 2 Lapwing chicks and the unhatched egg, all of which were being brooded by an Oystercatcher. Although the area contained a high density of both Lapwings and Oystercatchers, interspecific competition was not readily apparent. Acknowledgements Access to the land on which these nests were studied was freely given by the committee of Newtonmore Golf Club, Mr John and Mrs Eira Drysdale of Raliaand Mr Donald Munro of Pitmain Farm. Pat French died in 1999 and Hugh Insley is indebted to Mrs Ann French for making the records and diaries of his Lapwing study available. Pat French and Hugh Insley, 1 Drummond Place, Inverness IV2 4JT Manuscript accepted February 2000 Roof nesting Oystercatchers using a Herring Gull nest On 26 May 1998 we were checking a known site for roof nesting Oystercatchers Haematopus oestralegus at Cauldeen School, Inverness when we were surprised to find a Herring Gull Larus argentatus nest containing 2 eggs together with an Oystercatcher egg. A second Oystercatcher egg lay on the flat roof near the side of the gull nest, which was asubstantial structure used the previous year. We assumed ‘that what we were seeing represented an Oystercatcher clutch laid in an old Herring Gull nest which had been repossessed by the returning gulls. When we revisited the site on 18 June we found the nest occupied by a Herring Gull chick approximately 7 days old together with an unhatched egg, and an Oystercatcher chick about 3 to 4 days old, which jumped out of the nest cup on our approach and ran to hide elsewhere on the roof. There was no evidence of any other nest site on the roof. Since Birds of the Western Palearctic Volume 3 gives the incubation period for Herring Gull as 28 to 30 days and that for Oystercatcher as 24 to 27 days it was clear that the Oystercatcher had laid its clutch in the nest while it already contained one or both of the gull eggs. The age of the Oystercatcher chick also suggested that it had somehow been finding food. No adult Oystercatchers were seen at the site on either visit, so there was no evidence to suggest that they had been feeding the chick. Scottish Birds (2000) Short Notes DS) Because of the nature of the site it was not convenient to continue observations to discover whether the Oystercatcher. Weare grateful to Mr Bill Reid, janitor at Cauldeen School, who allowed us access to the site. Hugh Insley, 1 Drummond Place, Inverness IV2 4JT Jonathan C Brain, 11 Rathad Mhic Eoin, Balivanich, Isle of Benbecula HS7 5NG Manuscript accepted March 2000 Observations ona Lesser Whitethroat in a Dundee suburb in winter On Sunday 30 January 2000 at 1410hrs my attention was drawn to a warbler in the garden. I immediately identified it as a Lesser Whitethroat Sylvia curruca. Using 8 x 42 binoculars, at distances between 15 and 20 metres, I was able to get very good views. I have had experience of the species locally during the breeding season and ringed several juveniles between late June and early August. This individual was a very clean looking bird with a uniformly grey head and nape, no obvious rufous on the wings and distinct dark ‘eye mask’ on the ear coverts. The bird foraged in and around the flowers of a Mahonia bush. Observations lasted about 10 minutes before the bird left the garden. The bird reappeared about 20 minutes later and again its activities were centred on the flower heads of the Mahonia. Anunsuccessful attempt to photograph the bird resulted in it leaving the garden and it was not seen again that day. On 5 February at 0900hrs the bird was again observed feeding on the flower spikes of the Mahonia. At 1000hrs I walked to within 3m of the bird; shortly after it flew off into adjacent gardens. I have observed wintering Blackcaps Sylvia atricapilla foraging in a similar manner but Blackcaps tend to remove whole flowers. It 1s possible that foraging visits to Mahonia are associated with the sequential maturatuion of the flowers on this shrub. The bird was seen on 2 occasions before 1040hrs on 6 February but at no other time that day. On 9 February I had made some porridge using fine oatmeal and a little honey. The mixture was put out on the grass at O800hrs for the awaiting Blackbirds Turdus merula and Starlings Sturnus vulgaris. I observed the Lesser Whitethroat leave the /eylandii hedge, make its way along the top of the stone boundary wall and then fly down onto the grass, procure some of this porridge and immediately fly back to the cover of the hedge. This was the last observation of the Lesser Whitethroat, although! spent some time watching forit. Although Lesser Whitethroats are regularly recorded on the east coast in late autumn the latest records (Dundee and Angus Bird Report 1998) were of 2 singles on 3 October, both on the Angus coast. I know of no other records of this species during the winter, at least locally. Bruce Lynch, 27 Luke Place, Broughty Ferry, Dundee, DD5 3BN Revised manuscript accepted March 2000 56 Scottish Birds 21:56-59 SB 21(1) OBITUARIES Robert Marshall Craig Lambie 1935-1999 Big Bob, as he was known to his friends, was a man of many parts. Keen golfer, hillwalker, naturalist, artist, traveller and, above all, bird photographer and lecturer. Born in Carmunnock on 10 September 1935 he spent his life in this small village on the south side of Glasgow. Encouraged by his parents, he and his brother took an early interest in natural history, collecting birds’ eggs when it was still socially acceptable but denouncing it in later life as keen conservationists. He loved to walk in the hills, glens and islands of his beloved Scotland and knew intimately many of its wildest places. As a young man he joined the railways and spent the rest of his working life with them, taking full advantage of an early retirement deal when nearing 40 years service. Sport, particularly golf and football, had always played an important part in his active life. For 36 years he was a member of his local Cathkin Braes Golf Club, playing off 9 and in his retiral years played almost daily whenever weather and health permitted. To further his birdwatching interest he joined W K Richmond’s University of Glasgow class in the late 1960s. With Kenneth at the helm the group became known by the somewhat unflattering name ‘Richmond’s Raiders’, a title not wholly inaccurate. Eventually the University class disbanded and reformed as The Glasgow University Birdwatchers Club and in 1976 simply Glasgow Birdwatchers Club which continues to this day. Bob was a leading light within this group and also the SOC where he served on the Clyde branch committee and was for a short period its branch treasurer. By his own admission he was not a committee man, preferring the freedom of open spaces, particularly during the spring and summer. Bob was a good organiser and when in May 1988 he hatched up a plan to visit far off St Kilda I was invited to join his select group. A boat was chartered and a memorable trip ensued with much hilarity, some interesting meals and the exhilaration of reaching and exploring our goal. The return leg was very rough and most of the party were seasick; it was a dishevelled, wet and weary group which disembarked at Oban. Apart from numerous short trips to many parts of Scotland he caught the foreign travel bug and made visits to South Africa, Goa, Australia, New Zealand, Arizona, British Columbia and Nepal. Another of his talents, not generally known, was his ability as an artist. He produced good watercolours and some wood carvings which adorned his home. Above all else Bob will be remembered as a very good wildlife photographer. His camera was always to hand to record as much of the wildlife and scenery as he was able. I think I was instrumental in fostering his interest in bird photography some time in the early 1970s, but he soon far surpassed anything I could produce as he had an artistic eye and far more patience. This was to become a major pasttime and he spent countless hours in uncomfortable hides in all sorts of conditions and places at ungodly hours just to capture that “special picture’. As a result of all these pictures he was to become a very popular speaker on the lecture circuit and his superb images and self effacing commentary were much in demand. Visiting photographic clubs he would announce, ‘I present myself as a Scottish Birds (2000) Obituaries a7 naturalist’ but with a naturalist audience he would say, ‘ | am an ignorant photographer’. This was typical of the man, always modest and eager to learn from others. His photographs appeared in several books. The Sunday Post newspaper, The Scots Magazine and many other places. In recent years he won a major prize in an international photographic competition with a presentation at Vane Farm which gave him much pleasure and encouragement. One facet of his life which caused much amusement to his friends and family was the fact the Bob was always hungry! Many will remember his opening remark when returning to camp after many hours Bob Lambie in the field, ‘What’s fer eatin’, and stories abound about his prodigious appetite. Throughout his active life Bob was supported by his wife Una and they journeyed together to the far reaches of the globe in search of wild places and wildlife. When he became seriously ill a few years ago she nursed him back to apparent good health. Sadly, a further short spell of illness ended with his death in hospital at the age of 64. The “whistling big yin’ will be remembered for his quiet and friendly manner and thought for others. He was a good companion, always cheery, and he will be sorely missed by his family and friends. David Clugston Photographer unknown 58 Obituaries SB 21(1) Gerard Lionel Sandeman 1909-1999 People smile when they speak about Gerard Sandeman. ‘Courteous’, ‘kind and helpful’, ‘unassuming’ and ‘compassionate’ are words they use. They evoke a figure familiar to generations of birdwatchers in the Forth area, striding along a beach or crouched on areservoir bank, wrapped in mackintosh and hat. Austere, whiskery, craggy features, weatherbeaten by daily exposure to wind and rain, would often break into a grin anda laugh. He was deeply caring about people and passionate about recording birds. This passion began early for him and his brother Pat, first on teenage golfing holidays and later when they were dragooned by their cousin George Waterston into the Inverleith Field Club founded by 6 young former Edinburgh Academy pupils in 1929. Not that entry was immediate; Gerard, a Watsonian, attended meetings initially only as a guest. The study of birds in those prewar days was handicapped by the lack of good identification guides, and a benign view was still held of egg collecting. Members of the IFC would take a string held low between them over, say, the Red Moss of Balerno to put up sitting birds, take the eggs, and then make up nests out of moss and twigs in a chest of drawers at home. Having fun was the highest priority and many a boisterous weekend was spent in tents or bothy around the Highlands and Islands, birdwatching by day and piping, singing and story telling by night. The bagpipes, the violin and drums played a prominent part in Gerard’s life. The Midlothian Ornithological Club grew out of the IFC in 1933, limited to 10 members who were enthusiastic early ringers as well as recorders. Gerard confessed to being terrified by the ferocious keenness of some of the other members. It was during the 1930s that he started the series of records of his frequent trips to half a dozen areas in and around Edinburgh which marked his unique contribution to Scottish ornithology. Not for him dry, academic studies of bird biology or behaviour, or indeed the analysis or even publication of the very information he gathered, to the irritation of some people. His genius lay simply in observation, and noting and counting, and in the disciplined and meticulous recording process itself, combined with a great joy at the natural wonder of birds and experiencing the sight and sounds of them. The scale and detail of his records, now in the SOC archives, almost beggar belief: Count Registers of Leith Docks 1946-97, Penicuik to Hermitage 1933- 96, Aberlady and Threipmuir 1931-96, Isle of May and Gladhouse 1934-97, Joppa to Tyninghame 1931-98, and Lothian reservoirs 1931-97. And all these are backed up by 47 thick Bird Record Books 1931-98. On each of thousands of day trips every single species seen was counted and entered neatly in notebook and cross referenced even more neatly in register: certainly hundreds of thousands, perhaps more than a million observations. Nor did this relentless avian scrutiny diminish when he was off to the Second World War, serving in North Africa and Italy, having been commissioned in the Territorial Royal Artillery. Further notebooks record all his sightings, (kit bag laden with Coward 3 volumes, Dresser 2 volumes, and the unillustrated The Birds of Europe and North Africa by Colonel Wardlaw Ramsay). Observations included 3 subspecies of Great Grey Shrike (subspecies were a great preoccupation of everyone in those days), and his single views of Slender-billed Gull and White-spotted Bluethroat, together with innumerable columns of commoner species. After the War he and his brother Pat joined forces with their elder brother Norman in the Sandeman family wine and spirits business. This operated Scottish Birds (2000) Obituaries 59 separately from, but in association with, the more widely known London based Sandeman port and sherry firm, for whom they acted as agents in Scotland. While Pat looked after sales, Gerard took up whisky blending and became one of the leading exponents in the business. Working with up to 30 single whiskies, malts from Islay, lowlands and highlands, and grain spirits, he had the exacting task of producing consistent, high quality blends, at which he excelled. The 2 Sandeman firms eventually combined, Gerard becoming Managing Director ofits Scottish operation until he, a little thankfully one suspects, took early retirement in 1969 at the crest of his company’s success. Slightly diffident by nature, he found ornithology somewhat more congenial than the hard nosed business of the drinks industry. Now he was free to fillup more of those notebooks, travelling often 6 days a week round his familiar haunts, his piercing gaze focused through an old pair of binoculars, never a telescope. For some years he was frequently accompanied on these daily jaunts by Bill Harper, the SOC Honorary Librarian, and following these excursions he was a regular attender at the Wednesday evening discussion groups at 21 Regent Terrace. There he would take up his usual place hunched in the maroon armchair in a corner of the Waterston Library and quietly reveal more ornithological activity during the previous month than any of the other and younger people present. Retirement is a great thing when you know how to fill it. Most years he would spend some days at the Isle of May Bird Observatory, where he had been one of the pioneers in the 1930s, but travelling over always on a Monday rather than on the regular Saturday sailing to ensure his presence at home on the Sabbath. At the centre of his life was his deep religious conviction as a committed member of the Glasites, a Christian sect founded in 1730 by a Church of Scotland minister John Glas, deposed from his ministry during the continuing feuds over the secular content of the covenants. It placed emphasis on Christian life and love rather than dogma, and practised its weekly devotions without clergy. Services at its Meeting House were taken by an elder, such as Gerard, who also led the singing of the psalms: no hymns, no organ. John Glas’s son in law Robert Sandeman took this form of worship to London and then America, where its adherents became known as Sandemanians. His abiding faith sustained Gerard after the sad death of his first wife, Isobel, and later buoyed him throughout his long and happy marriage to Peggy, with their 3 children. He cared about people. When you talked to him you felt that his whole attention was concentrated in a direct, genuine concern for you. He was a friendly, able, conscientious, old fashioned, gentle, man. John Arnott Gerard Sandeman Judith Sandeman 60 Advice to contributors SB 21(1) Advice to contributors Authors should bear in mind that only a small proportion of the Scottish Birds readership are scientists, and should aim to present their material concisely, interestingly and clearly. Unfamiliar technical terms and symbols should be avoided wherever possible and, if deemed essential, should be explained. Supporting statistics should be kept to a minimum. All papers and short notes are accepted on the understanding that they have not been offered for publication elsewhere and that they will be subject to editing. Papers will be acknowledged on receipt and are normally reviewed by at least 2 members of the editorial panel and, in most cases, also by an independent referee. They will normally be published in order of acceptance of fully revised manuscripts. The editor will be happy to advise authors on the preparation of papers. Reference should be made to the most recent issues of Scottish Birds for guidance on style of presentation, use of capitals, form of references, etc. Papers should be typed on one side of the paper only, double spaced and with wide margins and of good quality; 2 copies are required and the author should also retain one. Weare also happy to accept papers on disc; or by email at, mail@the-soc.org.uk however, please state the type of word processing package used. If at all possible use Microsoft Windows 97. Contact Syivia Laing on 0131 556 6042 for further information. Headings should not be underlined, nor typed entirely in capitals. Scientific names in italics should normally follow the first text reference to each species unless all can be incorporated into a table. Names of birds should follow the official Scottish list (Scottish Birds 1994 Vol 17:146- 159). Only single quotation marks should be used throughout. Numbers should be written as numerals except for one and the start of sentences. Avoid hyphens except where essential eg in bird names. Dates should be written:...on 5 August 199]......but not on the 5th (if the name of the month does not follow). Please do not use headers, footers and page numbers. Please note that papers shorter than c700 words will be treated as short notes, where all references should be incorporated into the text, and not listed at the end, as in full papers. Tables, maps and diagrams should be designed to fit either a single column or the full page width. Tables should be self explanatory and headings should be keptas simple as possible, with footnotes used to provide extra details where necessary. Each table, graph or map should be on a seperate sheet, and if on disc each table, graph, map etc should be on a seperate document. Please do not insert tables, graphs and maps in the same document as the text. Maps and diagrams should be either good quality computer print out and in black and white (please do not use greyscale shading) or drawn in black ink , but suitable for reduction from their original size. Contact Sylvia Laing on 0131 556 6042 for further details of how best to lay out tables, graphs, maps etc. , 3 : = all se = 4 , = —— rf \ 7 ae Scottish Birds Volume 21 Part 1 June 2000 Contents Main papers Amendments to the Scottish List. Ronald W Forrester. SBRC Breeding birds of the Isle of May, Firth of Forth, 1972-99. M P Harris, S Wanless, I Darling & C Gallacher Wintering seafowl in Scapa Flow, Orkney, October 1998 to March 1999. EJ Williams Greylag Geese breeding in Shetland. M G Pennington East Ross Rookeries in 1998-99. Henry McGhie Short notes Breeding success of Merlins in Galloway. R C Dickson Mute Swans rearing Greylag Geese. Gavin Paterson A new record of successful tree nesting Peregrines. Fiona Leckie & Steve Campbell Post breeding congregations of adult Black-throated Divers in Wester Ross. Chris Donald Swifts in Speyside. Harvey Burton Mallard eating Eel. Tom Dougall Jackdaw killing and carrying newly fledged Blackbird in its feet. RC &A P Dickson Merlins selectively preying on Chaffinches. RC Dickson Ringed Plover using covered nest site. C J Booth Interaction between Greylag Goose, Great Northern Diver and Bottle-nosed Dolphin. A J Leitch Common Crossbill eating adelgids. David C Jardine Oystercatchers usurping Lapwings’ nest. Pat French & Hugh Insley Roof nesting Oystercatchers using a Herring Gull nest. Hugh Insley & Jonathan C Brain Observations on a Lesser Whitethroat in a Dundee suburb in winter. Bruce Lynch Obituaries Robert Marshall Craig Lambie. D Clugston Gerard Lionel Sandeman. J Arnott Front Cover Puffins, Fidra, Firth of Forth. Jan Andrews Advice to contributors Published by the Scottish Ornithologists’ Club, 21 Regent Terrace, Edinburgh EH7 5BT. © 2000 ISSN 0036 9144 60 JAN 3.0 200i \) aa Decline in House Sparrows in central Edinburgh nmer status and distribution of Greylag Geese in north and west Scotland Storm Petrels on Priest Island Egg sizes of crossbills in Scotland Density and habitat associations of Barn Owls in East Ross Breeding success of Lapwings in part of Strathspey Wintering Ring Ouzels in the High Atlas Mountains, Morocco Scottish Birds The Journal of the Scottish Ornithologists’ Club Editor: Dr S da Prato Assisted by: Dr I Bainbridge, Professor D Jenkins, Dr M Marquiss, Dr J B Nelson, and R Swann Business Editor: The Secretary SOC, 21 Regent Terrace Edinburgh EH7 5BT (tel 0131-556 6042, fax 0131 558 9947, email mail @the-soc.org.uk). Scottish Birds, the official journal of the Scottish Ornithologists’ Club, publishes original material relating to ornithology in Scotland. Papers and notes should be sent to The Editor, Scottish Birds, 21 Regent Terrace, Edinburgh EH7 5BT. Two issues of Scottish Birds are published each year, in June and in December. Scottish Birds 1s issued free to members of the Scottish Ornithologists’ Club, who also receive the quarterly newsletter Scottish Bird News, the annual Scottish Bird Report and the annual Raptor round up. These are available to Institutions at a subscription rate (1997) of £36. The Scottish Ornithologists’ Club was formed in 1936 to encourage all aspects of ornithology in Scotland. It has local branches which meet in Aberdeen, Ayr, the Borders, Dumfries, Dundee, Edinburgh, Glasgow, Inverness, New Galloway, Orkney, St Andrews, Stirling, Stranraer and Thurso, each with its own programme of field meetings and winter lectures. The Waterston Library at the Club’s headquarters at 21 Regent Terrace, Edinburgh EH7 5BT is the most comprehensive ornithological library in Scotland, and is available for reference during office hours (Monday to Friday, 0930-1630 hrs but please phone beforehand). Check out our website for more information about the SOC and other bird related organisations: www.the-soc.org.uk SOC annual membership subscription rates Direct Debit Other methods Adult £18.00 £20.00 Family (2 adults and any children under 18) living at one address £27.00 £30.00 Junior (under 18, or student under 25) £7.00 £8.00 Pensioner/unwaged £10.00 £11.00 Pensioner Family (2 adults living at one address) £14.50 £16.00 Life £360.00 £400.00 Life Family £540.00 £600.00 All subscriptions may be paid by Direct Debit and Covenanted. Subscriptions paid by Direct Debit greatly assist the Club. Please ask for a Direct Debit form by phoning the Secretary at the above address. Published by the Scottish Ornithologists’ Club, 21 Regent Terrace, Edinburgh EH7 SBT Printed by Meigle Print, Block 11, Units 1 & 2, Tweedbank Industrial Estate, Galashiels TD1 3RS Scottish Birds (2000) 21:61-68 61 A major decline in House Sparrows in central Edinburgh HE M DOTT & A W BROWN A recent British national decrease in House Sparrows is now recognised, though evidence comes mainly from non urban census plots in southern Britain. Counts presented here show that House Sparrows in the centre of Edinburgh have become about 10 times less numerous over a recent 15 year period, and have contracted in distribution, while other bird species in the same place have shown no such change. Possible causes are discussed but no specific reason for the decrease has been determined. Introduction The House Sparrow Passer domesticus increased greatly with the spread of human population (Cramp & Simmons 1994) though overall numbers of the species in Britain reached their highest level about the early 1970s and then a slight decrease began (Summers-Smith 1988). From 1978-88 the numbers visiting suburban gardens in Britain decreased by 15-20% (Marchant etal 1990). Recent Statistics reveal that the decline of numbers in Britain has continued since the early 1980s (Crick etal 1998, Glue 1998), and the New Atlas (Gibbons etal 1993) shows acontraction of range especially in Scotland and Ireland. However, these national trends are largely based on census data derived from farmland and woodland plots, the majority of which are in southern Britain; birds of urban and suburban habitats have not been adequately monitored (Balmer & Marchant 1993, Crick 1998, Summers- Smith 1999b). Local surveys have also shown declines but these too have been mainly in south Britain (Marchant et al 1990, Summers-Smith 1993, Easterbrook 1999) and little information has been published for Scotland and northern England (da Prato 1989, Summers-Smith 1999a). This paper presents evidence of a serious decline of House Sparrows in the centre of Edinburgh, Scotland, which has been referred to in Murray et al (1998). Study areas The ‘Core Study Area’ is Princes Street Gardens, in the centre of the City of Edinburgh. Princes Street Gardens cover c20 ha (0.2km7’) and consist of areas of mown grass, sloped rough grass, asphalted walkways, akeeper’s house, bandstand buildings, plentiful mature and young trees, many evergreen and deciduous shrubs, rose bushes, perennial herbs and annual bedding plants. A railway runs through the Gardens in the lowest ground and a street, The Mound, divides the Gardens into West Princes Street Gardens and East Princes Street Gardens (see Figure 1). Adjoining the Core Study Area are further ‘green’ areas comprising the grounds of 2 churches to the west, Edinburgh Castle and crags to the south, private gardens at the southeast corner of West Princes Street Gardens, and gardens of a bank at the south edge of East Princes Street Gardens (Figure 1). None of these ‘green’ areas held House Sparrows in 1997-99. Apart from these areas, Princes Street Gardens are surrounded by busy streets, tall commercial buildings anda large railway station. In terms of bird habitats, these conditions including the layout within Princes Street Gardens have notaltered significantly since the 1950s or earlier. The ‘Wider Study Area’ covers a more extensive c350 ha (3.5km?) part of central Edinburgh and 62 HEM Dott & A W Brown SB 21(2) Figure 1 The Core Study Area: Princes Street Gardens in central Edinburgh and some surrounding streets and features. a 8 8 & boundary of Princes Street Gardens = Core Study Area (shaded), other boundaries, — p private gardens. Square auye surrounds the Core Study Area. It comprises mainly urban habitat of old and recent commercial and administrative buildings, 4-5 storey tenement housing of 100-200 years old with small rear communal gardens or yards, and other features such as some municipal squares and gardens and the open spaces of Calton Hill and the north half of The Meadows (Figure 2). Methods House Sparrows were searched for and counted in the whole of Princes Street Gardens (the Core Study Area) in both the early 1980s and in the late 1990s. In the winters of 1982-84, AWB counted House Sparrows and other birds up to 4 times per month. Over counting could have occurred due to double counting of mobile birds, and under counting due to the presence of concealed birds or birds which may have been on buildings facing om Edinburgh Ay Castle Johnston Te Streets or — railway, b bank gardens, ch church grounds, k keeper’s house, Surasace sacs seve seve ma onto the Gardens, although AWB believes none of these factors were significant. These counts were done near mid day in variable weather conditions. In the 1990s HEMD noticed that a major change in numbers appeared to have occurred, and during most of 1997-99 he counted House Sparrows at least once every month throughout Princes Street Gardens with subsidiary observations on many additional dates. These counts were made in variable weather but not in very wet, windy or misty conditions, between 0800 and 1800 GMT though better counts were usually obtained between 0900 and 1500 GMT. Counts were of birds seen and heard. Double counting of individual birds was negligible as movements of the small numbers present could be watched, and it is believed that no birds were on buildings facing onto the Gardens during most of the counts. Under counting must have occurred on some days due to birds remaining Scottish Birds (2000) A major decline of House Sparrows in central Edinburgh 63 Figure 2 The Wider Study Area in central Edinburgh, surrounding the Core Study Area (shaded). Only selected streets are shown. streets or other boundaries, e@ eee e@ locations (1-10) where House Sparrows occurred February 1998 - July 1999. meme ome [jit of area within which no House Sparrows were found February 1998 - July 1999 except at locatins 1-10. HP Holyrood Palace, 1 West Princes Street Gardens, 2 Rutland Square-Atholl Crescent, 3 Dewar Place-St David’s Terrace, 4 Chuckie Pend, 5 Tollcross School, 6 Lauriston Gardens, 7 Meadow Lane-Boroughloch, 8 N Gray’s Close-Chalmer’s Close, 9 S Gray’s Close-Blackfriars Street, 10 Campbell’s Close. The *°002 © seme 2 cme 5 concealed and silent, though bread thrown by the counter or other people helped to draw birds from cover On some occasions. Counts presented are the highest obtained in each month, thus minimising any problem of under counting. Less regular counts were made of all other bird species as well. In the Wider Study Area in 1997-1999 HEMD observed House Sparrows and other bird species throughout the year, so that all parts of the Area were visited at least on a few occasions and some on many occasions. These observations aimed to detect presence or absence and approximate Meadows A numbers of House Sparrows and other species. Results The counts of House Sparrows recorded in West and East Princes Street Gardens, the Core Study Area, are presented in Table 1. A large decrease in numbers between the early 1980s and the late 1990s is immediately obvious. In the 1982-84 period, the counts on 8 and 15 Dec 1982, and 16 Feb 1983 are among the lowest, and more rain occurred on these dates than the others. 64 HEM Dott & A W Brown SB 21(2) Table 1 Numbers of House Sparrows counted in Princes Street Gardens, Edinburgh - the Core Study Area. A. All counts made in 1982-84. Date West Princes East Princes Street Gardens Street Gardens Total 8 Dec 1982 2D 26 148 15 Dec1982 a2 16 168 22 Dec 1982 164 333} 197 7 Jan 1983 papi) 102 Sy 20 Jan 1983 156 90 246 26 Jan 1983 182 32 234 2 Feb 1983 2 85 DSi 10 Feb 1983 95 62 Sy7/ 16 Feb 1983 99 27 126 25 Feb 1983 333) 47 380 4 Mar 1983 230 24 254 5 Jan 1984 139 39 178 B. Highest counts obtained in each month in 1997-99. Date West Princes East Princes Date West Princes East Princes Street Gardens Street Gardens Street Gardens Street Gardens Mar 1997 10 0 Jun 1998 19 0 Apr 1997 16 0 Jul 1998 28 0 May 1997 16 0 Aug 1998 WE 0 Jun 1997 8 0) Sep 1998 19 0) Jul 1997 2D 0) Oct 1998 7 0 Aug 1997 38 0 Nov 1998 14 0 Sep 1997 35 0 Dec 1998 16 0 Oct 1997 9 0 Jan 1999 10 0 Nov 1997 WS} 0 Feb 1999 16 0 Dec 1997 Di 0 Mar 1999 16 0 Jan 1998 29 0 Apr 1999 12 0 Feb 1998 i17/ 0 May 1999 9 g) Mar 1998 10 0 Jun 1999 26 0 Apr 1998 5) 0 Jul 1999 20+ 0 May 1998 1 0 The mean of the 1980s counts excluding those 3is likely to have been the true population in Princes 248, suggesting that a minimum of c250 birds is Street Gardens in the winters of the early 1980s. Scottish Birds (2000) In the 1997-99 period, the highest counts obtained in most months are probably a good indication of the true numbers, as already indicated. The 1990s numbers show evidence of seasonal variation. In July-August 1997, June-July 1998 and May-July 1999 up to c12 juveniles, judged by traces of yellow gape, were noted, so the higher counts obtained during May to September each year included young birds. The June1997 count was late at 1645-1730h which probably accounts for its low figure. Fairly high numbers appeared to continue during the 1997-98 winter until at least January, after which numbers declined until the following summer period of new recruits. This may contrast with 1982-83 when there were substantial counts in January and February, though whether any seasonal pattern occurred then is uncertain as counts were not continued. These data show that from at least 250 House Sparrows in Princes Street Gardens in the winters of 1982-84, numbers declined to c15-30 birds in the winters of 1997-99; a level just 10% of those present 15 years earlier. Numbers were slightly higher in summers than in winters in 1997-99 including newly fledged juveniles. The distribution of House Sparrows in Princes Street Gardens also diminished. In the early 1980s they occurred throughout the West and East Gardens, although they were most numerous along the principal walkway with seats and refuse containers. During 1997-99 House Sparrows were never seen in East Princes Street Gardens except on one day on 13 May 1999, when 3 were present but not seen on days before or after. In the West Gardens their range had contracted to the close vicinity of the keeper’s house (Figure 1), where they entered under the roof eaves of this building to roost and nest, and where they made much use of evergreen bushes. Occasionally, birds moved further along the main walkway from the keeper’s house but were never seen elsewhere A major decline of House Sparrows in central Edinburgh 65 inthe Gardens. Onrare occasions they perched on and called from buildings in Princes Street overlooking the keeper’s house, usually in early morning before traffic noise reached its high daytime level. In 1999 one pair nested in an air vent on one of these buildings; otherwise the only known nests were in the eaves of the keeper’s house. On 2 occasions House Sparrows were noted moving out of the Gardens. On 7 October 1997 one bird separated from a group of 7 and flew high southeast over The Mound and beyond. On 8 January 1998, of 4 birds which flew up, 3 returned to the hedge below while the fourth flew high westwards and disappeared from sight over buildings near the west end of Princes Street. Whilst House Sparrows have declined greatly in Princes Street Gardens there is no evidence that other common species of birds there have decreased. Wren Troglodytes troglodytes, Dunnock Prunella modularis, Robin Erithacus rubecula, Blackbird Turdus merula, Song Thrush T philomelos, Blue Tit Parus caeruleus, Great Tit P major, Carrion Crow Corvus corone, Starling Sturnus vulgaris, Caffinch Fringilla coelebs and Greenfinch Carduelis chloris all bred in the Gardens in 1982-84 and 1997-99 and some additional species may have also. Both authors’ observations indicate that the abundance and distribution of these species and others were broadly similar in the 1970s-80s and the late 1990s. Feral Pigeons Columba livia, which feed inthe Gardens and breed on surrounding buildings, were estimated as up to 560 birds in 1982-83 and up to 340 birds in 1997-99, although counts varied from month to month. Apart from the recent appearance of Lesser Black-backed Gulls Larus fuscus breeding on overlooking buildings (Dott 1996), there has been no obvious trend of change in numbers of any common breeding birds in Princes Street Gardens over the last 15-20 years, except for the 10 fold decrease in House Sparrows. 66 HEM Dott & A W Brown SB 21(2) In the Wider Study Area House Sparrows were largely absent. They were present only at locations mostly isolated from each other by considerable distances, shown in Figure 2. These locations held only small numbers of House Sparrows ranging from 2-7 birds per site, except for one site, abandoned overgrown rear gardens between North Gray’s Close and Chalmer’s Close, with cl6. The total for the whole Wider Study Area was c45. A group not included in the above was c5 House Sparrows seen between Johnstone Terrace and Grassmarket on 2 January 1998 which apparently moved or disappeared, as none was found there on several searches in 1998-99. Earlier information for the Wider Study Area is lacking, although AWB has records of up to 9 House Sparrows at Queen Street Gardens and St Andrew Square in winter 1982-83 where none was seen in many visits in 1997-99, and the authors and others are in no doubt that Sparrows were more numerous and widespread in the Wider Study Area in the 1960s-80s than now. Discussion Itis not obvious why House Sparrows should have decreased drastically in central Edinburgh. In Princes Street Gardens there has been an increase in evergreen and broadleaved shrubs, and habitat for many birds may have marginally improved as these have matured over the last ten years. Pesticide use is minimal and has decreased; bark mulch is now used to suppress weeds and herbicides are no longer used in the Gardens except occasionally along concrete edges (City of Edinburgh Council pers comm). On account of these factors alone, it would be unlikely that invertebrates are less available to birds now than before the Sparrow decline. However, the general level of city centre pollution remains an uninvestigated potential cause of invertebrate decrease, and measurements of vehicle effluent gases show that the highest levels in Edinburgh occur in city centre streets near Princes Street Gardens and that these levels are well above national average figures (City of Edinburgh Council 1999). People discard scraps and deliberately feed birds in Princes Street Gardens to an apparently similar extent now as in recent decades. It could be that maintenance of central city buildings has improved so that fewer holes or crevices are available for Sparrows to nest in. However maintenance staff of at least one large property overlooking Princes Street Gardens do not recall that Sparrows were ever a problem, in contrast to Feral Pigeons which have necessitated preventative measures for many years. In the Wider Study Area there is similarly no indication that the habitat has become less suitable for Sparrows; on old and new buildings there are abundant ledges, ventilators, external pipes, rain gutters and crevices on which Feral Pigeons and Starlings bred in 1997-99 throughout the Area, suggesting that potential nesting opportunities for House Sparrows must be widespread. Also Chaffinches, Robins, Blackbirds, Dunnocks and Blue Tits held territories widely through the Wider Study Areain heavily built up places with minimal plant growth, indicating that these small passerines could find nest sites and food to attempt to rear young in many urban sites where House Sparrows were absent. Predation on Princes Street Gardens House Sparrows seems to be slight or insignificant. Sparrowhawks Accipiter nisus were infrequent visitors to the Gardens in 1997-99, and small birds other than House Sparrows showed no apparent decrease in numbers. Signs of avian predators were rarely found; Feral Pigeon remains were seen 4 times in 1997-99. Tawny Owls Strix aluco were apparently absent. Domestic cats Felis catus were never seen in 1997-99 by the authors in the Gardens or surrounding streets, and Gardens staff know only of one cat in the church grounds west of the Gardens that is fed by people. Foxes Vulpes vulpes have increased greatly in Edinburgh in the last 2 decades though no indications of their Scottish Birds (2000) presence in the Gardens were noted in 1997-99. Up toc15 Grey Squirrels Sciurus carolinensis live in the Gardens and up to 2 Brown Rats Rattus norvegicus were seen occasionally, but itis highly unlikely that any of these mammals would cause Sparrows and not other birds to decline. House Sparrow distribution is not known to correlate negatively with that of other birds in Britain though this is the case in suburban habitat in Australia (Woodall 1996). Feral Pigeons can be presumed not to have affected Sparrow numbers through competition in central Edinburgh as their numbers have probably decreased (see above) and Summers-Smith (1999a) also concludes that Pigeons are not generally a cause of Sparrow decline. The recently arrived Lesser Black-backed Gulls in Princes Street (Dott 1996) do not seem to impinge on the life of the local Sparrows in any way. It has been shown that proximity to highway traffic noise can reduce the density of breeding songbirds up to a distance of several hundred metres (Reijnen et al 1995). In the Gardens close to Princes Street traffic noise is highly intrusive to the human ear and it could be that the vocalisations of House Sparrows are less effective at communicating through such noise than those of other birds. The House Sparrow decline in central Edinburgh could be a result of a more general decline. A recent analysis of national data shows a reduced survival rate of first year Sparrows during 1976- 94 when the national population was in decline, compared with the rate during the previous period of stable population (Siriwardena et al 1998); however this analysis relates particularly to farmland and factors operating in urban habitats may be different. In Europe, studies have shown some urban Sparrow populations to be non sustainable and dependent upon immigrant surplus birds from neighbouring suburban populations A major decline of House Sparrows in central Edinburgh 67 (Heij & Moeliker 1990), and it would follow that if suburban Sparrows ceased to produce surplus then the urban Sparrows would decline (Summers- Smith 1999a). This is consistent with observations in Edinburgh. Throughout Edinburgh’s older suburbs of c50-100 years of age Sparrows have declined (Murray et al 1998). House Sparrows now have a discontinuous distribution through these older suburbs, and, on average, are resident in probably well under 50% of streets (HEMD pers obs). It is possible that the high winter numbers in Princes Street Gardens in 1982-84 included birds that had moved there from surrounding older suburbs and that the decline in the latter is reflected in the low numbers now present in central Edinburgh. In Edinburgh’s more recently built outer suburbs we know of no evidence cf change in House Sparrow numbers. In towns outside Edinburgh withrecent low rise housing, suchas at Musselburgh, Livingston and Tranent, House Sparrows presently appear common (HEMD pers obs). In Tranent there were c350 Sparrow “territories” per km? in the 1980s (da Prato 1989) and numbers visiting a bird table there have remained fairly similar from then until now (SRD da Prato pers comm). Density of House Sparrows in the Core Study Area dropped from c1250 birds per km? in 1982- 84 to 75-190 birds per km? in 1997-99. These levels are comparable with the highest and lowest densities previously known for House Sparrows in towns (Summers-Smith 1988). In the Wider Study Area in 1997-99 densities were as low as 20 birds per km’. As Sparrows were also found to be absent in many places beyond the limits of the 350 ha Wider Study Area, the actual House Sparrow density in an area greater than 3.5 km’ of the centre of Edinburgh was below 20 birds per km? in 1997- 99. The decline documented in this paper is far more drastic than national figures reveal or other studies suggest, although one recent piece of evidence froma Glasgow suburb (Summers-Smith 68 HEM Dott & A W Brown SB 21(2) 1999a) hints that there could be a serious decline in that city also. Acknowledgements Weextend our gratitude to P Vandome for allowing the use of acomputer and facilities for production of drafts and for comments on the text, and to G Gainey for help in production of the maps. We thank I Hay and J Hunter of the City of Edinburgh Council for information on management of Princes Street Gardens and monitoring of air pollution levels respectively. References Balmer D & Marchant J British Birds 86: 631-633. City of Edinburgh Council 1999. Review and assessment of air quality in the City of Edinburgh, Stages I] & 2. City of Edinburgh Council, Environmental & Consumer Services. Cramp D & Perrins C M (eds) 1994. The Birds of the Western Palearctic Vol VU. OUP, Oxford. Crick H 1998. The BTO’s Danger List: warning to Government about 140 species. BTO News 218: 16. Crick H Q P, Baillie S R, Balmer D E, Bashford R I, Beaven L P, Dudley C, Glue D E, Gregory R D, Marchant J H, Peach W J & Wilson A M_ 1998. Breeding birds in the Wider Countryside: their conservation status (1972-1996). BTO Research Report No 198. BTO, Thetford. Dott HEM 1996. Roof-nesting by gulls in Lothian, 1994. Lothian Bird Report 1994: 113-115. Easterbrook TG 1999. Population trends of wintering birds around Banbury, Oxfordshire, 1975-96. Bird Study 46: 16-24. 1993. The sparrows fall. Gibbons D W, Reid JB & Chapman R A (eds) 1993. The New Atlas of Breeding Birds in Britain and Ireland : 1988-1991. Poyser, London. Glue D 1998. Siskins star - but where are the House Sparrows? BTO News 218: 8-9. Heij C J & Moeliker C W 1990. Population dynamics of Dutch House Sparrows in urban, suburban and rural habitats. In: Pinowski J & Summers-Smith J D (eds) Granivorous Birds in the Agricultural Landscape. PWN, Warsaw. Marchant J H, Hudson R, Carter S P & Whittington P 1990. Population Trends in British Breeding Birds. BTO, Tring. Murray R D, Holling M, Dott HE M & Vandome P 1998. The Breeding Birds of South-east Scotland, a Tetrad Atlas 1988-1994, SOC, Edinburgh. da PratoS RD 1989. The breeding birds of some built -up areas in south-east Scotland. Scottish Birds 15: 170-177. Reijnen R, Foppen R, Ter Braak C & Thissen J 1995. The effects of car traffic on breeding bird populations in woodland. III Reduction of density in relation to the proximity of main roads. Journal of Applied Ecology S2ASi—2 02: Siriwardena G M, Baillie S R & Wilson J D 1998. Variation in the survival rates of some British Passerines with respect to their population trends on farmland. Bird Study 45: 276-292. Summers-Smith J D 1988. The Sparrows. Poyser, Calton. Summers-Smith J D 1993. House Sparrow. In: Gibbons et al, The New Atlas of Breeding Birds in Britain and Ireland. Poyser, London. Summers-Smith JD 1999a. Current status of the House Sparrow in Britain. British Wildlife August 1999: 381- 386. Summers-Smith D 1999b. Can the BTO tell us what is going on with the sparrows? BTO News 225: 8. Woodall P F 1996. Limits to the distribution of the House Sparrow Passer domesticus in suburban Brisbane, Australia. [bis 138: 337-340. Harry E M Dott, Stonecroft Cottage, Main Street, West Linton, EH46 7EE Allan W Brown, 61 Watts Gardens, Cupar, Fife, KY15 4UG Revised manuscript accepted April 2000 Scottish Birds (2000) 2]:69-77 69 The summer status and distribution of Greylag Geese in north and west Scotland C MITCHELL, D PATTERSON, P BOYER, P CUNNINGHAM, R McDONALD, E MEEK, J D OKILL & F SYMONDS The numbers and distribution of Greylag Geese were recorded in north and west Scotland in late August 1997. The census found a minimum of 10,000 geese, comprising 1258 adult pairs, 3220 goslings and 4264 non breeding adults. Principal concentrations were found on the Uists (3311), Coll & Tiree (2366), Sutherland (1262) and Orkney (1114). The current population estimate is discussed in relation to the historical status of Greylag Geese in this area of Scotland, and other groups of the same species in Britain. Introduction The Greylag Goose Anser anser used to breed in the wild in the East Anglian Fens, Lancashire, the Lake District and probably many other parts of Britain before the reed marshes and fens were reclaimed for agriculture in the 17-19th centuries (Owen et al 1986). By the early 20th century the species was restricted to north and west Scotland, but between 1930 and 1970, flocks were again established in many parts of the country, especially in west Galloway and in England (Atkinson- Willes 1963, Owen et al 1986) and many were derived from eggs or goslings from the native stock on the Uists. Most of the indigenous birds are now restricted to the Uists, Harris/Lewis, Coll/ Tiree, and the northern and westernmost areas of mainland Scotland and associated coastal islands. Greylag Geese also breed in Shetland and Orkney although proven breeding is a recent occurrence and the provenance of these birds is unknown. Atthe end of the 19th century, the Greylag Goose still bred in considerable numbers in Scotland in the Outer Hebrides, the northwest coast and in Caithness/Sutherland (Berry 1939) but not in Shetland and Orkney (Holloway 1996). It was, however, subjected to almost continuous persecution. For many years it appears that few nests escaped destruction, and birds were killed both in and out of season. A dramatic decrease in numbers and contraction in range began at the end of the 19 century and continued for the first 30 years of the 20" century. The chief causes were persecution by crofters, whose corn and oats the geese damaged, especially in autumn, and excessive sport shooting on estates. Increasing motor traffic, egg collecting and summer trout fishing on previously undisturbed lochs and, in the 1930s, an increase in the numbers of Great Black-backed Gulls Larus marinus which can kill broods of young geese, may also have contributed to the reduction in numbers. Certainly by 1920, the species had ceased to breed on North Uist, and Berry (1939) reported that ‘..in Scotland as a whole, the Greylag appears in danger of extinction as a breeding species..’. Clearly small pockets of Greylag Geese survived in the north and west of Scotland, and since the 1960s, numbers of geese have shown a period of gradual increase (Owen et al 1986, Thom 1986). Changes in legislation in Britain, beginning with the 1954 Protection of Birds Act, reduced the number of ways in which Greylag Geese could be taken or shot, and at the same time, a number of 70 C Mitchell et al SB 21(2) protected areas were established (eg Loch Druidibeg, South Uist). Also during the same period, Greylag Geese began to take advantage of the higher quality herbage available on improved grasslands. These changes occurred concurrently and the net effect has been to reduce winter mortality. The remnant groups of native Greylag Geese restricted to the very north and west of Scotland through persecution at the end of the last century became isolated, and in terms of numbers, probably reached a low point in the first half of the 20th century (Berry 1939). A small degree of dispersal from the remnant stock was sufficient to colonise either new areas, or possibly former breeding areas, albeit in a rather restricted band from Colonsay, along the west coast and probably as far north as the northern isles. On cultivated islands, recent increases in the numbers of Greylag Geese is thought to be partly due to greater breeding success and recruitment, resulting from an increase in the quality and quantity of improved pasture since the 1960s (Paterson 1991) and also reduced persecution during the close season. Recent attention paid to migratory populations of geese in Scotland (eg Scottish Office, 1996) exposed an apparent gap in our knowledge of the status and distribution of Scotland’s only native breeding goose. Hugh Boyd attempted to count breeding Greylag Geese in Scotland using an aerial survey in 1959. However, the sheer scale of the task in such remote areas, and the retiring nature of the families, precluded a full assessment being made. No real attempt at a coordinated census has been attempted since. Surveys of Greylag Geese in 1989-91 (Brown & Dick 1992) and in 1990 (Delany 1993) did not focus on the north and west of Scotland, and breeding survey fieldwork in 1988-1991 (Gibbons et al 1993) did not aim to establish numbers in post breeding flocks. This paper describes the results of a survey of post breeding Greylag Geese in late summer 1997. Methods Most Greylag Geese moult in July close to the breeding areas, and small flocks tend to gather in remote upland or secluded coastal areas. After the moult, the geese move often only short distances to agricultural feeding areas. Thus it seemed sensible to undertake a survey during the late summer when birds had moved away from more inaccessible areas. Counts of Greylag Geese took place during the last 2 weeks of August 1997. Most monitoring involved checking suitable post breeding habitat on foot or by car. Monitoring took place north and west of Glen Mor between Fort William and Inverness, although counts were also undertaken in west Argyll as far south as the Kintyre peninsula and on islands to the west. Local knowledge of the status and distribution of summer flocks was sought and used wherever possible. Casual observations were sought from birdwatchers and hill walkers through the publication of requests for information in appropriate magazines and newsletters. When a flock of geese was located effort was made to identify goslings (through plumage characters) in order to assess the proportion of young. Observers were asked to note the habitat type the geese were encountered on. Overall coverage was comprehensive with few, if any, seemingly suitable mainland areas missed. Large areas of north and west Scotland were unchecked due to their apparent unsuitability (essentially land higher than 200m, or coastal areas lacking suitable feeding areas). Many small offshore unihabited islands were not checked either, although where Greylag Geese were known to occur through a priori knowledge of local counters, every effort was made to check these sites. For example there are several hundred small Scottish Birds (2000) Summer status & distribution of Greylag Geese in north & west Scotland71 islands in the Sound of Harris where occasional pairs of Greylag Geese are known to breed. Yet it also known that by mid August the nesting islands are largely abandoned in favour of one or 2 large inhabited islands where the Greylags feed on managed grasslands. However, we are conscious that Greylag Geese in north and west Scotland can be found in remote areas, thus the results of the census must be treated as an absolute minimum. Results The total counted was a minimum of c10,000 birds, including approx 3320 young of the year. The distribution of Greylag Geese is shown in Figure | and Table 1. Principal concentrations were found on the Uists (3311), Coll & Tiree (2366), Sutherland (1262) and Orkney (1114). The distribution of Greylag Geese in August 1997 was largely concurrent with the breeding distribution reported for the 1988-91 Breeding Birds Survey (Gibbons et al 1993) and 58% of 10km squares that held breeding geese in 1988- 91, held post breeding flocks in 1997. Comparing the two distribution maps shows a shift away from remote upland areas to lower lying agricultural land, although the distances involved are relatively small. Greylag Geese were primarily encountered on agricultural land 55%, Table 2, with most favouring improved grasslands. However, many geese were also found close to natural wetlands. The average flock size was 62 (n=66 flocks), and the overall proportion of young recorded in sample flocks was 32.2% (n=1391 aged). The average brood size was 2.56 goslings (n=154 broods), and assuming that each brood was accompanied by 2 parents, this equates to approximately 1258 adult pairs and 4264 non breeding adults. The breeding success was similar to the long term average for the Uists (27%, 1987-93, Mitchell 1999). Mitchell (1999) noted that early breeding at arelatively low altitude and the absence of a long Figure 1 The distribution of Greylag Geese re- corded in north and west Scotland during late August 1997. Key to symbols ; a @ >1000 (Sea @ >500 oes? | @ >250 SAV @ >100 ° 1-100 migration may increase gosling survival compared to Greylag Geese breeding in Iceland forexample. Loch Loyal (Sutherland) is an important moulting site for non breeding geese from other parts of north Scotland (see below). Inner Hebrides Coll/Tiree The presence of Greylag Geese in summer on Coll/Tiree appears to be rather poorly documented. According to local crofters, the presence of geese in the summer has only been noticeable in the last 30-40 years (I McDonald pers comm) There appears to be no historical records of breeding prior to the early twentieth century, although after an increase during the 72 C Mitchell et al SB 21(2) Table 1 The numbers of Greylag Geese recorded in north and west Scotland in late August 1997. Inner Hebrides Coll 291 Tiree 2075 Colonsay 86 Treshnish Isles 104 Mull 154 Eigg, Small Isles 72 Muck, Small Isles 200 Canna, Small Isles 16 Outer Hebrides North Uist 1670 Benbecula 595 South Uist 1046 Harris/Lewis 268 Northwest Scotland Acharacle, Lochaber 40 Eilean Tioram, Lochaber 14 1940s post breeding flocks reached 50-100 birds (Owen etal 1986). In 1938, a pair bred on Coll for the first time in many years, and further records suggest numbers slowly increased from that time (Boyd 1958, Sharrock 1976). Winter numbers on Coll/Tiree appear to have increased from c670-920 individuals in 1985-87 (Stroud 1988) toc2900 in 1997 (c22% per annum, Figure 2). The combined (Coll/Tiree) count of 2366 in August 1997 is some 500 birds fewer than that recorded in the previous winter. With the addition of several hundred goslings hatched in 1997, this is somewhat surprising. Charlie Self, counting on Coll, noted that the number of geese he found was lower than he had expected. He had previously encountered 443 adults and c300 goslings earlierin the summer and had expected to find 700+ geese during the time of the census. The geese were probably still in the more remote areas of this island and thus led to an underestimate Kirton, Lochaber 94 Plockton, Lochaber 46 Isay, Skye 100 Achiltibuie, Wester Ross 40 Inversdale, Wester Ross 1g) Loch Carron, Wester Ross 56 Longa Island, Wester Ross 12 Mungasdale, Wester Ross 63 Tournaig, Wester Ross 56 Tournapress, Wester Ross 11 Caithness/Sutherland 1262 Northern Isles Orkney 1114 Mainland, Shetland 20 Unst, Shetland 100 Total (minimum) 9618 of the real number summering there. On Tiree, principal post breeding concentrations occured on Loch Rhiagain, Loch an Eilean, and Loch an Phuill. On Coll, large post breeding gatherings were found on Loch Cliad, Ballyhaugh, Loch nan Cinneachan, Loch Anlaimh and on the headlands around Crossopol Bay. Table 2 Habitat type of Greylag Geese flocks recorded in north and west Scotland in late August 1997 (sample size 3924). Improved grassland 37.4% Barley stubble 8.3% Unimproved grassland 5.2% Set aside 5.2% Natural loch 14.4% Saltings/mud/marsh 14.4% Sea loch 12.6% Scottish Birds (2000) Summer status & distribution of Greylag Geese in north & west Scotland73 Other islands Small numbers of Greylag Geese breed on other Hebridean islands (eg 25 adults and 7 young on Colonsay in 1995; 6 broods on Mull in 1995, Argyll Bird Report). The combination of secluded, Figure 2 Numbers of Greylag Geese recorded in winter on Coll & Tiree, 1982-1997. Data from annual Argyll Bird Reports and Alan Leitch, pers comm. 0 ,UUY Coll = 3 2, = : = Tiree x< u 2 82 84 86 88 90 92 94 96 undisturbed offshore islands, together with nearby grazing meadows favours small pockets of geese. On Islay, records of Greylag Geese in the summer have been few although the potential for colonisation appears to be only a matter of time. Outer Hebrides The Uists Records of Greylag Geese go back to the 18th century, when attempts were made by crofters to keep the geese away from their autumn cereal crop (McKay 1980). By 1920, some 200 pairs were still reckoned to breed on South Uist, but it seems the numbers had been even higher (Baxter & Rintoul 1953). Certainly by the same year the species had ceased to breed on North Uist. The reasons for the decline were partly the changes in land tenure and growing persecution by crofters and partly the overshoot on the estate in the years up to 1914. At that time, the geese on South Uist were afforded no close season, and large numbers were shot in March and April. The only legal protection was a ban on the taking of eggs (Owen et al 1986). Numbers appear to have reached a low point by the 1930s (Berry 1939) but from the 1940s onwards there seems to have been a noticeable increase. Counts of Greylag Geese were undertaken on the Uists in 1968-72, when c700-800 individuals were recorded (Newton & Kerbes 1974). Numbers increased to 1676-2000 by 1982 and the number of breeding pairs increased from about cl40 (Sharrock 1976) to c 200-300 over a similar period (Thom 1986). Regular counts on the Uists since 1986, suggest an increase from at least c 1600 birds in the mid 1980s (Paterson 1986) toc 3300 individuals in 1997 (cl2% per annum, Figure 3). Breeding occurs on coastal sites in heather Calluna vulgaris, grass/rush stands and low scrub. Principal post breeding concentrations occured on machair areas on the west of the archipelago at Balranald/Clettreval, Ath Mhor, Berneray, on Benbecula and several sites on South Uist. Harris/Lewis On Harris/Lewis a small dispersed group now numbers over 200 individuals. No systematic count during the summer had been carried out previously thus the rate of change cannot be estimated, however, itseems likely from previous observations that the numbers now are larger than in the last 30 years. During the Second World War and the following 20 years or so Greylag Geese were harassed by random and indiscriminate shooting of migrant geese on almost all estates but more so on Barvas, Galson and Stornoway Trust Land. Illegal, untimely and careless burning of moorland and heather clad islands on freshwater lochs destroyed nests. The 74 C Mitchell et al SB 21(2) Figure 3 Numbers of Greylag Geese recorded in late August on the Uists, 1986- 1997. Data from Mitchell (1999) and Roderick McDonald. 3,500 3,000 2,500 Count 2,000 1,500 1,000 86 87 88 89 90 91 92 93 94 95 96 97 Post breeding (August) census taking of eggs for rearing geese for Christmas was prevalent during the same period. The introduction of the American Mink Mustela vison in the mid fifties may also have led to some local decreases innumbers. Breeding seemed then to be confined to Eilean Mor on Loch Orosay, south of Stornoway, and even there local butchers were said to be taking flightless young for the pot. In Harris, the main breeding grounds are now on Lochs Steisevat and Moracha and associated waters behind Leverburgh and on some of the islands in the Sound of Harris, such as Pabbay, Killegray, Sleicham and Heisker. Families have been reported also from the east coast of south Harris at suitable lochs such as Plocrapool. In Lewis, a similar Anserine diaspora has taken place with a diminution of pressure on the local geese in winter and spring and Greylag Geese seem to be increasing in number. Families may now be found on many freshwater and sea lochs where they were hitherto unknown to the present generation. The widespread establishment of reseeded moorland, especially in Lewis, since the 1950s, initially intended for cattle but degraded by sheep has doubtless been to the Greylags advantage. An assessment of the breeding distribution has not been undertaken and is not considered practicable owing to the size and diversity of the area. Northwest Scotland There have been infrequent records of Greylag Geese inhabiting the coast of Wester Ross and Sutherland during the last 40 years. For example, Greylag Geese appear to have been present in very small numbers on the Summer Isles and Skye (H Boyd pers comm, Holloway 1996), with breeding occurring on several offshore islands. In 1997, Greylag Geese were still thinly distributed, although flocks of up to 100 birds were noted in several places. Many of the coastal areas are unsuitable for supporting large flocks. Although low, offshore breeding islands are relatively plentiful, the availability of suitable lowland, undisturbed grazing areas, and a safe roosts is limited and this may contain future population growth. Caithness/Sutherland Due to the size and inaccessible nature of Caithness/Sutherland, summer counts have not achieved full coverage, although numbers there were thought to be stable at c 2200 for the last ten years (F Symonds, pers obs). Alan Wood recorded 1437 moulting Greylag Geese in July 1992 at Loch Loyal and, in 1994, Ian Stenhouse found 1100 Greylag Geese at 6 moulting lochs during a road transect, including 846 at Loch Loyal. In 1997, 896 Greylag Geese were recorded moulting at Loch Loyal. The count for Caithness/Sutherland in 1997 (1262 geese) therefore, appears low although this may be consistent with a reduction in the size of the moulting flock at Loch Loyal. Scottish Birds (2000) Summer status & distribution of Greylag Geese in north & west Scotland ™] Uy Northern Isles Orkney There are no historical records of Greylag Geese breeding or summering in either Orkney or Shetland (Holloway 1996, E Meek and D Okill. own data), and nesting was first recorded as recently as the mid 1980s (Pennington 2000). It seems likely that the current numbers summering there derive from birds originating from Sutherland, or possibly, small numbers of over summering Iceland Greylag Geese (see Pennington 2000 for a summary). In Orkney. these have been augmented with a few birds deliberately released for hunting (C Booth, in litt). Two Greylag Geese marked in Sutherland in July 1996 were recorded in Orkney in winter 1996/97, indicating that there may be some, albeit modest, connection between the mainland and Orkney. Summer counts suggested c 50 pairs in 1993/94 (and c200 non breeding birds) on Orkney and c200 birds in total on Shetland. The 1997 count of 1114 Greylag Geese on Orkney is the largest ever recorded and is probably not fully comprehensive in light of the fact that breeding is occurring on small. uninhabited islands which were not visited during the survey. Nesting is now taking place on moorland areas and around lochs on the Mainland, Shapinsay. Gairsay, Stronsay, Sanday, Burray, Copinsay and probably Rousay, as well as on a series of small islets, in the waters offshore from these larger islands. Post breeding concentrations occur on freshwater bodies close to the breeding sites, and in the case of the smaller islets, on the sea. The greatly increased wintering population of Greylag Geese in Orkney in recent years has begun to cause considerable worries amongst the local farming community and at least one farming organisation has called for action to be taken to reduce numbers. The presence now of considerable numbers outside the winter period is raising the degree of antagonism even further. Shetland Pennington (2000) documents the increase in numbers and range of breeding Greylag Geese since the 1980s and a concurrent increase in reseeded hillsides during the period of colonisation has evidently favoured the species. In light of recent. more thorough coverage it seems the count of 120 birds in 1997 was low. Pennington (2000) suggested the colonisation of Shetland was by Icelandic breeders short stopping yet without any evidence from ringed individuals there remains the possibility that the birds breeding here originated from the mainland Scotland stock, or even a mixture of the 2. Dispersal from breeding areas Ringing studies on the Uists, Coll and Tiree and in Sutherland confirm the rather sedentary nature of the Greylag Goose in north and west Scotland. For example. of 500 Greylag Geese ringed on North Uist only 7 have been recorded away from the Uists. On Coll/Tiree, over 400 Greylag Geese have been ringed in 1998-99 and none have been seen away from the islands. However, some minor movement between the offshore island and groups on the west coast of mainland Scotland was sufficient to promote and retain genetic mixing and, through dispersal. further the establishment of new breeding colonies. Of the 7 records of movements away from the Uists, 5 birds were recorded on Tiree. one went to the north coast of Lewis and another moved south to Colonsay. The use of individual marks (plastic leg rings and collars) has shown that while the majority of birds move very little. a small minority do move sufficiently to promote genetic mixing with other stock. It seems likely therefore, that the colonisation of Coll/Tiree in the middle of the 20" century resulted from birds moving southeast from the Uists. 76 C Mitchell et al SB 21(2) Recoveries and sightings of Greylag Geese marked at Loch Loyal (Sutherland) in 1996-97 have been reported from Orkney and along the west coast of Easter Ross suggesting that this site is an important moulting ground for birds from an area greater than the Sutherland breeding sites. A record of 800-1000 Greylag Geese on Muck on 12 September 1997 when the summer population was estimated at 200 birds appears too early to be Greylag Geese from Iceland and suggests a major post breeding movement, although it is not known where these birds may have originated from or what caused the movement. Discussion The range of this stock of Greylag Goose, for now, 1s still restricted to areas of north and west Scotland. This would enable the development of a conservation plan for these geese to guide national conservation and management actions, since this would involve relatively few organisations. For management purposes, the Greylag Geese breeding in north and west Scotland may be regarded as the remnants of the native stock (see Mitchell 1999). They are relatively sedentary although some minor movement between islands and mainland areas aids dispersal to new areas. However, the status of the native groups needs to be fully examined in light of the various reestablishment schemes carried out from the mid 1930s. Delany (1993) found approximately 2000 reestablished Greylag Geese in Scotland in areas to the south and east of the Great Glen (eg Perthshire). These are derived from reestablishment projects carried out by landowners, the Nature Conservancy, the Wildfowl Trust and various shooting clubs (Sedgwick 1975). Many of these reestablished birds derived from eggs or goslings from the native stock on the Uists (Sedgwick 1975, Atkinson- Willes 1963). The gradual spread of the reestablished Greylag Goose north has been coincidental with the recent increase in numbers and spread of the native Greylag Goose south. However it may be safely predicted that eventually the 2 stocks will interbreed and Greylag Geese will once again nest over much of Britain. Morphologically, there appears to be no difference in measurements between the two stocks (WWT unpublished data) and it might be argued that since much of the reestablished stock derived from native birds they are genetically comparable too. Perhaps higher land surrounding the Great Glen merely remains as a physical barrier between future integration. The 1997 survey was the first successful attempt to cover the whole range of the indigenous stock. However, future monitoring of the distribution and numbers of breeding Greylag Geese will probably necessitate a full survey of not only the north and west of Scotland but also of the rest of Britain in order to assess the extent of integration of the 2 stocks. The continuation of ringing programmes should help to monitor the progress and pace of integration. Acknowledgements We wish to thank the following for providing valuable advice about the status of Greylag Geese in their local areas: R Adam, D Andrews, W Bews, N Bielby, C Booth, R Broad, P Carty, J Chester, A Currie, T Dean, S Delany, C Donald, J Dye, R Evans, K Hague, P Hollindrake, D Jardine, A Knight, A Leitch, D Matson, J McCutcheon, C McKay, B Neath, M Ogilvie, M G Pennington, J Plowan, M Ramage, B Ribbands, North Ronaldsay Bird Observatory, A Rothwell, C Self, M Smith, P Stanley, P Strachan, R Swann, C Tiarks, R Wilson. Apologies for any omissions or names misspelt. Peter Cranswick, Malcolm Ogilvie and Tony Fox kindly provided valuable comments on an earlier draft. Goose monitoring in the UK has for many years been supported by the Joint Nature Conservation Committee. Scottish Birds (2000) Summer status & distribution of Greylag Geese in north & west Scotland77 References Atkinson-Willes Britain. HMSO. BaxterEV & Rintoul LJ 1953. The birds of Scotland. Edinburgh. Oliver & Boyd. Berry J 1939. The Status and Distribution of Wild Geese and Wild Duck in Scotland. University Press, Cambridge. Boyd J M 1958. The birds of Tiree and Coll. British Birds 51:41-56, 103-118. BrownAW & Dick G 1992. Distribution and number of feral Greylag Geese in Scotland. Scottish Birds 16:184-191. Delany S N 1993. Introduced and escaped geese in Britain in summer 1991. British Birds 86:591-599. Gibbons D G, Reid J B & Chapman R A 1993. The New Atlas of Breeding Birds in Britain and Ireland: 1988-1991. BTO/SOC/IWC. Poyser. Holloway S 1996. The Historical Atlas of Breeding Birds in Britain and Ireland, 1875-1900. Poyser. London. McKay MM 1980. The Rev John Walker’s report on the Hebrides of 1764 and 1771. Donald. Edinburgh Mitchell C 1999. Greylag Goose (Scotland). In Madsen J, Cracknell GS & Fox AD eds. Goose populations of the Western Palearctic - a review of the status and distribution. Wetlands International Publication G L 1963. Wildfowl in Great No. 48 Wetlands International, Wageningen, The Netherlands. National Environmental Research Institute, Ronde, Denmark, 344pp. Newton I & Kerbes RH 1974. Breeding of Greylag Geese Anser anser on the Outer Hebrides, Scotland. Journal of Animal Ecology 43:771-783. Owen M, Atkinson-Willes GL & Salmon D G 1986. Wildfowl in Great Britain. Second Edition, University Press, Cambridge. Paterson I 1986. The status and distribution of Greylag Geese Anser anser in the Uists, Scotland. Bird Study 34:235-238. Paterson I 1991. The status and breeding distribution of Greylag Geese Anser anser in the Uists (Scotland) and their impact upon crofting agriculture. Ardea 79:243-252. Pennington M G 2000. Greylag Geese Anser anser breeding in Shetland. Scottish Birds 21:27-35. Scottish Office. 1996. Wild Geese and Agriculture: a discussion document. Edinburgh, UK. Sedgwick N M, Whitaker P & Harrison J G 1970. The New Wildfowler in the 1970s. London, Barrie & Jenkins Sharrock J T R 1976. The Atlas of Breeding Birds in Britain and Ireland. Poyser, Berkhampstead. Stroud D A 1988. Breeding waterfowl on Coll & Tiree. In: Stroud, DA ed The Birds of Coll and Tiree: status, habits and conservation. NCC/SOC, Edinburgh. Thom V 1986. Birds in Scotland. Poyser, Calton. Carl Mitchell, WWT, Slimbridge, Gloucestershire., GL2 7BT (address for all correspondence) David Patterson, WWT, Eastpark Farm, Caerlaverock, Dumfries, DGI 4RS Paul Boyer, 96, Carnan, Eochar, Lochboisdale, South Uist, Western Isles, HS8 5OX Peter Cunningham, 10 Barony Square, Stornoway, Isle of Lewis, PA87 2TQ Roderick McDonald, Western Isles Council, Balivanich, Isle of Benbecula, Western Isles, HS7 5LA Eric Meek, RSPB Orkney Office, Smyril, Stenness, Stromness, Orkney, KW16 3JX David OKill, Heilinabretta, Tronda, Shetland, ZE1I 0XL Fraser Symonds, SNH, Main Street, Golspie, Sutherland, KW10 6TG Revised manuscript accepted May 2000 78 Scottish Birds (2000) 21:75-84 SB 21(2) A survey of Storm Petrels on Priest Island in 1999 P MAYHEW, K CHISHOLM, H INSLEY & N RATCLIFFE The Storm Petrel population of Priest Island, Wester Ross was censused in July 1999 using the tape playback method. Habitat specific response rates were calculated for the 4 main breeding habitats on the island: boulder beach, stone wall, scree, heath/ grassland. Response rates in different habitats showed significant variation - 0.36 to 0.48. Three of the habitats were completely censused while, due to the large area of this habitat, heath/grassland was surveyed by sampling densities in randomly placed quadrats. The population estimate was 4,400 (95% confidence limits = 3,300 - 6,100) Apparently Occupied Sites, putting Priest Island amongst the largest counted colonies in Scotland. Over half the birds were at a low density in the extensive heath/grassland habitat. The accuracy of the technique, its resource requirements and potential improvements are discussed. Introduction The difficulties of censusing Storm Petrels are well documented; they are completely nocturnal and nest in cavities or burrows on remote islands (Ratcliffe et al 1998b). However, the proportion of the world population breeding in Britain and Ireland (estimated at 51-65%, Lloyd et al 1991) makes the development of accurate and repeatable census techniques extremely important. Much work has been carried out in the last 5 years to develop census methods using diurnal playback of the males’ purr songs (Mainwood ef al 1997, Gilbert etal 1998, Ratcliffe etal 1998a,b, Vaughan & Gibbons 1998). This work culminated in the publication of an agreed method of Storm Petrel monitoring which should be robust and repeatable (Gilbert et al 1999). One of the most important aspects of this method is that response rates to tape playback are estimated separately for each survey and breeding habitat, rather than using generic response rates. This paper reports the results of a survey of Priest Island, Wester Ross in July 1999 using the above method. Priest Islandis an RSPB reserve, managed principally for its Storm Petrel colony and is a key site for these birds in north west Scotland. Previous population estimates have varied from 2,300 pairs (Mainwood et al 1997) to 10,000 pairs (Dennis 1976). The paper also develops a new analysis of response rate from calibration plots. The practical application of the technique, whichis the first time ithas been used to census akey colony, and its use where Apparently Occupied Sites (AOS) are dispersed at low density over a large area, are discussed. Priest Island is located in the coastal waters of the Minch and is 137.5ha in size. The island encompasses a wide range of habitat types but a wet heath/acidic grassland mosaic predominates. There are also significant cliffs and freshwater lochs and some small areas of remnant woodland. The island usually hosts 11 species of breeding seabird, though most populations are small. Storm Petrels breed in 4 habitats: heath/grassland, stone walls, scree, and boulder beach. Boulder beach and stone walls were easy to define and locate. Scree was less easy to locate because it was often covered in dense vegetation (eg heather, bracken). The heath/grassland habitat was by far the most Scottish Birds (2000) A survey of Storm Petrels on Priest Island in 1999 79 Figure 1 Map of Priest Island showing calibration plots and survey sites. © Loch e Scree + Stone wall Oo Boulder beach () Calibration area, scree, stone wall or boulder beach ry Heath/ grass sample areas oe Heath/ grass calibration area difficult to survey. Storm Petrels here were found to be breeding in cracks in peat under deep vegetation throughout the island. Responses to tape playback suggested that AOS were located, seemingly at random, throughout the whole of this habitat. We did not find the concept of ‘subcolonies’, as has been used in previous studies appropriate in describing the distribution of AOS in this habitat. Methods Calibration Plots Only a proportion of Storm Petrels will respond to the tape playback of the male’s purr song in a single visit so in order to calculate correction factors, calibration plots were established in the 4 habitat types following the methods described in Gilbert et al (1999). The areas chosen were typical of the habitat type while the size of each calibration plot was intended to give a minimum of 50 AOS by the end of the calibration period (7 days). However, this was not possible for boulder beach due to the small total number of birds in this habitat. Plots varied in size from 10’s of metres in stone wall to around 2ha in heath/grassland. The tape recorders used were 2 Sony Walkman WM-EX182 with Saisho SP10 speakers and 2 Aiwa HS-GS 194 with JVC A-210 speakers. The recorder and speakers were mounted on a wooden board for easier use in the field. The purr song was played for 10 seconds every 2 metres throughout the stone wall, scree and boulder beach calibration plots. Speakers were held within 0.5m of the wall or ground. Due to the scale of the heath/grassland habitat, the purr song was played approximately every 10 m (see also main survey methods). All responding burrows were marked with sticky tape or pegs as an AOS. Over 7 days the rate of discovery of new AOS had fallen to 5% or less (Gilbert et al 1999). The cumulative numbers of AOS were plotted and response rates were estimated using the analysis described in Appendix 1. Main Survey The methods used were again those outlined in Gilbert et al (1999). The census survey technique (ie acensus of the complete habitat area) was used for the stone wall, scree and boulder beach habitats. In stone walls, the purr call was played for 10 seconds every 2m. In screes and boulder beaches up to 4 fieldworkers walked in transects 2m apart and again played the purr call for 10 seconds every 2m. The scale of the heath/grassland habitat (94.68ha - measured taking account of the topography of the habitat) meant that it was not practical to survey the whole of this habitat. Instead, the sample survey technique was used in which approx. 20% of the habitat is surveyed. The habitat was divided into 100 one ha (100m x 100m) squares. Twenty squares were then chosen using random numbers. If a square had a significant (greater than SO P Mayhew, K Chisholm, H Insley & N Ratcliffe SB 21(2) 10%) area of water within it then the closest square without water was used instead. The squares were walked by 4 fieldworkers 6m apart, thus covering an approximate 25m transect in each sweep. Thus a one hectare square could be covered in 4 sweeps. The tape recorder was played every 10m for 10 seconds. Thus, the furthest from an AOS that a tape was played was 5m. While the intensity of tape playback was not as great as in the other habitats the time involved in a more intensive method would have been prohibitive. Ratcliffe et al (1998b) found a significant decline in response rate with lower volume of recording and also an indication (not significant) of reduced response with distance from AOS, though they only tested up to 2m. In this study responses were heard up to 10m from the tape recorder and it may be that the main limiting factor is the ability of the fieldworker to hear the response, particularly in a breeze. Thus, a criticism of this method is that it may produce an underestimate of AOS. Results Calibration Plots Figure 2 shows the cumulative number of AOS over time for the calibration plots within each habitat. The asymptote for each habitat was calculated to give the total number of AOS for each plot (see Appendix 1). Table 1 shows the response rates (with 95% confidence intervals for each habitat). The response rates in scree and boulder beach were significantly higher than those in walls and grassland. Figure 2. The cumulative total of Storm Petrel AOS detected in each habitat plotted agains visit. Points represent observed data, lines represent the fitted values derived from substituting the values in Table A into equation I (see Appendix 1). Circles and solid line = boulder beach; squares and dashed line - scree; triangles and dotted line = wall; diamonds and dotted/dashed line = heath and grassland. Cumulative number detected Visit number Main Survey Table 2 shows the total number of responses recorded for the 3 census survey habitats and the number of responses recorded in the heath/ grassland random squares. The extrapolated figure for the whole heath/grassland habitat has been calculated from the average density in the 20ha of sampled squares multiplied by the 94.68ha of this habitat on the island. Bootstrapping (Greenwood 1991, Ratcliffe et al 1998a) was used to calculate Table 1 Estimates of response rates (proportion of the asymptote detected on the first visit) with the upper and lower 95% confidence intervals (UCI, LCI). Habitat Response rate Boulder beach 0.472 Stone wall 0.418 Scree 0.475 Heath/grassland 0.359 LCI UCI 0.400 01535 0.379 0.454 0.448 0.500 0.273 0.435 Scottish Birds (2000) A survey of Storm Petrels on Priest Island in 1999 81 Table 2 Calculation of AOS from the main survey for the 4 breeding habitats plus the total population estimate for the island. extrapolation AOS (after 95% confidence 95% confidence Habitat total no to total correction for limits (after limits of responses habitat area responses rate) bootstrapping) correction factors Boulder beach 25 (25) Stone wall Re @2) Scree 679 (679) Heath/grassland 206 975 Total - WS the 95% confidence limits for the heath/grassland habitat. The number of AOS for each habitat has been calculated based on the response rates in Table 1. The heath/grassland and scree habitats together held 95% of the breeding population. The total population of Storm Petrels on Priest Island from this survey is 4,370 AOS (95% CL 3,338 - 6,069) Discussion Ratcliffe et al (1998b) reported that response rates could vary from 0.11 to 0.56 depending on a variety of variables. They recommended that a colony specific correction factor (ie measured at the colony that year but not taking habitat into account) should be calculated for each survey. This recommendation was extended to habitat specific correction factors by Gilbert et al (1999). The results reported here confirm that this approach is necessary, given the range of response rates from the 4 different habitats. The population estimate from this survey is roughly twice that of Mainwood et al (1997) and half that of Dennis (1976). The latter estimate was little more than a guess based on trapping rates in mist nets compared to those recorded for Skokholm. The former estimate was the first systematic survey 53 — 47-63 172 — 159-190 1429 — 1358-1516 2716 2149-3270 1774-4300 4370 = 3338-6069 of Priest Island and, interestingly, the total number of responses recorded (1637) was very similar to that recorded in this study (1751). The difference in the estimated population is due to the correction factors used. Mainwood etal (1997) used a factor of 1.37 based on published male and female response rates plus a correction factor equation developed by James & Robertson (1985). Thus the correction factor was roughly half that of those used in the current study. It is encouraging that the total responses recorded in the field are broadly similar in both studies. Clearly, the crucial difference was due to the correction factor and the colony and habitat specific approach taken in this study, plus the mathematical calculation of the total number of AOS in the calibration plots. Therefore, the best estimate to date of the Priest Island Storm Petrel population is about 4,400 AOS. This suggests that Priest Island is one of the largest colonies in Scotland. Recent estimates for other key colonies are: Auskerry, Orkney - 3,600 (Wood 1997); Treshnish Isles - 5,000 (Gilbert et al 1998); Mousa, Shetland - 6,800 (Ratcliffe et al 1998a). The suggestion from previous studies that Storm Petrel breeding distribution is clumped into “sub colonies” eg Mainwood et al (1997), Gilbert et al 52 P Mayhew, K Chisholm, H Insley & N Ratcliffe SB 21(2) (1998) is not borne out by the experience of this survey. Storm Petrel AOS were found throughout the heath/grassland habitat ina seemingly random fashion and it is suggested that a randomised survey of at least 20% of the habitat is the most appropriate way to survey petrels breeding in this type of habitat. In our view, it is not possible to locate all burrows or “sub colonies” in this habitat, as suggested in Gilbert et al (1999), unless a systematic tape playback method 1s used. The amount of work required in this type of survey is significant. Approximately 42 man days were required to complete this one survey. This is principally because correction factors must be calculated for each habitat over a period of about 7 days. Tocarry out similar surveys on key colonies to assess baseline populations and then monitor them would require major resources. The results in Fig 2 suggest it may be worthwhile investigating whether a shorter calibration period (eg 5 days) would give an acceptable level of accuracy for habitat specific correction factors. Whilst the method employed is, without doubt, the most accurate available to date, some concern must remain over its accuracy. For example, the least accurately censused habitat (heath/grassland) contained over half the population. Further work to improve the main survey technique in large scale, low density habitat would be helpful. Initial results from a smaller Storm Petrel colony on Eilean Hoan, Durness, suggest that tape playback methods result in a population estimate of approximately one third of that from mark/ recapture mist net results conducted in mid June when the number of non breeders at the colony should be minimal. While there are significant difficulties associated with interpreting results from mark/ recapture studies at breeding colonies, the du Feu method can be used to estimate the number of birds that have used the site during the trapping period (du Feu et al 1983, Amengual etal 1999, M. Hounsome, pers comm). During the next 2 years, itis hoped to carry out studies on both Eilean Hoan and Priest Island to compare results from the 2 methods. Acknowledgements Weare extremely grateful to Dave O’ Hara, David Russell, Mike and Wendy Lofthouse and Jacquie Heaton who, with 3 of the authors, made up the team carrying out this survey. Dr Neil Cowie advised on topographical area measurements. Tony Mainwood and Drs Gillian Gilbert and Mike Hounsome kindly commented on the manuscript. Morag Mackenzie typed the many drafts. References Amengual J F, Gargallo G, Suarez M, BonninJ, Gonzalez J M, Rebassa, & McMinn M 1999. The Mediterranean Storm Petrel Hydrobates pelagicus melitensis at Cabrera archipelago (Balearic Islands, Spain): Breeding moult, biometry and evaluation of the population size by mark and recapture techniques. Ringing & Migration 19:181- 190. Dennis, R H 1976. Scottish Bird Report, 1975. Scottish Birds 9:181. du Feu C, Hounsome M, Spence1 1983. A single session Mark Recapture Method of Population Estimation. Ringing & Migration 4:211-226. Gilbert G, Hemsley D & Shepherd M 1998. A survey of Storm Petrels on the Treshnish Isles in 1996 Scottish Birds 19:145-153. Gilbert G, Gibbons D W & Evans J 1999. Bird Monitoring Methods - a manual of techniques for key UK species RSPB, BTO, JNCC, ITE & The Seabird Group. Greenwood JJ D 1991. Estimating the total numbers and its confidence limits. Appendix to: Shrubb M & Lack PC The numbers and distribution of Lapwings V. Vanellus nesting in England and Wales in 1987. Bird Study 38:20-37. James P C & Robertson H A 1985. The use of playback recordings to detect and census nocturnal burrowing seabirds. Seabird 8:18-20. Lloyd C, Tasker M L, Partridge K 1991. The status of seabirds in Britain and Ireland. T & A D Poyser. Scottish Birds (2000) A survey of Storm Petrels on Priest Island in 1999 83 Mainwood A R, Ratcliffe N, Murray S & Mudge P 1997. Population status of Storm Petrels (Hydrobates pelagicus) on islands off north-west Scotland. Seabird 19:22-30. Ratcliffe N, Vaughan D, Whyte C & Shepherd M 1998a. The status of Storm Petrels on Mousa, Shetland. Scottish Birds 19:154-159. Ratcliffe N, Vaughan D, Whyte C & Shepherd M 1998b. The development of playback census methods for Storm Petrels (Hydrobates pelagicus) Bird Study 45:302-312. Vaughan D & Gibbons D W_ 1998. The status of breeding Storm Petrels (Hydrobates pelagicus) on Skokholm Island in 1995. Seabird 20:12-21. Wood D 1997. An estimate of the numbers of Storm Petrels (Hydrobates pelagicus) breeding on Auskerry, Orkney. Seabird 19:40-46. Peter Mayhew, Kenna Chisholm, The Royal Society for the Protection of Birds, Etive House, Beechwood Park, Inverness IV2 3BW. Hugh Insley, 1 Drummond Place, Inverness IV2 4JT. Norman Ratcliffe, The Royal Society for the Protection of Birds, The Lodge, Sandy, Bedfordshire SG19 2DL. Revised manuscript accepted May 2000 ~ ee = — —* ~ > me * ~ ia ee eal > gga ag ‘ : a ar os ie* aj" an aie Pt ee = — F oo _ Storm Petrel at breeding cavity. Hugh Insley 84 P Mayhew, K Chisholm, H Insley & N Ratcliffe SB 21(2) Appendix 1 The cumulative number of AOS detected for each of the 4 calibration plots was modelled using the equation: via cleress) Equation | This predicts the number of AOS detected (y) on a given visit (x) according to the exponential (e) proportional rate of increase (b) to the asymptote (a). The coefficient of (a) is the total number of AOS present in the study plot if sampling was continued for longer. The values of the coefficient were estimated by minimising the sum of the squared deviations of the observed values from those predicted by iteratively varying the values of the coefficients in the equation. The values for (a) and (b) in each habitat are given in Table A (+lse). The values of the coefficients for (b) were substituted into the equation |-e” in order to estimate the proportion of the total population detected on the first visit (ie the response rate). The upper 95% confidence intervals can be determined from the equation 1-e?“*"!”® and the lower 95% CI from the equation 1-e?***"!°® | where SE is the standard error given in Table A. q g Table A Coefficients of (a) and (b) estimated from regression equations (+1 standard error). The values in brackets in column (a) are the total number of AOS detected by visit 7. Subtracting these values from the estimate of (a) gives a figure for how many AOS were present but not detected by visit 7. Habitat a b Boulder Beach 30.48 + 0.85 (30) 0.64 + 0.065 Wall 58.03 + 1.08 (57) 0.54 + 0.033 Scree 0395)-=0!5516)) 0.64 + 0.026 Heath and Grassland G2 10.22 3),22, (ll) 0.44 + 0.064 Scottish Birds (2000) 21:85-87 Egg sizes of crossbills in Scotland H A McGHIE & R W SUMMERS The Scottish Crossbill is intermediate in size between the Common Crossbill and Parrot Crossbill and would be expected to lay eggs of an intermediate size. However, examination of the sizes of crossbill eggs collected in the Highlands of Scotland confirms that they are not significantly different from Common Crossbills. Thus, Scottish Crossbill eggs are either similar in size to those of Common Crossbills or many of the eggs collected in the Highlands actually refer to Common Crossbills. Currently it is thought that 3 species of crossbill breed in Scotland: the Common Crossbill Loxia curvirostra, Scottish Crossbill L scotica and Parrot Crossbill L pytyopsittacus, though the latter is rare (Gibbons ef al 1993). However, because of similarities in plumage and overlap in biometrics (Knox 1976), identification is not always straight forward (Knox 1990a). The Scottish Crossbill has a bill depth and wing length intermediate between the Common Crossbill and the Parrot Crossbill, though it is closer to the Common Crossbill in these measurements (Knox 1976). It has been argued that the Scottish Crossbill must be regarded as a full species because it maintains this intermediate size despite repeated invasions of the other crossbill species into Scotland (Knox 1975, Voous 1978). Although it is accepted as a full species (BOURC 1980), there are many aspects of its status and biology that are unknown because it is difficult to identify in the field (Knox 1990a), so ithas been categorised as data deficient (Tucker & Heath 1994). One might expect the egg sizes of Scottish Crossbills to be intermediate between Common and Parrot Crossbills, given that the Scottish Crossbill is intermediate in wing length and presumably body size. However, Jourdain, cited in Nethersole-Thompson (1975), gives mean values for 100 Scottish Crossbill eggs (mean length x mean breadth = 21.64 x 15.9mm) which are lower than those of 100 Common Crossbill eggs (22.12 x 16.11mm). This prompted an examination of crossbill eggs in museum collections. Methods Crossbill eggs were measured from the collections in Inverness Museum and Art Gallery, the National Museum of Scotland (Edinburgh) and the Natural History Museum (Tring). Full clutches from Scotland, England and Scandinavia were examined. Lengths and breadths were measured to the nearest 0.1 mm with Vernier callipers; all measurements were taken by HM. Mean lengths and breadths were calculated for each clutch to account for within clutch variations in size. An index of egg volume was derived from length x breadth’. Crossbill clutches from England were accepted as being from Common Crossbills because Scottish Crossbills are unknown in England and Parrot Crossbills are rare (Gibbons etal 1993). Likewise, Parrot Crossbill clutches collected in Scandinavia were accepted as being correctly identified. Species identification of crossbill clutches from Scotland was not attempted but only clutches collected from the Highlands of Scotland, within the supposed range of the Scottish Crossbill (Nethersole-Thompson 1975, Gibbons etal 1993), were used in analysis. The crossbill clutches were mainly collected in the early part of the 20th S86 H A McGhie & R W Summers SB 21(2) century, but there is no evidence to suggest that the Scottish Crossbill has undergone a significant change in range since then. Results There were significant differences in length and breadth between the 3 groups of crossbills (Table 1). Parrot Crossbill eggs were significantly longer and broader than those of Common Crossbills (t= 4.9, P<0.001 andt=4.0, P<0.0001, respectively), though there was overlap (Table 1). Likewise, the lengths and breadths of crossbill eggs from Scotland were significantly different from those of Parrot Crossbills (t=4.9,P=0.001 andt=2.7, P = 0.007, respectively) but not from Common Grossbillsi(G=0:85.P=<0and t—se2 9520): The mean length of the Scottish eggs was actually smaller than the Common Crossbills. A similar pattern emerged using indices of egg volume. Egg volumes of Parrot Crossbills were signifcantly greater than those from Common Crossbills (¢ ='0'55, P’=0 59)" Giiablem) Discussion The egg measurements of the Common Crossbills from Scotland were similar to those quoted by Nerhersole-Thompson (1975) and confirmed that those from Scotland were not intermediate between Common and Parrot Crossbills. It is possible that eggs measured by Jourdain were included in the present study. Egg size variation within a species, and presumably between closely related species, is mainly related to the size of the female (O’Connor in Campbell & Lack 1985). Therefore, the results are unexpected. One possibility to account for this anomaly is that Table 1 Lengths, breadths (mm) and indices of volume (length x breadth? (cm’)) of eggs from Parrot and Common Crossbills and crossbills from the Highlands of Scotland. Parrot Crossbill Common Crossbill Highland Crossbills Number of clutches 20 79 105 Length Mean 22.96 21.96 21.84 SD 0.79 0.83 0.96 Min PINT 2.027 24.37 Max 24.19 24.18 24.37 Breadth Mean 16.44 16.00 16.10 SD 0.50 0.43 0.52 Min Se 14.75 14.78 Max 17.46 208 17-29 Index of Mean 6222 5.63 S67) Volume SD 0.49 0.43 6.85 Min 5.5) 4.53 4.21 Max 7.31 659 6.85 ANOVA on egg length: F (2,201) = 13.3, P<0.001 ANOVA on egg breadth: F (2,201) = 6.5, P<0.002 ANOVA on egg volume: F (2,201) =:13.1, P<0.001 Scottish Birds (2000) Egg sizes in crossbills in Scotland 87 many of the clutches collected in Scotland were primarily from Common Crossbills, given that Scottish and Common Crossbills breed sympatrically (Knox 1990b). At least 78 of the 105 clutches were from one site: Fairburn Estate, near Dingwall, Ross-shire. Harvie-Brown (undated) gives an indication of the scale of breeding and nest finding at Fairburn: upwards of 50 nests were found in 1901, only 2 in 1902, but over 50 in 1903. Although the main egg collector did not differentiate between Common and Scottish Crossbills in the collection catalogue, other ornithologists who collected clutches from Fairburn classified them as Scottish Crossbill eggs. The clutches were collected from old plantations of Scots Pines Pinus sylvestris and Hybrid Larches Larix x eurolepis, from Scots Pines lining drives and avenues, and remnant Caledonian pinewoods (clutch record cards). This concords with Pennie’s (1950) description of Fairburn in the early part of the 20th century. These tree species and stand types are known to be used by both species of crossbill. For example, Common Crossbills in Norfolk make use of Scots Pines (Nethersole-Thompson 1975) so one cannot rely on habitat as a means of identification. Even if the sample from the Highlands did contain Common Crossbill clutches, one would still expect the average measurements to be higher. Given that the crossbills from the Highlands had eggs which were only 0.7% greater in volume from the Common Crossbills, it does suggest that there is little difference in egg sizes of Common and Scottish Crossbills. However, the only way to determine the true sizes of Scottish Crossbill eggs is to obtain measurements from clutches where one is absolutely sure of the identification of the adults. Acknowledgements We thank Stephen Moran of Inverness Museum and Art Gallery, Bob McGowan of the National Museums of Scotland, and Robert Prys-Jones and Michael Walters of the Natural History Museum, Tring, for granting HM access to their collections. The drafts were commented upon by IP Bainbridge, D W Gibbons, D Jardine, M Marquiss, Capt R Stirling and M Thompson. References BOU Records committee 1980. Records committee: tenth report. bis 122:564-568. Cambell B & Lack E 1985. A Dictionary of Birds. Poyser, Calton. Gibbons D W, Reid J B A & Chapman RA 1993. The New Atlas of Breeding Bird in Britain and Ireland: 1988-1991. Poyser, London. Harvie-Brown J A undated. Manuscripts. Museums of Scotland. Knox A G 1975. Crossbill taxonomy. In, Nethersole- Thompson D, Pine Crossbills. Poyser, Berkhamsted. Knox A G 1990a. The identification of Crossbill and Scottish Crossbill. British Birds 83:89-94. Knox A G 1990b. The sympatric breeding of Common and Scottish Crossbills (Loxia curvirostra and Lexia scotica) and evolution of crossbills. Jbis 132:454:466. Nethersole-Thompson D 1975. Pine Crossbills. Poyser, Berkhamsted. Pennie I D 1950. The history and distribution of the capercailzie in Scotland. Scottish Naturalist 62:157- 178. Tucker G M & Heath M F 1994. Birds in Europe: Their Conservation Status. Cambridge, UK: Birdlife International (Birdlife Conservation Series No3). Voous K H 1978. The Scottish Crossbill: Loxia scotica. British Birds 71:3-10. National Henry McGhie, Manchester Museum, Oxford Road, Manchester, M13 9PL Ron Summers, RSPB, Etive House, Beechwood Park, Inverness, 1V2 3BW Revised manuscript accepted June 2000 8S Scottish Birds (2000) 21:88-97 SB 21(2) Density and habitat associations of Barn Owls in East Ross H A McGHIE An area of 146km? of Mid Ross was searched for Barn Owls between 1995-9. A total of 27 home ranges was located and the mean occupancy rate during the breeding season was 95%. The mean altitude of nest sites was 48m asl and 93% of pairs were below 100m asl. Breeding sites were regularly spaced below 100m asl, and the mean nearest neighbour distance was 1461m; density was high locally at 18 pairs per 10km square. The majority of nest sites were in tree cavities, mainly in Ash and Beech, although a wider range of structures was used for roosting. Areas that held breeding Barn Owls had a higher number of potential nest sites than areas that did not, but there was no difference in the amount of feeding habitat. Nest sites and roost sites had lower levels of disturbance than unused sites and a greater area of rough grassland within 400m; unused sites were significantly further from nest sites than were roosts. Introduction The populations of Barn Owls Tyto alba around the Moray Firth are of special interest as they represent the most northerly populations in the world of this cosmopolitan species (Glue 1976). Notwithstanding the marginal nature of these populations, Thom (1986) drew attention to the existence of a population at high density in the Inner Moray Firth and Black Isle areas in East Ross and East Ness. Shawyer (1987), however, considered the Inner Moray Firth population to be critically low and in need of primary protection as part of a nation wide strategy for Barn Owl conservation. Between 1995-99 I carried out a systematic search with the intention of establishing the density and habitat associations of part of this population. Study area and methods The study area consisted of 146km° of Mid Ross with | 15km/? of agricultural land, 22km/ of forestry plantations and 9km* of moorland. The exact limits are not given in order to protect Barn Owl nest sites, which are vulnerable to disturbance, and in accordance with the wishes of landowners who kindly granted full access to land. The area was mainly developed over Old Red Sandstone, (ORS) mainly below 75m above sea level (asl) in the east; this increased to 250m asl in the west over Moinian Schist with low ground restricted to the floors of alluviated glacial valleys. Good quality farmland, both mixed and arable, was widely distributed to the east but was limited to valley floors in the west. Sport shooting interests were found in most areas providing numerous coverts, shelter belts and estate plantations. Exotic conifer plantations largely separated agricultural areas at low altitudes from higher ground, where the marginal hill ground and moorland were used as sheep grazing. The area was intensively and systematically searched for the presence of Barn Owls in 1995 and 1997-99. All suitable structures (ie relatively quiet buildings, cliffs, quarries and areas with old trees away from the centre of forests) were checked for roosting or nesting Barn Owls. The methodology was essentially the same as that of Project Barn Owl organised by the BTO/Hawk and Owl Trust. Information on ‘white’ or ‘screech’ owls was alsorequested from landowners, farmers, gamekeepers, naturalists and residents. Each apparently suitable site was checked at least twice per year to establish occupancy in the breeding Scottish Birds (2000) Density and habitat associations of Barn Owls in East Ross 89 (April-September) and non breeding (October- March) seasons. Occupancy of sites was determined from sightings of Barn Owls entering or exiting sites and from signs including pellets, droppings, feathers and white fluff around tree holes or on beams at potential sites. A mirror and a torch mounted on the end of a walking stick was used to check unoccupied sites to ensure there was a suitable cavity available. The number of home ranges was established by observing pairs in and around confirmed or strongly suspected breeding sites; checks of several sites per night were made to demonstrate simultaneous occupancy of sites by different pairs. Disturbance was kept to a minimum and birds were not flushed from sites; nest contents were not checked, so that breeding outcome was often not established. Confirmation of breeding was taken from eggshell fragments below nests, adults carrying prey into apparently suitable sites and sounds and sightings of young. Suitable roosting and nesting sites and potential feeding areas consisting of rough grassland patches, wide verges, woodland edges and field edges, ditches and riverside vegetation were recorded onto 1:25,000 Ordnance Survey maps. Disturbance at each apparently suitable site (ie structurally capable of holding roosting or breeding Barn Owls) between April-September was scored on a scale of 0-2. A score of 0 denoted no known human disturbance; a score of | denoted regular but non-persistent human disturbance, as where vehicles were stored in buildings; a score of 2 denoted regular and prolonged human disturbance, as when buildings were used as workshops or for livestock. Whilst this system was largely qualitative it was felt that some measure of disturbance was better then none. It was appreciated that sites that were heavily disturbed during the day could be used as nocturnal roosts by Barn Owls. Habitat features within an 800m radius of confirmed or strongly suspected nest sites were compared with features within 800m radius of 40 randomly selected points in farmland, 30 of which were below 100m as] and 10 of which were above 100m asl. This was done to determine whether there were any differences in habitat between areas that held breeding Barn Owls and areas that did not. Habitat features were measured from the 1:25,000 maps. Verges were given an approximate value of 2m in width and wide verges were double counted. Barn Owls are known to range further than 800m but as this area would encompass the main home range itis probable that habitat features within this area would be important in habitat selection (Taylor 1994, Cramp 1985). Habitat features within 400m of nest sites, roost sites and unused sites were compared within each home range, to establish selection of habitat features for nesting and roosting. Only one roost site was known to be more than 800m from the main nest site, thus falling outside the 800m radius defined above; this site was included in calculations. For each habitat feature, the mean values of all roosts and all unused sites within each home range were used so that there was one value for each nest sites, roosts and unused sites for each habitat feature for each home range. This was necessitated by the varying number of roosts and unused sites in different home ranges: each home range had to carry equal weighting as the intention was to compare habitat features around the different types of site (nest sites, roosts, unused sites) within, rather than between, home ranges. Statistics follow Fowler and Cohen (BTO Guide 22) throughout. The nearest neighbour distance for each known or strongly suspected breeding site was measured from 1:25,000 maps. Random points were generated using Microsoft Excel and plotted on maps in order to test the actual spacing of nests against arandom distribution. Expected distances could not be simply calculated from the area and number of points alone as not all of the area was available to breeding Barn Owls because of the 90 HA McGhie SB 21(2) lack of feeding and/or nest and roost sites in areas of forestry and moorland (see Table 1). The first 27 randomly generated points which lay within 1km? containing apparently suitable nest sites and feeding were used; points which fell in 1km/? lacking sites and feeding were rejected as these were not available to Barn Owls for breeding. each used for one year only in 1984 and 1971 respectively. Breeding was confirmed at least once between 1995-99 in 15 of the 27 home ranges and strongly suspected from the remaining 12 sites on the basis of the persistent presence of pairs over several years. Occupancy of home ranges was very high throughout the period of Table 1 Use of lkm/’ of National Grid by Barn Owls 1995-99. Number of 1km? <100m asl Feeding Suitable Breeding Summer Winter |Unused Breeding Summer Winter only Sites (not breeding) Farmland Y WY, 18 16 Ve N 9 N N Forestry Y Ye | Y N | Moorland Y WC Y; Results Number of pairs and occupancy Barn Owls were confirmed or strongly suspected of breeding in 27 home ranges between 1995-99. This produced a maximum density (ie with 100% occupancy) of one pair per 5.41 km? or 18 pairs per 10km square. One of these areas was used for breeding in 1995 and 1996 only, although this area had previously held a breeding pair in the 1970s. Evidence for former breeding was received from three additional areas: 2 former home ranges became amalgamated following the destruction of nest sites in each home range c1993. Two higher altitude nest sites, at 130m and 160m asl, were >100m asl Unused (not only breeding) | 8 10 6 Ws 8 5) 2 14 1 2 + 22 2 10 study, varying from 83% (winter 1998/9, 24 checked) to 100%. The occupancy rate of known or strongly suspected breeding sites between April- September was 100% in 1995 (20 checked), 100% in 1997 (23 checked), 96% in 1998 (26 checked) and 85% in 1999 (27 checked). Distribution Barn Owls were well distributed below 100m, using 71% of 1km?’ during the breeding season for breeding or hunting; an additional 6% of 1km? were used in winter only (Table 1). Barn Owls were much less widely distributed above 100m asl, using 42% of km’. Barn Owls were much more widely distributed in farmland than in Scottish Birds (2000) woodland or moorland, where suitable sites were absent or uncommon. Barn Owls used 77% of farmland 1km*below 100masl during the breeding season but only 40% of farmland 1km? above 100m as]. A higher proportion of farmland |km? above 100m as] was used only in winter when compared with farmland 1km* below 100masl. Below 100m asl, 95% of squares with suitable sites and feeding were used during the breeding season whilst 49% of squares above 100m as! were used. The mean altitude of nest sites was 48m asl (SD=38m, n=27); only 2 nest sites were above 100m (110m and 180m asl) and occupancy at both of these sites was irregular. Spacing between pairs The mean nearest neighbour distance was 1461m (SD=421, n=27) and a significantly higher proportion of nests were separated by 1200-1600m than would have been expected if nests had been randomly distributed (P<0.01, chi-squared test, 2 degrees of freedom after grouping to bring expected values above 5)(Table 2). Only 2 pairs of sites were separated by less than 1000m. One of these sites was used for 2 years only (breeding confirmed at both sites in both years), and a busy road that was bordered by thick woodland which was unused by Barn Owls separated the other pair of sites. Pairs were very regularly distributed below 100m asl, where there were many areas of suitable feeding with apparently suitable nesting and roosting sites. Spacing was much less regular above 100m as! where there was less suitable Density and habitat associations of Barn Owls in East Ross 91 habitat and where owls were not so widely distributed. Use of structures Barn Owls used tree cavities as the main nest site in 24 of the 27 home ranges (89%)(Table 3). This is a higher proportion than quoted by Blaker for England (1933, 43%), Sharrock for Britain (1976, 39%), Bunn et al (1982, 32%) or Taylor for South Esk (1994, 9.9%) but similar to Taylor’s (1994) value of 74% fora small study areanear Edinburgh. Of 68 little disturbed sites examined (scores 0 and 1, see Methods) 79% were in tree holes so that the high proportion of tree nesting found in this study was a consequence of the relative abundance of large, old trees and the relatively small number of quiet buildings. There was no evidence for selection in favour of tree sites or against buildings (contra Shawyer 1987). When all alternative nest sites known to have been used at any time within the study area were considered, the proportion in trees fell to 69% (n=39) although those in buildings would be more likely to be found in the absence of intensive fieldwork. Beech Fagus sylvaticus and Ash Fraxinus excelsior were equally important as nest sites, together constituting 63% of all nest sites used from 1995- 99 (Table 3). Glue (in Bunn et al 1982) found Beech and Ash to comprise 20% of nest sites in an English study compared with 56% in Oak and Elm. Table 2 Nearest neighbour distances between confirmed or strongly suspected nest sites. Number of confirmed or strongly suspected breeding sites 0-400m 400-800m 800-1200m Observed 2 2: Expected! 2, 7 7 1200-1600m 1600-2000m >2000m 16 3 4 2 5 4 1 Based on measurements between 27 randomly generated points in 1km/? with suitable sites and feeding habitat (see Methods) 92 HA McGhie SB 21(2) Table 3 Use of structures within 800m of confirmed or strongly suspected nest sites by Barn Owls 1995-99. Number of structures Noofhome Confirmed ranges in breeding which present Ash IS 5 Beech 15 6 Horse Chestnut 2 | Oak 4 l Farms 26 2 Disused buildings 9 2 Cliff/quarry 2, Box hedge | Grey Poplar | Wych Elm | Sycamore 3 Cherry | Lime 2 Hay barns 19 Main nest sites in trees were along field edges (7), in or close to the edge of small clumps of broadleaved trees (7), in quiet road or trackside avenues (5), in solitary trees (3) and in scattered parkland trees (2). The mean distance to the nearest A or B road was 407m (Table 5) and none nested or roosted within 20m of these roads. Almost all trees used for breeding were plantings from the Victorian era although some were probably older. Nest sites in solitary and parkland trees were all within 200m of cover (eg thick conifers). Of the 24 main nest sites in trees, 22 (92%) were in the trunk, 1 was in a large hollow branch and one was in a large suckering growth (‘witches’ broom’). All trees were alive; 3 did not have full crowns but were sheltered by the crowns of adjacent trees. The mean height of nest entrances in trees was 2.98m (SD=0.58, n=24, maximum=4m) which was lower than the modal height of 3.5-6.0m found by Glue. This was probably attributable to differences in structure of Ash and Beech and Elm Suspected Roosts Winter Unused breeding only 5 14 3 3 4 1] 1 l 1 2 | 9 26 1 3 D 3 2 1 Il I 2 1 l l 15 13 and Oak, rather than to differences in selection by Barn Owls. Regularly used nest sites in buildings were in a loft, a ruin and in the chimney of a disused building. A wider variety of structures were used as breeding roosts. Roosts known to have been used during the breeding season were adjacent (<20m) to nest sites in 12 home ranges; in 4 home ranges with greater spacing between possible roost sites, the mean distance to 7 roost sites was 600m, with a maximum distance of 900m. In one home range, roosts were 600m, 600m, 800m and 900m from the nest site, and the site that was 900m distant was used as an alternative nest site in 1998. Barn Owls were also thought to roost in thick conifers (see Taylor 1994), at least occasionally, but these were not searched. All of the roost sites used during the breeding season were considered to have been located in 23 Scottish Birds (2000) Density and habitat associations of Barn Owls in East Ross 93 Table 4 Habitats in which Barn Owls were recorded hunting. Number of home ranges in which recorded Important feeding habitat Of which recorded as main feeding habitat Patches of rough grassland’ 20 12 Woodland edge 8 0 Open scrub 10 l Waterside and ditches 17 2 Tracks and roadside verges 11 1 Field edges 16 4 Young forestry plantations 1 it 1 Includes marshy areas and field headlands Table 5 Habitat features within 800m of nest sites and randomly selected points within farmland. Nest sites Randomly selected areas Significant within farmland differences! a b c Below 100m asl Above 100m asl n=27 n=30 n=10 Mean SD Mean SD Mean SD No of quiet suitablesites*» 2.70 0.87 P37. 35 0.90 1.29 a>b a>c b>c Waterside & ditch (km)° 2.95 0.11 121 0190 1.16 0.10 a>c Distance to nearest main road (m)? 407.19 2.59 384.86 3.30 936.92 1.81 None Total rough grassland (ha)° 12.59 AO MO MOF 2797 11.00 3.26 None Woodland edge (km) 3.54 1.90 BS) PAI) 3.40 2.04 None No of disturbed sites**> aasyih 1.48 DAD Wes 2.04 leit None 1 P<0.01, one tailed Mann-Whitney U tests for differences in medians (performed on untransformed data) 2 Number of sites with disturbance scores 0 and 1 (see Methods) 3 Distance to nearest A or B grade road 4 Number of sites with disturbance scores of 2 (see Methods) 5 Mean and standard deviation figures are back transformed from the log(x+1) distribution 94 HA McGhie SB 21(2) home ranges, based on observations of emerging adults. Barn Owls used only one site, including the nest site, in4 home ranges (2 of which were in large spacious buildings with plenty of room for breeding and roosting). Barn Owls used 2 sites in 10 home ranges, 3 sites in 7 home ranges and 4 sites in 2 home ranges (mean=2.3, SD=0.88, n=23). The majority of roosts used during the breeding season were not used outside the breeding season. Barn Owls commonly resorted to hay barns in winter, normally between September and February. When all structures used throughout the year for breeding and roosting are considered, Barn Owls used one site in 4 home ranges, 2 sites in 7 home ranges, 3 sites in 6 home ranges, 4 sites in 4 home ranges, 5 sites in one home range and 7 sites in one home range (mean=2.78, SD=1.44, n=23). Habitat usage Observations and records of hunting Barn Owls were collected from 24 of the 27 home ranges (Table 4) and although these were not done on a systematic basis the results were considered to give a good indication of habitat usage as records were collected from a large number of land users over a number of years. Barn Owls hunted over a variety of grassland and open habitats but were not known to hunt commonly over forestry plantations, cultivated land or settlements. Barn Owls tended to frequent areas with quiet suitable roost and nest sites significantly more than other areas but there were no significant differences in the amount of feeding habitat between used and unused areas (Table 5). This was taken to indicate that the distribution of Barn Owls was limited by the availability of suitable quiet sites rather than by the availability of feeding habitat. Within each territory, nest sites and roost sites did not differ significantly in terms of any of the habitat features quantified. However, in 12 of 13 home ranges where there was more than one cavity available, Barn Owls used the largest cavity for breeding, the exception being where Barn Owls nested in a tree with a smaller cavity but with much better shelter. Barn Owls roosted in trees that were significantly closer to the nest site than were unused sites (Table 6). Potential competitors and predators Five Tawny Ow] Strix aluco nests were located in the course of the present study; Tawny Owls generally nest in more enclosed woodland and in smaller cavities (Mikkola 1983) and only one of the sites was considered suitable for Barn Owls. Tawny Owls and Barn Owls nested less than 10m from each other in 1995 and 1996 in one area. Pine Martens Martes martes were known to have bred successfully in a tree normally used by Barn Owls in 1994 but the owls subsequently reoccupied the site in 1995. One site was taken over by feral cats in c1994 but was found to have been reoccupied by Barn Owls in 1999. Wildcats Felis silvestris bred close to Barn Owls in another home range but there was no evidence of animosity and the owls were known to have bred successfully in most years between 1994-99. Discussion This study found Barn Owls to occur at 18 pairs per 10km square, a very high density (Shawyer 1987). This density was thought to extend throughout the Old Red Sandstone district of Mid Ross and some distance around the Cromarty and Beauly Firths. Barn Owls were very thinly distributed west of the Old Red Sandstone district: less than 5 sites were probably occupied regularly between 1995-99, although fieldwork was less extensive and not systematic in that area. The Barn Owl would appear to have been rare or absent from the study area in the 19th Century (Holloway 1998) and 2 lines of evidence suggest Scottish Birds (2000) Density and habitat associations of Barn Owls in East Ross 95 Table 6 Habitat features within 400m of nests, roosts and unused sites within each home range. Nests n=27 Roosts n=19 Unused sites Quiet unused Significant n=26 sites! n=11 difference a b c d Mean SD Mean SD Mean SD Mean SD Disturbance score! 0.29 1.20 OSU ES! i] weet ac b>c Distance to nest (m)* 88.41 6.88 595.32 1.57 628.41 1.65 bc Waterside vegetation (m)°710.38 4.01 SOS Te - 22h 95997 Te4ai2 651.4 1.70 None No of farms? 0.71 0.38 0.41 0.48 0.74 0.26 0.12 0.26 None 1 See Methods. Quiet sites defined as those with disturbance scores of 0 or | 2 P<0.01, one-tailed Wilcoxon’s test for matched pairs (performed on untransformed data) 3 Mean and standard deviation figures are back transformed from the log(x+1) distribution that this was still the case in the early part of the 20th Century. Firstly, there are no Barn Owl eggs inthe comprehensive William Stirling of Fairburn egg collection (c1890-1910) or in any other collection in Inverness Museum and Art Gallery (eg McGhie 1994), although Pollock (1902) recorded the species in a short list of birds in nearby Kilmorack parish (Invernessshire). Secondly, only 3 Barn Owls were submitted to an Inverness firm of taxidermists between 1912-69 from the area presently under study (McGhie 1999). Other birds of prey were being submitted in considerable numbers at this time and this suggests that the Barn Owl was rare, especially as it is one of the most popular species for taxidermy. The Barn Ow] was not recorded for East Ross in the 1920s, although it is marked as resident in all surrounding vice counties (Baxter and Rintoul 1928). Baxter and Rintoul (1953) later described the species as “by no means rare in Rossshire’. Two sites just outside the study area which were occupied by Barn Owls in the 1950s were still occupied in the 1990s and were though to have been occupied more or less continuously between these dates. Barn Owls were recorded for several squares in the Moray Firth basin during the period of the The Atlas of Breeding Birds in Britain and Ireland. (Sharrock 1976). There would seem to have been an increase in the number of pairs of Barn Ow] in the Inner Moray Firth since Headlam noted the presence of ‘a few pairs’ (in Bunn eft al 1982). Shawyer (1987) estimated that there were 10 pairs in the whole of Ross-shire and considered the population to be critically low. The New Atlas (Gibbons ef al 1993) shows how Barn Owls spread in all directions into 10km/? in which they were not encountered during the 1968-72 Atlas period and Crooke (1999) stated that the number of pairs had increased steadily throughout the 1990s. Many of the nest sites included in the present study were known by land 96 HA McGhie SB 21(2) users and residents to have been occupied by Barn Owls before the 1990s and these sites would appear to have been previously overlooked, although breeding was rarely confirmed. Some new sites have certainly been occupied by Barn Owls since the late 1980s however. The Barn Owl is one of several species that approach the northern limit of their British, and sometimes world, distributions in the Moray Firth area; these species may be presumed to be especially sensitive to cold winters (Percival 1991). Mild winters in recent years may have increased over winter survival in Barn Owl, and also Kingfisher Alcedo atthis (see Dennis 1995), resulting in the spread of Barn Owls into previously unoccupied areas in the Great Glen, the Black Isle and northwards into south east Sutherland. Barn Owl populations are known to have increased breeding success and overwinter survival since the mid 1970’s (Percival 1991), which may also be linked with declines in pesticide levels in Barn Owls (Newton 1991). Barn Owls can be threatened by disturbance or destruction of nest and roost sites, and the loss of feeding habitat. Following disturbance in the present study, birds in 3 home ranges switched from buildings to nearby tree sites. Nest boxes were provided in 3 other home ranges to try to retain Barn Owls when the main nest site had become unsuitable or unstable. Most sites were considered tobe fairly secure for the immediate future, excepting tree felling, as Barn Owls were not breeding in very dead trees. Few hardwood trees have been planted since the Victorian era and this could lead to a future shortage of suitable nesting trees. Road casualties were recorded from 8 home ranges (30%) between 1995-99 and probably occurred undetected at more. In another 8 home ranges they were unlikely to occur because of their remoteness from busy roads. Barn Owls were killed almost annually at one site but the site was always reoccupied and often succeeded in producing some young. Housing developments threaten Barn Owls in many home ranges: green field sites on steep banks at the edge of woodland, which are very attractive to developers, are also important as a feeding habitat for Barn Owls. All landowners were informed of the presence of Barn Owl nest and roost sites on their land, and that these were some of the most northerly Barn Owls in the world. This was done to try to reduce the risk of tree nest sites being felled inadvertently. The provision of further nest boxes in areas with suitable feeding habitat but insufficient quiet sites may attract further pairs of Barn Owls and increase the population. Acknowledgements I am extremely grateful to all the farmers, gamekeepers, residents and other land users who offered me information and to all the landowners who kindly granted me free access to their land. I am grateful to D Aiton, M Canham (Forest Enterprise), R Graham, M Hancock (RSPB), J Heaton, P Mackay (Forest Enterprise) and H Paton for information on some pairs of Barn Owls. I am likewise grateful to C Crooke (RSPB), B Etheridge (RSPB), D McAllister, S Moran (Inverness Museum and Art Gallery) and W Sinclair for useful discussion and assistance. I am grateful to SNH, and to C Bewsley in particular, for granting me licences to visit nest sites. References Baxter EV & Rintoul LJ 1928. The Status of Birds in Scotland. Oliver and Boyd, Edinburgh. Baxter EV & Rintoul LJ 1953. Birds in Scotland. Oliver and Boyd, Edinburgh. Blaker GB 1933. The Barn Owl in England- Results of the Census. Bird Notes and News 15:169-172; 207-211. Bunn DS, Warburton AB & Wilson RDS 1982. The Barn Owl. Poyser, Calton. Scottish Birds (2000) Density and habitat associations of Barn Owls in East Ross 97 Cramp S (ed) 1985. The Birds of the Western Palaearctic, Vol 4, Oxford University Press, Oxford. Crooke C (ed) 1999. Highland Bird Report 1997. RSPB, Inverness. Dennis R 1995. Birds of Badenoch and Strathspey. Colin Baxter, Grantown. Fowler J & Cohen L. Statistics for Ornithologists (BTO Guide 22). BTO, Tring. Gibbons DW, Reid JB & Chapman RA (eds) 1993. The New Atlas of Breeding Birds in Britain and Ireland: 1988-1991]. Poyser, London. Glue DE 1976. Barn Owl. In Sharrock JTR (ed) 1976. The Atlas of Breeding Birds in Britain and Ireland. BTO/Irish Wild Bird Conservancy, Tring. Holloway S 1995. The Historical Atlas of Breeding Birds in Britain and Ireland 1875-1900. Poyser, London. McGhie HA 1994. Discovery of the first British clutch of Slavonian Grebe eggs in a museum collection. Scottish Birds 17:166-167. McGhie HA 1999. Persecution of birds of prey in north Scotland 1913-69 as evidenced by taxidermists’ stuffing books. Scottish Birds 20:98-109 Mikkola H 1983. European Owls. Poyser, Calton, Percival SM 1991. Population trends in British Barn Owls- areview of some possible causes. British Wildlife 2:131-140. Pikula J, Bekloca M & Kubik V 1984. The breeding bionomy of Tyto alba. Prirodoved Pr Ustavu Cesk Akad Ved Brne. 18:1-56. Pollock JR 1902. Guide to Beauly and District. Pefferside Press, Dingwall. Sharrock J T R 1976. The Atlas of Breeding Birds of Britain and Ireland. Poyser, Berkhamsted. Shawyer CR 1987. The Barn Owl In The British Isles: Its Past, Present and Future. The Hawk Trust, London. Taylor IR 1994. Barn Owls: predator-prey relationships and conservation. University Press, Cambridge. Thom VM 1986. Birds in Scotland. Poyser, Calton. Henry A McGhie, The Manchester Museum, University of Manchester, Oxford Road, Manchester M13 9PL Revised manuscript accepted July 2000 Fronces Codires Barn Owl 98 Scottish Birds (2000) 21:98-108 SB 21(2) The breeding success of a population of Lapwings in part of Strathspey 1996-1998 P FRENCH, H INSLEY, G SIRIWARDENA & N BUXTON During a 3 year study of breeding Lapwings from 1996 to 1998 on 287ha in Strathspey, Scotland, 630 nesting attempts were monitored. Mean clutch size (3.6) and brood size at hatching (3.0) remained constant over the study period. Nest failure rate did not vary significantly between years and nest survival from laying to hatching was around 60%. However, there were very highly significant differences in the daily failure rates for nests across the 8 different study sites used, which ranged in habitat from drained, improved pasture to marshy rough grazing. Combining all 8 sites, survival of chicks from hatching to fledging varied between 9.5% (in 1996) and 17.5% (in 1997). These survival rates indicated a production of between 0.29 and 0.46 chicks per nesting attempt, which, allowing for replacement clutches, would at most have equated to between 0.38 and 0.77 chicks per breeding pair. Introduction Lapwing populations have been falling across many parts of western Europe for several decades (eg Reichholf 1996, Broyer & Benmergui 1998) and these declines have consistently been associated with changes in farming practice leading either directly or indirectly to reduced breeding success (Baines1989 & 1990, Shrubb 1990, Berg et al 1992, Triplet et al 1997). In Britain the most recent survey indicated that breeding Lapwings in England and Wales have declined from an estimated 123,000 pairs in 1987 to just over 63,000 pairs in 1998, a fall of nearly 50% (Wilson 1999). Although Scotland was not included in the 1987 survey, an estimated 92,000 pairs were found in the survey of breeding waders in lowland Scotland carried out in 1992 and 1993 (O’Brien 1996) and the 1998 BTO Lapwing survey, which attempted to estimate the population for the country as a whole, calculated the Scottish breeding population as 69,800 pairs (Wilson & Browne 1999). Since the confidence limits for these 2 surveys overlap no decline can be imputed from the lower estimate derived from the 1998 survey. Nevertheless, the Lapwing population in Scotland isnow apparently higher than that in all of England and Wales, and there are clear indications of problems there. Results from Breeding Bird Survey squares in Scotland indicate a statistically significant 28% drop in numbers between 1994 and 1998 (Noble et al 1999). This study of Lapwings breeding in the immediate surroundings of the village of Newtonmore in Strathspey, at an overall density of up to one pair per hectare, attempted to determine whether the apparently strong breeding populations of Lapwings in the Highlands were changing over time by studying their breeding success and trying to determine if this was sufficient to maintain the local population, or even to export juveniles to other British breeding populations. Scottish Birds (2000) Breeding success of Lapwings in Strathspey 1996-1998 99 Methods During the 3 breeding seasons 1996 to 1998, an intensive study was carried out of the breeding Lapwings on 287 hectares of Strathspey close to the village of Newtonmore. The study area (which was divided into 8 non contiguous sites) was watched intensively and searched in the period between the last week in March and the second week in June each year to locate and monitor all Lapwing nesting attempts. Mostnests were located by remote observation of sitting birds using a telescope with, in the more distant or difficult cases, the observer being directed to the nest by an assistant using hand held radios. The same techniques were used to locate chicks. Individual nests were marked and numbered, and breeding success was monitored through repeat visits while the nests were still active (ie until hatching or failure was observed). 68% of nests were visited sufficiently close to hatching for chick production to be measured directly (ie when the chicks were still hatching or were still present or nearby). When this brief period was missed, success or failure was assessed on the basis of the presence or absence of eggshell and on whether its appearance suggested hatching or failure. Damaged or crushed eggshells were clear indications of failure, whereas the presence of fine fragments or hatched eggshells were taken to indicate successful hatching. A completely clean empty nest was taken as a sign that the nest had failed with the eggs being removed by predators either before or after desertion. As seasons progressed it became increasingly difficult to distinguish between late first clutches and repeat clutches by pairs replacing earlier failures. Rather than introduce any false precision by trying to split the results for first clutches from those for repeat nesting attempts, all recorded breeding attempts have been included in the analysis. We consider the implication of this practice for our results in the discussion. Because of the difficulty of following the post hatching progress of nidifugous young, attempts were made to monitor progress by examining all groups of Lapwings in the study area until early July in an attempt to count the number of fledged young produced. In the absence of colour ringing, the total number of fledged young necessarily had to be taken as the sum of the maximum counts recorded at each flock or group location, with several counts being made at each site during late June and early July. These counts of juvenile birds were assumed to represent the minimum production from each of the sites, and it was accepted that they would underestimate the total production of fledged young. As observed in other studies (Redfern 1982, Galbraith 1988, Johansson & Blomqvist 1996) chicks were found to move considerable distances, often immediately after hatching, and although no attempt was made to colour ring chicks it was assumed that there was little chance of movement in and out of the study sites prior to fledging. This was not unreasonable because the study sites were selected because they represented islands of suitable habitat containing loose colonies of breeding Lapwings. Where study sites were adjacent to other equally suitable areas there was usually a physical barrier to interchange such as the A9 trunk road, the railway, the River Spey or an extensive area of unsuitable habitat. Both immigration and emigration are unlikely, therefore, to have been significant. Description of the study area The study area lay between the Cairngorm Mountains to the east and the Monadhliath Mountains to the west in the Central Highlands of Scotland. It stretched along approximately 10km of the River Spey (approximately 230m AOD) in the Badenoch district of Inverness-shire. The 8 study sites ranged in size from 7ha to 86ha and are mapped and described in Fig 1, Table 1, respectively. 100 P French, H Insley, G Siriwardena & N Buxton SEZ) Figure 1 Map of the area of Strathspey in which breeding Lapwings were studied between 1996 and 1998. Newtonmore 3 km Table 1 Description of the sites around Newtonmore, Strathspey, used to study Lapwings between 1996 and 1998. Site Area(ha) Altitude (m) Land cover and drainage status above river flood plain A 68 0-10 Golf course on light sandy soils with closely mown greens and fairways and semi natural floristically rich rough grading into agricultural ground to the west where the ground becomes more acid and peaty, changing through unim- proved wet heath to damp improved grassland B 38 10 Flat semi improved grassland, with wetter, more rushy areas to the north. Much of the area is uniform vegetation, with an isolated clump of trees. Grazing regime Golf course ungrazed but closely mown. To west heavily grazed by sheep throughout the year. Heavily grazed by sheep and cattle. Divided into a number of smaller paddocks probably used most intensively at lambing Scottish Birds (2000) € 18 10-20 H 86 >40 Breeding success of Lapwings in Strathspey 1996-1998 101 To the south a fenced paddock of improved mossy grassland. Slightly lower ground to the north is wetter grass and rush pasture. An extensive area of semi natural grassland on the western floodplain of the Spey, interspersed with a number of wetter, more rushy areas. East of the railway semi natural grassland, on the river floodplain. To the north lies a small boggy area and pools formed from relict river channels. West of the railway the area is fairly wet improved grassland. Flat improved grassland lying either side of a marshy burn. Open wet sedge and grassland adjacent to the river with trees along one side. To the east mixed arable and improved grass- land, but grades upwards to the north west through semi natural wet grassland and felled Intensively stocked with sheep at the south end. The north end was grazed and poached by by cattle and sheep. Grazed by sheep. Probably most intensiv- ely during lambing. Heavily grazed through out the year and poach- ed by sheep and cattle in winter. Used as a lambing park in spring. Grazed but not heavily. Heavily grazed and tracked by sheep. Arable ungrazed during the breeding season. Improved grass heavily woodland into wet moorland grazings and grouse MOOr. Analysis Breeding productivity per nesting attempt across the 8 sites and 3 years was assessed by estimating clutch and brood sizes, hatching success and daily nest failure rates. The latter were calculated using a formulation of the Mayfield method (Mayfield 1961 & 1975) which takes into account the number of days over which anest is monitored and therefore avoids any bias resulting from the absence from a data set of nests that failed before they could be found. All variables were calculated as a property of individual nesting attempts. Nests in which no eggs hatched (for any reason) were treated as whole nest failures and hatching success was calculated as the proportion of the clutch that hatched successfully in nests that did not fail. All the variables taken from the nest data were analysed using generalized linear models in the GENMOD procedure of SAS (SAS Institute, Inc 1996), with site and/or year as categorical independent variables. grazed by sheep and rabbits. Grouse moor less heavily grazed. Daily nest failure rates were estimated using a logit linear model with a binomial error term, in which success or failure over a given number of days (as a binary variable) was modelled with the number of days over which the nest was exposed during the egg period as the binomial denominator (Crawley 1993; Etheridge et al 1997; Aebischer 1999). The number of exposure days during the egg period was calculated as the mid points between the maxima and minima possible given the timing of nest visits recorded (note that exposure days refer only to the timespan for which data were recorded for each nest and do not represent the full length of the egg period). Hatching success was also modelled with a logit link and binomial errors, brood size forming the numerator and clutch size the binomial denominator. Individually, clutch and brood sizes were modelled with identity links and normal errors. The significance of the variation between sites was tested by comparing the fit of each model incorporating the sites with that of an intercept 102 P French, H Insley, G Siriwardena & N Buxton SB 21(2) only (constant) model using a likelihood ratio test (SAS Institute, Inc 1996). Results Table 2 summarises the results for the 630 nests located over the 3 years 1996 to 1998. Although the sites searched were the same in all 3 seasons there was a sharp decrease from 254 nesting attempts located during the 1996 breeding season to 181 in 1997 and 195 nests located in 1998. In 1997 a heavy snowfall during the first week of April disrupted most of the early nesting attempts, and made accurate recording of the first clutches difficult, so that the number of nesting attempts recorded may not have been a complete record. The mean clutch size laid and number of chicks hatching at successful nests did not vary significantly within or between seasons. The average number of young hatched in nests surviving to hatching ranged from 3.01 in 1998 to 3.19 in 1996. There was, however, great variation in apparent chick survival to fledging, but because of the methods used to count fledged chicks these data have to be treated with caution. The proportion of hatched chicks surviving to fledging ranged from 9.5% in 1996 to 17.5% in 1997. When all 3 years’ data were combined, comparison between the 8 different sites (Table 2) revealed very highly significant differences in the daily failure rate of nests. Average nest survival was calculated as |- (the daily failure probability to the power of the average length of the incubation period in days), converted to a percentage. An incubation period of 27 days (Cramp & Simmons 1983) was assumed for all nests: although this ignores any variation in the length of incubation, it produces figures that are more easily interpretable than daily failure rates. For all 3 years combined nest survival in sites C, Fand H was above 70%, whereas in sites B, D, and Eit was below 50%, with sites A and G intermediate at 63%. When all sites were lumped together comparison of nest failure rates between years showed that the daily failure rate was remarkably consistent. There was a very slight trend towards increasing nest survival, from 58.7% in 1996 to 64.9% of nests surviving to hatching in 1998 (Table 2). However, lumping the data in this way masked considerable variation in nest failure rates between and within sites in different years. When the significance of the interaction between sites and years was tested, to see whether different temporal variation had occurred across the different sites, the interaction term was highly significant (Table 3). Nest failure as a result of predation was usually difficult to identify with certainty because the evidence was usually an empty nest with no sign of what had caused the egg loss. Identifiable predation accounted for 3.8% of all nest failures, and the true figure was almost certainly much higher since many unidentifiable cases would have been attributed to the unknown category (Tables 4 & 5). In 1998 when 40 separate nesting attempts were found in site B only one was lost as a direct result of agricultural activity, being crushed by a tractor, a further 17 simply disappeared and although attributed to ‘unknown’ were probably predated. The high number of clutches found in that site and year was almost certainly the result of replacement clutches laid after these losses. Site D was adjacent to an active Common Gull (Larus canus) colony and suffered a particularly high failure rate in 1997. Although this was again noted as being due to unknown causes it was almost certainly due to the gulls taking clutches. The high rates of clutch loss at some sites in 1997 may have been due to delayed laying. As noted above heavy snowfall during the first week of April resulted in the loss of most early clutches and may have resulted in the subsequent replacement clutches coinciding with a peak in the food requirements of predating species such as corvids. In the Dombes Region of France, Broyer & Benmergui (1998) found that early clutches were less susceptible to predation and had higher hatching rates than late clutches. In site E, 9 out of 20 clutches were destroyed as a Scottish Birds (2000) Breeding success of Lapwings in Strathspey 1996-1998 103 Table 2 Summary of the data collected on Lapwings breeding around Newtonmore, Strathspey between 1996 and 1998 and comparison of nest failure rate across the 8 study sites over the 3 years (Likelihood Ratio Test: Chi squared=33.12, 7df, p=0.0001). Methods used to calculate the various parameters are described in the text. Site (Fig 1) A B € D E F G H Total 1996 No clutches 70 29 24 15 20 32 25 39 254 Mean Date Found 24.4 30.4 © 2 29.4 4.5 18.4 14.4 55 26.4 Daily Nest Failure Rate 0.025 0.035 0.011 0.012 0.046 0.014 0.007 0.015 0.020 % Nest Survival 49.8 38.3 73.4 TES 27.8 68.4 S73 67.4 Steel! Mean Clutch Size 3.60 S02 3.50 3.80 3.45 392. 3.88 3.56 3.62 Mean Brood Size 3.15 3.00 2.80 525 325 331 3.40 32 319 Total Young Hatched 154 A9 59 45 39 81 TS 108 610 % Hatching Success# 85.4 87.1 82.4 81.3 86.7 86.0 89.5 87.9 86.1 Total Fledged Young 6 5 8 10 0 8 6 15 58 % Chick Survival 290 10.20 13.56 DEDD 0.00 9.88 8.00 13.89 Oral Prod per Attempt 0.09 0.17 0.33 0.67 0.00 0.25 0.24 0.38 0.23 1997 No clutches 47 30 8 11 11 33 11 30 181 Mean Date Found 23.4 23.4 12.4 2S 25.4 20.4 15.4 22.4 22.4 Daily Nest Failure Rate 0.019 0.020 0.019 0.076 0.015 0.000 0.073 0.010 0.018 % Nest Survival* 59.0 58.0 60.0 RES 66.2 100.0 13.0 76.3 61.6 Mean Clutch Size 3.83 Seah 3.50 3.45 spa 3.82 3013 3.70 3.69 Mean Brood Size 2.19, 2.63 2.50 4.00 3.00 3.18 3.00 Bye 3.02 Total Young Hatched 2 60 33 8 23 126 12 106 480 % Hatching Success# 70.9 67.7 71.4 100.0 75.0 87.1 75.0 84.3 78.8 Total Fledged Young 20 8 5 4 3 16 - 20 84 % Chick Survival 17.86 132335 PANDA: 50.0 1304" 1270 33.33 18.87 17.50 Prod per Attempt 0.43 O27 P3 0.36 0.27 0.48 0.36 0.67 0.46 1998 No clutches 44 40 20 12 9 25 6 39 195 Mean Date Found 28.4 26.4 30.4 4.5 SD) 20.4 eo) 30.4 28.4 Daily Nest Failure Rate 0.005 0.031 0.009 0.021 0.071 0.030 0.010 0.008 0.016 % Nest Survival* 86.6 42.6 78.0 56.6 Bs 43.6 76.5 81.1 64.9 Mean Clutch Size 3.59 Se) 5 3.70 BHI 3.56 3.40 3.67 3.64 3.63 Mean Brood Size 522. 2.88 2.80 3.00 3.00 3:31 3.00 2.83 3.01 Total Young Hatched 133 53 42 21 21 43 19 94 443 % Hatching Success# 87.3 75.4 75.0 Ta TSO 87.8 81.8 78.1 81.0 Total Fledged Young 9 7 5 0 8 13 3 14 57 % Chick Survival 6.77 13.21 5.08 0.00 38:10. 3023 Sai9 14.89 12.87 Prod per Attempt 0.20 0.18 0.15 0.00 0.89 0.52 0.50 0.36 0.29 Failure/survival rates 1996-98 combined Daily Nest Failure Rate 0.0167 0.0281 0.0119 0.0284 0.0412 0.0123 0.0169 0.0105 Lower Confidence Limit 0.0124 0.0207 0.0064 0.0177 0.0264 0.0079 0.0099 0.0068 Upper Confidence Limit 0.0225 0.0381 0.0220 0.0452 0.0636 0.0190 0.0289 0.0163 % Nest Survival* 63.4 46.3 P23 45.9 S271 71.6 63.1 5c # Hatching success was calculated as the % of eggs hatching in nests surviving to hatching. * Nest survival was calculated as the Mayfield daily survival rate to the power of the length of the incubation period in days, So is not equal to the percentage of nests that were successful. 104 P French, H Insley, G Siriwardena & N Buxton SB 21(2) Table 3 Comparison of nest failure rate between years (1996 to 1998) and study sites for Lapwings breeding around Newtonmore, Strathspey (Likelihood Ratio Test: Chi-squared = 70.079, 14df, p = 0.0001). Percentage nest survival was calculated assuming our incubation period of 27 days (see text for details). Site Year Daily nest failure rate A 1996 0.0255 1997 0.0194 1998 0.0053 B 1996 0.0349 1997 0.0200 1998 0.0311 (C 1996 0.0114 1997 0.0188 1998 0.0092 D 1996 0.0125 1997 0.0759 1998 0.0209 E 1996 0.0462 1997 0.0152 1998 0.0714 F 1996 0.0140 1997 0.0000 1998 0.0303 G 1996 0.0072 1997 0.0727 1998 0.0099 H 1996 0.0145 1997 0.0100 1998 0.0077 direct result of agricultural practices in 1996, 5 during field rolling and 4 by stock trampling the eggs. Stocking of farm animals in site E was again intense in 1998 when, out of 9 clutches found, 3 were trampled by stock and 3 simply disappeared. Comparing the causes of nest failure between sites (Table 4) and between years for all sites combined (Table 5) showed that, overall, agricultural operations or stock directly accounted for only 5.0% of nest failures. Discussion Variation in Lapwing nest failure rates has Lower confidence limit Upper confidence Nest limit survival % 0.0173 0.0374 49.8 0.0113 0.0331 59.0 0.0022 0.0127 86.6 0.0199 0.0604 38.3 0.0108 0.0367 58.0 0.0197 0.0488 42.6 0.0043 0.0300 joe 0.0061 0.0565 60.0 0.0030 0.0281 78.0 0.0040 0.0380 (ales, 0.0400 0.1396 11.9 0.0087 0.0492 56.6 0.0264 0.0797 27.8 0.0038 0.0585 66.2 0.0300 0.1604 13.5 0.0073 0.0266 68.4 0.0000 0.0000 100.0 0.0168 0.0538 43.6 0.0027 0.0190 82.3 0.0368 0.1387 13.0 0.0014 0.0666 76.5 0.0076 0.0276 67.4 0.0041 0.0237 76.3 0.0035 0.0171 81.1 repeatedly been related to intensity of agricultural management, leading either directly to clutch destruction through damage by agricuitural machinery and operations, or trampling by stock (eg Shrubb 1990; Berg et al 1992; Triplet et al 1997; Broyer & Benmergui 1998), or indirectly to high predation rates on fields where nests were exposed through cultivation or intensive grazing pressure (Baines 1990), although Galbraith (1988) found that predation rates were higher on rough gazing than on arable land. Although there were significant differences in the nest failure rates within sites between years and between sites within years in this study, overall nest survival, for all Scottish Birds (2000) Breeding success of Lapwings in Strathspey 1996-1998 105 Table 4 Causes of failure for Lapwing nests on 8 study sites around Newtonmore, Strathspey over the 3 years 1996 to 1998 (all 3 years combined). Note that the simple percentage success rates given here are not equivalent to the Mayfield derived nest survival rates in Tables 2 and 3. Area Numbers and percentage of nests Successful Failing due to specific causes Total Agric Trampled Other Predation Unknown A No 118 6 0 5 14 18 161 % 13 Se7/ 0 Bel 8.7 pie2 B No 59 2 1 3 3 31 99 % See! 2.0 1.0 3.0 3.0 31.3 C No 53. 0 3 2 2 3 52 % 80.8 0 5.8 3.8 3.8 5.8 D No 21 1 0 1 1 14 38 % 55:3 2.6 0 26 2.6 36.9 E No 22) 5 5 0 0 9 40 % S22) 125 125 0 0 DDS F No 70 3 2 0 1 14 90 % 77.8 33 DD. 0 es 15.6 G No 29 0 0 3 0 10 42 % 69.1 0 0 Tell 0 23.8 H No 88 1 2 2 3 12 108 % 81.5 0.9 1.8 1.8 2.8 2 Total No 448 18 13 16 24 111 630 % las! 2.9 7 DS 3.8 17.6 sites and years combined, was around 60%. It is not known whether this rate of nest survival was adequate to ensure sufficient reproduction of young Lapwings to replace post fledging and adult mortality. However, while 40% of nesting attempts failed between laying and hatching, the apparent loss of chicks between hatching and fledging (from 82.5 to 90.5%) appears to have been far more significant and appeared to be the main factor determining success in Lapwings in this study. Analysis of Lapwing mortality rates in Denmark using ringing recovery data has indicated that, to maintain a stable population, each breeding pair needs to produce an average of 1.18 fledged young per annum (Bak & Ettrup1982), while an estimate of mortality rates in British Lapwings using ringing recovery data from 1963-1992 (Catchpole etal 1999) calculated that 0.56 fledged young per year were required for the population to maintain itself (note that this assumes that 50% of females do not breed until they are 2). Over the 3 years of this study the productivity ranged from 0.23 young per attempt in 1996 to 0.46 in 1997. Lapwings are single brooded, although they will replace clutches lost during the first half of the incubation period (Hegyi & Sasvari 1998). Overall nest survival was around 60% so that at most 40% of the nesting attempts recorded could have been replacements. Adjusting the figures to allow for this potential margin of error indicates that even if 40% of all nests were replacements, the productivity was no greater than 0.38 in 1996, 0.48 in 1998 and 0.77 chicks per breeding pair in 1997. These estimates suggest that only in 1997 was productivity sufficient to replace mortality. For the study population to have been self sustaining, our estimates of productivity per nesting attempt suggest that each pair must have been failing at least once, and probably twice, on average, before making a successful breeding attempt — 106 P French, H Insley, G Siriwardena & N Buxton SB 21(2) Table 5 Comparison of causes of nest failure between years (1996 to 1998) for Lapwings breeding around Newtonmore, Strathspey (all sites combined). Note that the simple percentage success rates given here are not equivalent to the Mayfield derived nest survival rates in Tables 2 and 3. Year Numbers and percentage of nests Successful Failing due to specific causes Total Agriculture Trampled Other Predation Unknown 1996 No 176 6 3 5 12 Sy) 254 %o 69.3 2.4 12 1.9 4.7 20.5 1997 No 131 6 5 6 32 181 %o 72.4 33 2.8 0.5 333 7a! 1998 No 14] 6 5 10 6 aT] 195 % 253 Sol 2.6 Del oll 13.8 Total No 448 18 13 16 24 eal 630 % Tile! 2.9 Dail 25) 3 Ji 2. 8 17.6 this seems unlikely, not only because the length of the breeding season is limited, but also because small chicks were not found late in the summer. The nest survival rate for nests over the 3 year period remained consistent at around 60%, which suggests that the decrease in nesting attempts recorded was likely to have reflected a decline in the number of nesting pairs present. The alternative hypothesis that the number of pairs present remained constant, but that each pair made fewer nesting attempts per annum was not credible. Baines (1990) showed that Lapwing breeding densities decreased by 74% on pastures and by 56% on meadows following agricultural improvement, and that fledgling production was 63% lower on improved areas. Although the level of stocking in the study areas was intensive over all 3 years, no changes in agricultural practice were apparent. Examination of the causes of nest failure (Tables 4 & 5) suggest that agricultural practices led directly to the destruction of only around 5.0% of all nesting attempts, although it was very likely that they also led indirectly (through lack of cover) to losses by predation. In Baines’ (1990) study almost twice as many simulated clutches were predated within 24 hours on improved pastures as on unimproved. At lower breeding densities, the strength of communal nest defence is reduced such that all nests become more vulnerable (Berg, 1996). However, a direct nest failure rate of around 40% does not appear sufficiently low to account for the observed population declines. Indeed, clutch size, brood size at hatching and nest survival all appeared to be sufficient, across the years considered, to maintain the breeding population (but not to increase it). Chick mortality, however, does appear to have been high. Although there was no evidence to indicate the age after hatching at which chicks were being lost, adverse weather is more likely to have affected the survival of young than older chicks. Younger chicks will have been more vulnerable to predation, although both these effects would be very difficult to measure in a nidifugous species like Lapwing. These factors were likely to have contributed to the between year variations in chick survival (Table 2). Although we have not attempted to relate farming practice to chick survival in this study, these relationships have often been investigated previously (eg Galbraith 1988; Shrubb 1990; Reichholf 1996; johansson & Blomgvist 1996; Triplet et al 1997) and it is clear that under the current agricultural regimes Scottish Birds (2000) Breeding success of Lapwings in Strathspey 1996-1998 107 practised in this part of Speyside, Lapwings are unable to breed successfully enough to maintain their numbers. Although there was an increase in the annual total of nesting attempts in the study area from 181 to 195 between 1997 and 1998, there was a 23% reduction in the number of nests found between 1996 and 1998. A 28% reduction in breeding Lapwing abundance was observed on Scottish Breeding Bird Survey squares between 1994 and 1998 (Noble et al 1999). While it may be difficult to make a valid direct comparison between the BBS figures and this 3 year study in a limited area, both figures suggest that the current situation where over 50% of all breeding Lapwings in Britain are found in Scotland (Wilson & Browne 1999) may be only a temporary phenomenon, and that, despite its currently relatively healthy size, the Scottish population cannot be regarded as a source potentially supplying other UK populations. Acknowledgements The fieldwork leading to this paper was carried out by Patrick French who died on March 22 1999. We are indebted to Mrs Ann French who made Pat’s data available to us so that this paper could be written. The study, which was carried out over 3 breeding seasons, required the cooperation of a great many people, and this was willingly given by landowners, especially John and Eira Drysdale of Ralia, Donald MacKenzie of Lochbuie Croft, David Morris of Ruthven Farm, Donald Munro of Pitmain and the committee and ground staff of Newtonmore Golf Club throughout the work. Fig 1 was drawn for us by John Williams. Bob Swann and an anonymous referee provided helpful criticism of the earlier drafts of this paper. Finally, the generous financial support provided for this study by the Scottish Ornithologists’ Club over the 3 years 1996 to 1998 is gratefully acknowledged. References Aebischer N J 1999. Multi way comparisons and generalised linear models of nest success: extensions of the Mayfield method. Bird Study 46: S22-S31. Baines D 1989. The effects of improvement of upland marginal grasslands on the breeding success of Lapwings Vanellus vanellus and other waders. Ibis 131: 497-506. Baines D 1990. The roles of predation, food and agricultural practice in determining the breeding success of the Lapwing Vanellus vanellus onupland grasslands. Journal of Animal Ecology 59: 915-930. Bak B & Ettrup H 1982. Studies on migration and mortality of the Lapwing Vanellus vanellus in Denmark. Danish Review of Game Biology 12: 1-20. Berg A, Lindberg T & GunnarKaellebrink K 1992. Hatching success of Lapwings on farmland: differences between habitats and colonies of different sizes. Journal of Animal Ecology 61: 469-476. Berg A 1996. Predation on artificial, solitary and aggregated wader nests on farmland. Oecologica 107: 343-346. Broyer J & Benmergui M 1998. Lapwing Vanellus vanellus breeding in the Dombes Region: productivity and factors of failure. Gibier Faune Sauvage 15: 135- 150. Catchpole E A, Morgan J T B, Freeman S N & Peach W J 1999 Modelling the survival of British Lapwings Vanellus vanellus using ring-recovery data and weather covariates. Bird Study 46: S5-S13. Cramp S & Simmons K EL (eds) 1983. The Birds of the Western Palaearctic. Volume 3 OUP, Oxford. Crawley M J 1993. GLIM for Ecologists. Blackwell Scientific, Oxford. Etheridge B, Summers R W & Green R E. 1997 The effects of illegal killing and destruction of nests by humans on the population dynamics of the hen harrier Circus cyaneus in Scotland. Journal of Applied Ecology 34: 1081-1105. Galbraith H 1988. Effects of agriculture on the breeding ecology of Lapwings Vanellus vanellus. Journal of Applied Ecology 25: 487-504. Hegyi Z & Sasvari L 1998. Components of fitness in Lapwings Vanellus vanellus and Black-tailed Godwits Limosa limosa during the breeding season: do female body mass and egg weight matter? Ardea 86:43-50. Johansson O C & Blomqvist D 1996. Habitat selection and diet of Lapwing Vanellus vanellus 108 P French, H Insley, G Siriwardena & N Buxton SB 21(2) chicks on coastal farmland in S.W. Sweden. Journal of Applied Ecology 33: 1030-1040. Mayfield H 1961. Nesting success calculated from exposure. Wilson Bulletin 73: 255-261. Mayfield H 1975. Suggestions for calculating nest success. Wilson Bulletin 87: 456-466. Noble D G, Bashford R T, Marchant, J H, Baillie, S R & Gregory R D 1999. The Breeding Bird Survey 1998. Report number 4. British Trust for Ornithology, Joint Nature Conservation Committee and Royal Society for the Protection of Birds. O’Brien M 1996. The numbers of breeding waders in lowland Scotland. Scottish Birds 18:231-241. Redfern C P F 1982. Lapwing nest sites and chick mobility in relation to habitat. Bird Study 29: 201-208. Reichholf J H 1996. Population crash of the Lapwing Vanellus vanellus in the Lower Bavarian Valley of the River Inn. Ornithologischer Anzeiger 35: 173- 179. SAS Institute Inc. 1996. SAS/STAT Software: Changes and Enhancements through Release 6.11.SAS Institute Inc. Cary, NC. Shrubb M. 1990. Effects of agricultural change on nesting Lapwings Vanellus vanellus in England and Wales. Bird Study 37: 115-127. Triplet P, Durant J & Bacquet Ss ioo7 ihe reproduction of the Northern Lapwing Vanellus vanellus and farming techniques: characteristics of breeding sites in maritime lowland Picardy (northern France). Alauda 65: 121-129. Wilson A 1999. Tumbling Lapwing populations. BTO News 224: 14. Wilson A & Browne S J 1999. Breeding population estimates for Lapwing, Oystercatcher, and Curlew in Scotland: results of the 1998 BTO Lapwing Survey. Scottish Birds 20: 73-80. The late Patrick French, Grianan, Strone Road, Newtonmore, Inverness PH20 1BA Hugh Insley*, 1 Drummond Place, Inverness IV2 4JT Gavin Siriwardena, British Trust for Ornithology, The Nunnery, Thetford, Norfolk IP24 2PU Nigel E Buxton, Westholm, Station Road, Newtonmore, Inverness PH20 IDH * corresponding author Revised manuscript accepted August 2000 — wwe Lapwing at nest scrape : aby ~ 4 ie Bobby Smith Scottish Birds (2000) 21:109-115 109 Observations of wintering Ring Ouzels and their habitat in the High Atlas Mountains, Morocco DS CARTHUR, PRELLIS, RG LAWIE & M NICOLL A study of the breeding biology of the Ring Ouzel in Glenesk, Angus, which involves the ringing of both chicks and adult birds, has yielded one recovery from Morocco. The High Atlas Mountains of North Africa are considered to be one of the Ring Ouzel’s more important wintering grounds. To date 11 British ringed birds, including 2 Scottish ringed chicks, have been recovered from Morocco and 2 from Algeria, one a chick ringed in Scotland. In January 2000 members of the Tay Ringing Group visited Morocco and located wintering flocks feeding on Juniper bushes along the arid northern foothills of the High Atlas Mountains. Introduction The principal wintering quarters of the Ring Ouzel (Turdus torquatus) have been identified as Southern Spain and Northwest Africa (Blondel 1962, Niethammer 1955, Zamora 1990). The first arrivals on wintering grounds in Northwest Africa reach the Middle Atlas Mountains from early October and the High Atlas from late October (Thevenot etal in prep). The information from the British Trust for Ornithology Ringing Scheme of North African recoveries of Ring Ouzel from Great Britain totals 11 from Morocco and 2 from Algeria. (Figure 1, Table 1). Although the sample size is small some conclusions can be drawn from the dates and ringing locations. The 3 Orkney and Kent ringed birds according to the criteria used by Durman (1976) are of possible Scandinavian origin. As far as the Lincolnshire, Isle of Man, Yorkshire and Sussex birds are concerned, no assumption of their origin could be made. Of the total recoveries only 5 (4 chicks and one juvenile) can definitely be attributed to the British population and 3 of the chicks and the juvenile were ringed in Scotland. The limited recoveries in North Africa are widespread and no conclusions are possible apart from their preference for the upland areas in the Atlas Mountain region. The recoveries are possibly limited by the terrain and the unrepresentative distribution of observers in the region. On the wintering grounds the birds occur in open coniferous woodland on bare stony slopes and are especially abundant in Phoenecian Juniper Juniperus phoenicea and Spanish Juniper Juniperus thurifera or in mixed woodland of Holm Oak Quercus ilex and Prickly Juniper Juniperus oxycedrus, often near sources of water (Thevenot et al, pers comm). Juniper berries are known to be the main food source in North Africa (Blondel 1962, Heim De Balsac 1931). A more comprehensive study of the birds’ winter feeding, carried out in the Sierra Nevada in South East Spain, showed that juniper berries from Juniperus communis make up around 90% of the diet (Zamora 1990). The breeding biology of the Ring Ouzel is currently being studied in the area around Invermark, Glenesk, Angus, Grid Reference NO4380 (Arthur 1994). Although the Invermark study area appears to hold a stable population, Ring Ouzels are in decline in other areas of Scotland, notably the 110 DSC Arthur. P R Ellis, RG Lawie & M Nicoll SB 21(2) gi tur fe Ring Ouzel Recovery sites from S)<- “ . Okn Morocco ard Algeria, 1928 - 1998. a : pee / oe Ques ae % +04 e Mees ‘ / ; Mecneric 0+ ‘2 { ‘ : ; =-Ksiba_ _ e Melt Morocco . Algeria @ Tounfite x €4st° ee ae > t-5 os Sg LU alka pueaiet : £ » Morraicect azlol 1 ‘ ae ene” on +3 * Caulorneder: oy Ae a +s “2 pee, «m 240 ae ne Se _- des 33 ae aes + Ring Dezel recovery sites ¢ see “atte (| ‘or detois southern uplands, as well as in Wales and partsof Aims England. Comparison between the 1993 BTO Breeding Atlas and the 1999 RSPB Ring Ouzel survey indicates a reduction in the occupancy of breeding territories of around 40% (Wotton pers comm). Inrecent years the leading UK Conservation Groups have placed the Ring Ouzel on their Amber List, the criteria being a decline between 25%-49% in the breeding population orrange over the previous 25 years. The reason for the decline is not apparent, but the studies currently underway in the Moorfoot Hills, Glenesk and Glen Clunie may give further information (Burfield pers comm, Rebecca 2000). The Angus Glens in the East Grampians are probably one of the most important breeding grounds in Scotland for Ring Ouzel. They currently appear to hold a stable population in a habitat which has not changed for some considerable time. While we are improving our knowledge of the breeding biology of this species, we felt it would be beneficial to extend our study to observing the birds in winter. As the High Atlas Mountains are considered to be one of the Ring Ouzel’s more important wintering grounds, a trip to Morocco was organised by 4 members of the Tay Ringing Group from 23 January to 2 February 2000. The objectives of the trip were to locate wintering flocks, study their habitat use, identify their main food source and record the altitude they were found at. Whenever it was practical, birds were checked for colour rings in the hope of sighting birds from any of the 4 British study areas: Angus, Aberdeenshire, Lothian and Shropshire. A further aim was to identify potential mist netting sites for a possible future follow up ringing study, should permission be granted by the Moroccan authorities. Logistics and methods Plans for the trip developed very quickly following informal discussions during the 1999 Scottish Ringers’ Conference in November. A great deal cels in the High Atlas Mountains, Morocco 111 > ~ Wintering Ring Ou Scottish Birds (2000) BLAs[Y “NOZNG ZI], “AOIg JOT 0661/01/r7 gyeul yjnpy AQUYIC ‘Avsp[euoy YON 686I/PO/ZI EI BLIO8[Y “BLIOYyooI] 6C61/€0/S I sulpsou = ABMOT[LD 2 solu ‘[EuUooyIy = YZHI/SO/BT TI ODDOIOW ‘SIAI ZNOPIOGY IN ‘noppe x WV 8661/1 1/SO SUIp]SoU atysusapsaqy “IIUNTD UID ~—s- BH6I/SO/ZZ_—T OI90I0J ‘epiNC IN ‘UOIsoy vpeioL C661/1 1/7 optuoans SnsUY ‘99J US[H “YILULIOAUT =» $6H6HI/60/0I_ OL OdIOIOY “[ET[IIN Hugg “eypeL, eqsey 6861/r0/01 ayeul yNpy KOUYIO ‘Avsp[euoy YON = 886I/P0/FT «6 OIDOIJOP ‘pers nog “sITTV [7 8861/£0/S1 ayeu yNpYy SSOUYPED “YOIM “UR SSON L86I/OI/IZ 8 OIDOIJO[Y “VIWUTUTNOD “Jonsnoury T861/T 1/17 suly}sou uviIpoT ‘yous 76/90/87 LL OIDOIOY ‘[RPIZY IN ‘IUNSNo[L], Z861/Z0/F1 SUT[}SOU SIEM “JUOMH ‘IesapauL TYHL OS6I/SO/ET 9 ODD0IOY| “JIUTY IN LLOITI/EI ayeur ofrusAne xossng = LL61/60/0l_ = § OID0JOJ ‘[PIIZY IN OLOL/TI/SI ayeuuay INpY Woy = SL6I/SO/8ST—s dv ODIOIOY “SO[-U-IZL, 8961/71/01 yeu ofruAne AIYSYIOXK —-NGI/60/9T_—_E OID0IOY ‘[PTIZY IN C96I/TI/SI geluloy o[tuoAne Ur JO AST JID OUL €96I/OI/EO Z oI90I0J ‘epino 7961/70/97 yeu yNpy AITYSUTOOUTT “WUTOg IEYCIGIQ = GGGT/6O/LI I uOlB00'T dueq xas/asV uoneo00'T 3ueq ON sjlvjap SUIpUL] s[lvjop Sursuly "DIY YJAON UWOAL SAIMIAOIAA J92NQ SUIY { IQeI, 112 DSC Arthur. P R Ellis, RG Lawie & M Nicoll SB 21(2) of research was done including gathering information from previous trips to the High and Middle Atlas (Ryall, 1993 pers comm, Smith, 1965, Allport pers comm). From the information available, target areas were chosen. These were: Tounfite (1941m altitude, lat 5 15°W, long 32 28°N), the pass of Tizi-n-Test (2092m altitude, lat 8 15°W, long 30 50°N), and Oukaimeden (2650m altitude, lat 7 50°W, long 31 7°N). Maps of a selection of areas of the High Atlas were obtained from Hamish Brown who operates the Atlas Mountains Information Services from his base in Burntisland, Fife. For general travelling we used the Michelin Morocco map 959. For more detailed work, maps at a scale of 1:100,000 and 1:200,000 were used. These were produced by the Moroccan military in the 1960s and were of varying quality but the best available. A 4 wheel drive vehicle with a driver and cook was hired through Hotel Ali in Marrakech. Their help with navigation, language, security, provisions, camping and hotel accommodation proved invaluable, allowing us to spend most of our time in the field. Survey work was carried out in 4 areas: Bou-mia, south of El-Ksiba, Tounfite, Oukaimeden and Tizi-n-Test. Observations were carried out around these areas in various habitats wherever vehicular access was possible. Regular stops were made on the routes chosen and excursions on foot, of one to 3 hours, were made to observe and record all birds seen. The route taken during the survey work is shown in Figure 2. Summary of sightings of Ring Ouzels and other thrushes in the High Atlas Mountains together with habitat notes 24 January 2000 Familiarisation tour from El-Ksiba south to El-Arba, east to Aghbala, north to E] Khemis and back to El-Ksiba (175 km). Several areas of Prickly Juniper located but no sightings of Ring Ouzel, one Redwing Turdus iliacus, 10+ Blackbirds, Turdus merula. 25 January 2000 13.00-17.00 20-30 km south of Tizi-n-Isly - along route 1903 to Imilchil (1500 to 1700m) 15+ Ring Ouzels encountered in open mixed woodlands, Prickly Juniper, Phoenician Juniper and Holm Oak with some Aleppo Pine Pinus halepensis. Also observed flocks of 20+ and 50+ Redwings, 3 Mistle Thrushes Turdus viscivorus and 4 Blackbirds. 26 January 2000 18.00 On late arrival at Tounfite area (1950 m) 2 Ring Ouzels plus 20+ Redwings, 2 Blackbirds north of Tounfite. Ring Ouzels seen on dry stony slopes with a scattering of Juniper, mostly Phoenecian along river banks. 27 January 2000 09.30-11.30 14.30 15.40 OnslopesofBouChouari, immediately south of Tounfite (c1980m) - 6 Ring Ouzel observed feeding on berries of Red-berried Mistletoe Viscum cruciatum growing on Hawthorn Crataegus sp among Prickly Juniper and Holm Oak on dry stony slopes. 2km north of Tounfite west side of road at c1920m (site ‘A’) - 80+ Ring Ouzels, 8 Mistle Thrushes, 5+ Redwings and 6+ Blackbirds actively feeding on Juniper, mainly Phoenecian some of which very heavily laden with berries. At water source close to previous location on the east side of road (site “B’) 10+ Redwings 10+ Song Thrushes 40+ Ring Ouzels drinking at water source, fed by spring in dry gully with steep rocky slopes sparsely covered by mainly Phoenician Juniper. Some of these bushes were heavily encrusted in droppings from Turdus spp, indicating regular use of the site. Scottish Birds (2000) Wintering Ring Ouzels in the High Atlas Mountains, Morocco 113 16.20 2.5km north of Tounfite 2+ Ring Ouzels with 20+ Redwings feeding on Phoenician Juniper and drinking at stream at side of road. 28 January 2000 07.30-10.00 10.30-11.40 12.00-13.00 13.10 14.00-14.40 Revisit to site A 30+ Ring Ouzels, 15+ Blackbirds, 12 Mistle Thrushes feeding on Juniper spp. Revisit to site B 30+ Ring Ouzels, 10 Redwings. 3km north of Tounfite following water course of the Oued Oudrhes downstream from ford, 20+ Ring Ouzels, 6+ Redwings, 6+ Blackbirds, 8+ Song Thrushes coming to water and perching in Hawthorn and other bushes overhanging small pools. Juniper spp and Holm Oak scattered on dry stony slopes. At principal water source 2 - 3km downstream from ford 40+ Song Thrushes, 5+ Redwings, 12+ Blackbirds and 30+ Ring Ouzels drinking from stream where cover of scrub and trees was at its most dense. 6km north of Tounfite - 2 Ring Ouzels 10km north of Tounfite - 3 Ring Ouzels, 3+ Redwings and 2+ Blackbirds. 31 January 2000 10.40-15.30 Route S501 to Tizi-n-Test (2000m) around noon between 118km marker and 128km marker from Marrakech 15 Ring Ouzels, 5 Mistle Thrushes and 2 Blackbirds observed feeding among scattered vegetation of Prickly Juniper and Holm Oak with some Phoenician Juniper and pine. 1 February 2000 8.00-12.30 Route $513 to Oukaimeden (2650m) Previous reports recorded Ring Ouzels in the Oukaimeden area but none were observed during our visit. Moderate amounts of Juniper both Prickly and Phoenician but many trees with few or no berries. 20+ Redwings observed among Phoenician on return from Oukaimeden, plus 5 Blackbirds and 7 Mistle Thrushes. No colour ringed birds were observed. Ring Ouzel survey area Qiteom Cesebiecs Mcrocco 2000. N. €23101#2000) ie h r eee j°--7 Seeyiga C Mace - i aS & 4 — a -# . b Me ¢ Tounfite teri Mee co Bou - a > c : * a ee 5 a > 4 D sand te Casabiancs rd (0210212000) SP esi ee { ee en se ‘ss % Pa Sse = 4 a } om oe . 4 —_— r is | Merrskech \ a io » N “ j i : eA ~ «a Astin Fe) ry 4 a é asOukaimeden “~~=- route of tre trip 4 4 at Po a >eoks over 3009m LS . &0 ~/ & A ih a ¥7izi-n-~ Test Figze 2 miles =) 114. DSC Arthur. P R Ellis, RG Lawie & M Nicoll SB 21(2) Conclusions Our preliminary research had led us to speculate that the destruction of Juniper habitat in parts of Morocco may be directly linked to the decline of the Ring Ouzel in the UK. Certainly, there are cases of deforestation in the Middle Atlas (Ryall and Green, 1994). Morocco is a poor country with an expanding population and wood is an important source of domestic fuel. However, our observations showed that in most cases the removal of timber was limited to the scale dictated by the use of basic hand tools and local transport by donkeys. A large percentage of the Juniper trees had evidence of branches being chopped away over along period of time, leaving partly pollarded, gnarled stumps, many probably of considerable age. Very little evidence was found of complete removal of trees. Likewise, the Holm Oak trees were being managed in a similar way but in this case the younger branches carrying the less prickly leaves higher up in the trees were being gathered as fodder for goats, sheep and cattle. Natural regeneration, although limited in scale, appeared to be sufficient to maintain a stable tree density. Overall the wooded areas visited seemed to be ‘managed’ ona sustainable level, and the Berbers were living in harmony with their environment. Due to the time scale and distances travelled, only a small number of sample sites were investigated and a relatively short time was spent in each area of suitable Ring Ouzel habitat. However, the weather conditions were excellent during the trip, which allowed useful observations to be made. Encouraging numbers of Ring Ouzels were observed once we had identified suitable habitats. These were in areas where we found Juniper to be plentiful, although well scattered, on arid, stony slopes (principally north facing) at altitudes between 1500m and 2000m. Two species of Juniper, Prickly and Phoenician, provided the favoured berries on which most of the Ring Ouzels were observed feeding. These berries seemed to form a very high percentage of their diet, and their dry, resinous nature seemed to necessitate frequent intake of water to aid digestion. Consequently, as water was generally in short supply, concentrations of birds were observed drinking at pools, especially where suitably located bushes and small trees afforded plenty of cover. The birds seemed to show a preference for the berries of Phoenician Juniper and this species was always present where larger numbers of the birds occurred. In some areas, juniper trees were encountered with little or no berry crop. The trees may well show annual variations of berry production with bumper crops followed by lean years. Once the birds have stripped the trees in an area they presumably disperse in their search for food, resulting in a mobile population with fluctuating numbers in any particular area throughout the winter months. Most of the birds in the Juniper scrubland were thrushes. Atleast 213 Ring Ouzels were found in the short time that was spent in suitable habitat, along with at least 161 Redwings, 58 Blackbirds, 58 Song Thrushes and 27 Mistle Thrushes. Whenever possible birds were checked for colour rings but none were seen. The largest concentrations of Ring Ouzels, totalling a minimum of 175, were observed just north of Tounfite. However, we consider it likely that similar numbers of birds are to be found wintering in suitable habitats anywhere along the northern foothills of the High Atlas between Tounfite and Tizi-n-Test some 350km to the southeast. Acknowledgements We would like to thank all those who assisted with information for our trip particularly the following Scottish Birds (2000) Wintering Ring Ouzels in the High Atlas Mountains, Morocco 115 people G Allport; C Bowden; H Brown; I J Burfield; M D Jones; B Lynch; G Rebecca; C Ryall; D Tattersfield, R Vernon and the British Trust for Ornithology Ringing Scheme for the ringing recoveries. References Arthur D S C 1994. Breeding Ring Ouzels Turdus torquatus in Glen Esk, Tayside, 1992-1994: Tay Ringing Group Report 1992-93: 2-3. Blondel J 1962. Migration prénuptiale dans les Monts des Ksours (Sahara septentrional). Alauda 30(1):1- 6,817. Durman R F 1976. Ring Ouzel Migration Bird Study 23:197-205. Gibbons D W, Reid JB & Chapman RA 1993. The New Atlas of Breeding Birds in Britain and Treland: 1988-1991. London, T. & A. D. Poyser. Heim de Balsac H 1931. Les lieux d’hivernage du Merle a plastron Turdus torquatus en Algérie Alauda 3:250-256. Niethammer G 1955. Das nordwestafrikanische Winterquartier der Ringdrossel Turdus torquatus Die Vogelwarte 18:22-24. Rebecca G W 2000. The contrasting status of the Ring Ouzel in 2 areas of upper Deeside north east Scotland between 1991 and 1998. Un press, Scottish Birds Summer 2001). Ryall C 1994. Preliminary report of an investigation of the ecology of wintering Ring Ouzels in Morocco. Department of Environmental Management, Farnborough College of Technology. Ryall C & Green M 1994. Le Merle a plastron (Turdus torquatus); espece europeenne hivernante peu connue au Maroc Porphyrio 6:3-6. Smith K D 1965. On the birds of Morocco. Ibis 107:493-526. Thevenot M, Vernon R, Bergier P (in prep) The Birds of Morocco. _BOU Check list. Zamora R 1990. The fruit diet of Ring Ouzels (Turdus torquatus) wintering in the Sierra Nevada (South-East Spain). Alauda 58(1):67-70. DSC Arthur, 12 Dundee Street, Carnoustie, Angus, DD7 7PD P R Ellis, Barry Mill, Barry, Angus DD7 7RJ RG Lawie, Eyrie, Knowes Loan, Barry, Angus DD7 7RF M Nicoll, 1 Banknowe Grove, Tayport, Fife DD6 9LH Revised manuscript accepted August 2000 Ring Ouzel at nest site William Brotherston 116 Scottish Birds (2000) 21:116-118 SB 21(2) SHORT NOTES Prey captured and attacked by Merlins in winter It was suggested in Dickson (1988 British Birds 81: 269-274) that more study was needed on winter prey of Merlins Falco columbarius in Britain. Inasubsequent quantitative note Dickson (1992 Scottish Birds 16: 282-284) detailed 270 hunts from 1965-92 where it was shown that 2 bird groups figured largely in the results by number: Skylarks Alauda arvensis and finches Fringillidae. The following is based on observed hunts in west Galloway to see if there has been any significant changes since 1992 in the composition of prey species attacked and captured by Merlins in winter. A further 104 hunts were recorded between 1992 and 2000 (Table 1). All the attacks by blue and brown Merlins were directed at species usually associated with low ground and open country in winter in various habitats, mostly in farmland. Most hunts involved similar techniques as those detailed in Dickson (1996 Scottish Birds 18: 165- 169) with low level attack flight predominating by both blue (79%) and brown (77%) Merlins. Blue Merlins were successful in 20% of hunts and brown Merlins in 16%, within the range of results given in Dickson 1996. The prey species regularly attacked and killed between 1965-1992 and 1992-2000 are remarkably consistent with Skylarks and finches being the most frequent targets in winter. Blue Merlins attacked more finches, averaging 21g in weight in both studies (55% and 82%) than any other species. Brown Merlins consistently attacked more Skylarks (40%) averaging 37g in weight, in 1965-1992 but more finches (60%) in 1992-2000 (Table 1). In 1965-92, 5 identified bird species formed more than 5% of kills in order of numbers: Skylark, Starling Sturnus vulgaris, Linnet Carduelis cannabina, Meadow Pipit Anthus pratensis, and Chaffinch Fringilla coelebs. Expressed by weight, Starlings emerged as the most important of kills (25%) with Skylark providing 21% of kills by weight, Lapwings Vanellus vanellus 17.5% and finches 11.8% (Scottish Birds 16: 282-284). In 1992-2000 only 4 species formed more than 5% Table 1 Percentage frequency of prey species attacked and killed by blue and brown Merlins in winter in west Galloway, 1992-2000. Prey species No attacks % by blue frequency Merlins Skylark 3 8.9 Meadow Pipit Starling 1 7253) Chaffinch 6 17.6 Linnet/Twite DD 64.7 Unidentified 2 Se) Totals 34 No attacks % Successful by brown frequency blue brown Merlins 23 32.9 2 2 2.8 2 1 1.4 2 41 58.6 4 7 4.3 1 70 i 11 Scottish Birds (2000) of kills: Skylarks, Meadow Pipits, Chaffinches and expressed by weight Linnets/Twites Carduelis flavirostris emerged as the most important of kills (87%) with Skylarks providing 6% of kills by weight, Chaffinches 4% and Meadow Pipits 3%. Thus blue (males) in winter concentrated on the kills of the smaller bird species (16-37g) while brown Merlins (females and juveniles) seem to kill smaller to larger species (16-214g), although males did not seem to select different habitats to females nor juveniles to adults (British Birds 81: 269-274). There are few other studies of winter diet apart from Warkentin and Oliphant (1990 Journal of Zoology, London 221: 509-563) North American study, which provides the only direct comparison, Short Notes 117 of a wintering population of Merlins in the city of Saskatoon, Canada. There they found that 3 species formed more than 5% (66% by weight) of kills House Sparrow Passer domesticus, Bohemian Waxwing Bombycilla garrulus (26% by weight) and Meadow Vole Microtus pennsylvanicus (5% by weight). They also found a large proportion of prey matched in summer and winter diets. The results in this study reveal a high degree of overlap between summer and winter diets in Galloway. A significant proportion of breeding season prey in Galloway (1965-1999) is similar in composition to that in winter in which 14.3% (11% by weight) of Skylarks and 7.6% (20% by weight) of finches feature behind Meadow Pipits (57%, 24% by weight) in the table of remains found at breeding sites (Dickson, unpublished data). R C Dickson, Lismore, New Luce, Newton Stewart, Wigtownshire DG8 OAJ Manuscript accepted May 2000 Nestling feeding rates by Peregrine Falcons There are few British data on the feeding rates of Peregrine Falcons Falco peregrinus throughout the breeding season. I obtained data at 2 sites on the coast and one inland in Wigtownshire between 1975-98 to provide an indication of the feeding rates in the nestling stage of the breeding cycle. The only other detailed account of nestling feeding rates is that given by Parker (1979, British Birds 72:104-114) of 64 items (number adjusted for delivered items only) in 249 hours of observation (0.26 deliveries/hour) in Wales. Overall, the pooled number of prey deliveries brought and fed to the young in the nestling stage, but not cached, was 23 in 40 hours of observation (0.58 deliveries/hour). Martin (1980, BSc dissertation in Ratcliffe 1993, The Peregrine, London) noted an irregular spread of feeding times throughout the day with no particular pattern. A similar trend was found in prey deliveries in Wigtownshire when it occurred 11 times in the morning, 9 in the afternoon and 3 in the evening with no significant differences in deliveries during the day (x* = 1.02, 2df, NS). During the second half of the nestling stage, when females assisted males from the third week onwards, males still seemed to 118 Short Notes SB 21(2) continue to act as the main provider ina male:female ratio of 4:1. In comparison, the nestling feeding rates in Wigtownshire are higher than those recorded in Wales. Data were also calculated from some dates detailed in Treleaven (1977, The Peregrine, Penzance; 1998, In Pursuit of the Peregrine, Wheathamstead) for Cornwall. His data (delivered items only) suggests that the feeding rate in the nestling stage was 12 items in 32.25 hours of observation (0.37 deliveries/hour) , lower than that obtained in Wigtownshire, but higher than the Welsh data. Interestingly, the nestling feeding rates for Peregrines in Wigtownshire at 0.58 deliveries/ hour, is identical to the feeding for Merlins Falco columbarius at 0.54 deliveries/hour in Galloway (Dickson 1995, Scottish Birds 18: 20-23). R C Dickson, Lismore, New Luce, Newton Stewart, Wigtownshire, DG8 OAJ Manuscript accepted May 2000 Peregrine family Edmund Fellowes a Scottish Birds (2000) 119 OBITUARIES Hetty Louisa Harper 1915 - 2000 Hetty Harper Geoffrey Harper Hetty Harper was born in London and when she was 6 her family moved to Devon where she grew up and went to school. She attended the University College of Exeter from 1933-36, gaining an external general degree from the University of London, and continued until 1937 to gain a teaching certificate in PE. She took up a teaching post in Bermuda in 1937 where she met Bill and they married in 1940. Billhad joined the Bermuda Meteorological Service in 1937 and they remained there until 1947. Bill was then posted to London Airport followed by a series of postings to Shetland, in England, Aden, Germany and, finally, Edinburgh after which he retired in 1977. While in these jobs they travelled widely and particularly enjoyed time on Fair Isle and in Africa and Arabia. Hetty acquired a wide knowledge of birds before she joined the staff of the SOC in March 1975 as a much needed additional part time assistant in our Bird Bookshop. She worked on the staff until 1983. One day Hetty mentioned that Bill, who retired in 1977, was looking for something to do. It was suggested that he might be interested in Library work, he took over from Irene Waterston later that year. Hetty worked with others in the Bookshop and they developed into a good team, which enjoyed its work and also the opportunity to meet members and other customers who came to buy books. A very good rapport was built up with customers due to the friendly welcome by Hetty, her help and advice when choosing bird books for many parts of the world they were either to visit or had come from. After 8 years on the paid staff, Hetty came in regularly with Bill and did much valuable work in the Library. With Bill she always attended conferences of both the SOC and the BTO, helping him with his second hand book sales in aid of Library funds or Bookshop sales. Her knowledge and advice on these occasions was greatly appreciated. When Bill, who had been made an Honorary Member of the SOC in 1987, died in 1995, Hetty was made an Honorary Member on her own merits. Hetty worked voluntarily in the Family Care bookshop in Edinburgh for several years until 1999. She was a very active member of the WRVS and joined the Tuesday morning team which ran the tea bar at the Princes Margaret Rose Hospital for about 15 years until 2000. One particularly enjoyable aspect of the Harper household was an invitation to dinner, normally after a Club committee meeting when Hetty was the perfect hostess. Apart from enjoying her excellent home cooking and hospitality she really enjoyed a good ‘blether’ on a wide range of topics and liked to be kept up to date with all the latest news in the bird world. 120 SB 21(2) Hetty was a remarkable and interesting lady, kind, gentle and caring and with a strength of character and determination which belied her small stature. This was aptly demonstrated by a strenuous trip last summer through Russia by local transport with her son Geoffrey. This was quite a feat for someone in her 80s, but she took it in her stride and enthused about her experiences on her return home. Sadly, recurring cancer took its toll and she died in September after a month in Fairmile Marie Curie Centre. The Club has lost a valued member who gave great service over 25 years, and who will not be forgotten by all her many friends and family. Alastair Peirse-Duncombe and David Clugston Ivan Hills OBE 1914 - 1999 Ivan Hills’ passion for birds was nurtured in the wilds of the Scottish Highlands. After schooling at Stow, where he was allowed to keep a pet Raven and Kestrel, he trained as aland agent and gravitated to the role of road surveyor in Sutherland. There he contributed to the construction of the roads from Lairg to Altnaharra, and by Loch Loyal to tongue and Melvich. In his spare time he searched for Golden Eagles, Peregrines, Greenshanks, Golden Plovers and Red-and Black-throated Divers, tracking down their nests on the cliffs of the hills and coast, and on the moorlands, flows and lochs. In the Grampians and Cairngorms he was amongst Dotterel and Ptarmigan and was one of the few to see nesting Snow Buntings in the years when they were Scottish breeders. At the outbreak of World War 2, Ivan enlisted in the Seaforth Highlanders and saw active service with the Eighth Army in North Africa, Sicily, Italy and Germany, rising to the rank of Major. After demob he joined the National Trust, becoming regional land agent for Kent, Surrey and Sussex, From 1967 to retirement in 1976, Ivan Hills was the Trust’s Chief land agent and based at the Trust’s London Headquarters, an unappealing location but with opportunity for a say in major policy, and for travel. In this role, he promoted programmes with important wildlife conservation spin off, such as Enterprise Neptune, with the acquisition of extensive coastal properties. He continued to visit Scotland on holiday, as the best part of Britain in which to watch birds and enlarge his knowledge of them, and was never happier than when in the north. On retirement, Ivan and his wife Mary turned their attention to Lapland, as a still more appealing northern region, and went there for 10 consecutive years in their camper van, enjoying the birds and plants of the Arctic wilderness. He became a hide photographer and sound recorder and pitted his skills as a nest finder against some of the difficult waders. His powers of endurance were tested by the 10 hours he waited by a marsh on acold day for a Bar-tailed Godwit to change over with its sitting mate, and he and Mary found no less than 6 nests of this elusive bird. Other notable nests were those of Broad-billed Sandpiper, Little Stint, Spotted Redshank, Hawk Owl, Arctic Warbler and Little Bunting. Ivan contributed observations on breeding behaviour, display, song and sonograms on some of these species to The Birds of the Western Palaearctic. Ivan lectured on his birding experiences to most branches of the SOC, where the beautiful slides that he and Mary had taken were much appreciated. A genial, kindly and outward going man, he was generous in sharing his knowledge with others; and he treated everyone as equal, with an Old World courtesy that is becoming increasingly rare. He is survived by his wife Mary, and children Tom and Lucy from his first marriage. DA Ratcliffe Scottish Birds (2000) IAT The Honorable Douglas Nigel Weir 1935 - 2000 With the sudden death of Doug Weir, Scottish ornithology has lost one of its most colourful characters and gentleman naturalists. He was the third of 6 children of the Viscount Weir. It is said that he was descended from Robert Burns; although reticent about this, he was proud that, as an infant, he had been dandled on Harry Lauder’s lap. Doug was educated at Eton and briefly (preferring birdwatching to academic studies) at Trinity College, Cambridge. During World War 2, his mother evacuated the family to her native Canada where he went to a preparatory school in the Okanagan Valley of British Columbia. His North American experiences had a profound effect on him and his great love of the outdoors found full expression there. He subsequently made a number of visits to Alaska, the first in 1952, where he based himself mainly at Nyac, a gold mining community which had been established by his Canadian uncle in the Kilbuck mountains. Later he worked as an associate ecological consultant with C C Hawley for the Klondike Miners Association on methods of land reclamation following open cast placer mining at Nyac. These environmental studies were of considerable importance. While there, he also worked extensively on the birds, especially raptors of the region. In 1974, he received a Winston Churchill Travelling Fellowship to continue this aspect of his fieldwork which resulted in a major joint publication in the North American Fauna Series (No 76). He was a skilled bird skinner and specimens collected in North America were deposited in the Royal Museum of Scotland, the Natural History Museum (Tring) and zoological museums in Alaska and California. In his more reflective moments he considered himself a prophet without honour in his own land, and it is certain that his work in North America was highly regarded there. In the early 1960s, he was the senior short term warden responsible for the day to day wardening of the RSPB’s “Operation Osprey’. Whenever he could, he also did extensive fieldwork, mainly on birds of prey in the Spey valley. He was especially active in this respect from 1964-1969 as aresearch assistant responsible to the RSPB’s Conservation Committee. He provided virtually all the Speyside data for the 1964-1968 Golden Eagle survey of Scotland. He was also a member of the Glenmore Mountain Rescue Team during the 1970s and, after he moved to Newtonmore, of the Lochaber Team. Despite being a chain smoker, few could match him on the hill at that time. In addition to a study of the inter relationship of the breeding biology of Eagle, Peregrine and Raven, he plotted all the Buzzard nesting territories from Aviemore to Newtonmore between 1964 and 1967. This was helped by having ready access to private estates through his friendship with the landowners. To capitalise on Doug’s efforts, the 2 of us worked together for 3 years on the social behaviour of the Buzzard in the - Spey valley from 1969. This resulted in 4 papers. one of which he copiously illustrated. The editor never returned the beautiful original sketches. Our close association during this period cemented a friendship begun in 1962 at the SOC annual conference at Dunblane. It was soon apparent that Doug's was a most unusual talent. He was a totally intuitive field worker but, whenever it was possible to check his claims, they always proved correct. He seemed inured to physical discomfort; if he had forgotten his wellington boots, he would cross bogs in his characteristic suede desert boots. Rather than take a long detour to a bridge, he once waded the Spey for a reviving dram at a local hotel after an arduous day on the hill. His knowledge of wildlife and the relevant literature was wide. His copious drafts, all typed on an old manual typewriter, covered the desk and floor of the 22 SB 21(2) basement office at Newtonmore which he shared with a Buzzard and his Golden Retrievers. Numerous references to his work in BWP and other major publications testify to the freedom with which he contributed unpublished data, especially on birds of prey, to leading scientists of the day. Doug was one of the 4 founding directors of the Highland Wildlife Park at Kincraig which opened in 1972. Although a commercial success, he felt there would be more security if it were part of Edinburgh Zoo and he pushed for the eventual take over of the Park by the Royal Zoological Society of Scotland in 1986. He served on its Animal Health and Management Committee (now Animal Welfare Committee) until the late 1990s. The committee’s early discussions included the then new activity of breeding endangered species in a coordinated and scientific way. Doug’s research on penguin breeding at the zoo contributed to the success of their new penguin exhibit - the largest in the world. He moved to Edinburgh in 1984 and was a frequent visitor to the Bird Section of the Royal Museum of Scotland. In 1993 his specialist expertise and Doug Weir with a colour marked Buzzard skills proved invaluable when the Museum’ s Bird Section undertook work on Braer oil spill bird casualties. Although contracted only for data gathering and specimen preparation, his commitment was such that he became senior author on most of the resulting publications. This determination led to the Section winning a contract from the Countryside Council for Wales to analyse bird casualties from the 1995 Sea Empress oil spill. In association with Museum staff, his last 2 projects involved Iceland Gulls and mortality in Buzzards, the former in press in The Journal of the Zoological Society of London. Most of us will remember him at conferences, pacing the floor and gently puffing a cigarette or stroking his beard, a drink always close to hand, discoursing knowledgeably, at length, on any variety of subjects, or drawing pithy cartoons on our dinner menus. He had anatural skill as an artist, and combined a Kodak mind for detail with a rare ability to capture the essence of an animal ina few strokes of the pen. Those privileged to know him well will miss a true and trusted friend of easy charm and great good humour. He will be greatly missed. Nick Picozzi Nick Picozzi Scottish Birds (2000) NS: aes Coe Mf! Fn ae Kors, ir ty > i y, x he J va) Wiper, : Sos = Tae ~~ IW ae pre PA : Ss. ~ ANN _— b L e oS “yt 1S Aww G Ea Goshawk one of Doug’s numerous sketches 124 Advice to contributors SB 21(2) Advice to contributors Authors should bear in mind that only a small proportion of the Scottish Birds readership are scientists, and should aim to present their material concisely, interestingly and clearly. Unfamiliar technical terms and symbols should be avoided wherever possible and, if deemed essential, should be explained. Supporting statistics should be kept to a minimum. All papers and short notes are accepted on the understanding that they have not been offered for publication elsewhere and that they will be subject to editing. Papers will be acknowledged on receipt and are normally reviewed by at least 2 members of the editorial panel and, in most cases, also by an independent referee. They will normally be published in order of acceptance of fully revised manuscripts. The editor will be happy to advise authors on the preparation of papers. Reference should be made to the most recent issues of Scottish Birds for guidance on style of presentation, use of capitals, form of references, etc. Papers should be typed on one side of the paper only, double spaced and with wide margins and of good quality; 2 copies are required and the author should also retain one. We are also happy to accept papers on disc; or by email at, mail@the-soc.org.uk however, please state the type of word processing package used. If at all possible use Microsoft Word 97. Contact Sylvia Laing on 0131 556 6042 for further information. Headings should not be underlined, nor typed entirely in capitals. Scientific names in italics should normally follow the first text reference to each species unless all can be incorporated into a table. Names of birds should follow the official Scottish list (Scottish Birds 1994 Vol 17:146- 159). Only single quotation marks should be used throughout. Numbers should be written as numerals except for one and the start of sentences. Avoid hyphens except where essential eg in bird names. Dates should be written:...on 5 August 1991......but not on the 5th (if the name of the month does not follow). Please do not use headers, footers and page numbers. Please note that papers shorter than c700 words will normally be treated as short notes, where all references should be incorporated into the text, and not listed at the end, as in full papers. Tables, maps and diagrams should be designed to fit either a single column or the full page width. Tables should be self explanatory and headings should be keptas simple as possible, with footnotes used to provide extra details where necessary. Each table, graph or map should be on a seperate sheet, and if on disc each table, graph, map etc should be on a separate document. Please do not insert tables, graphs and maps in the same document as the text. Maps and diagrams should be either good quality computer print out and in black and white (please do not use greyscale shading) or drawn in black ink , but suitable for reduction from their original size. Contact Sylvia Laing on 0131 556 6042 for further details of how best to lay out tables, graphs, maps etc. Scottish Birds Volume 21 Part 2 December 2000 Contents Main papers A major decline in House Sparrows in central Edinburgh. HEM Dott & A W Brown 61 The summer status and distribution of Greylag Geese in north and west Scotland. C Mitchell, D Patterson, P Boyer, P Cunningham, R McDonald, E Meek, J D Okill & F Symonds 69 A survey of Storm Petrels on Priest Island in 1999. P Mayhew, K Chisholm, H Insley & N Ratcliffe 78 Egg sizes of crossbills in Scotland. HA McGhie & R W Summers 85 Density and habitat associations of Barn Owls in East Ross. H A McGhie 88 The breeding success of a population of Lapwings in part of | Strathspey 1996-1998. P French, H Insley, G Siriwardena & N Buxton 98 Observations of wintering Ring Ouzels and their habitat in the High Atlas Mountains, Morocco. D S C Arthur, P R Ellis, R G Lawie & M Nicoll 109 Short notes | Prey captured and attacked by Merlins in winter. R C Dickson 116 Nestling feeding rates by Peregrine Falcons. R C Dickson 11% Obituaries Hetty Louisa Harper. Alastair Peirse-Duncombe & David Clugston 119 lvan Hills OBE. DA Ratcliffe 120 The Honorable Douglas Nigel Weir. Nick Picozzi 121 Advice to contributors 124 Front Cover House Sparrow. Tom Holden & BTO Published by the Scottish Ornithologists’ Club, 21 Regent Terrace, Edinburgh EH7 5BT. © 2000 4 f i a / a ii 4 i { ‘ } 5 , . iy j é i * n i f He me} t = f pags = . ‘ BEL Licht ks, by Vinge ae Pa tia bit Me | mM i 5 i] ‘ia iv 7 . rn ttm u es 7. j ( rn Via oan, ¥ ps JULY 02 ares! Lanne nae & woes Sp Ra a!