Wildfowl (1973) 24:111 121 Behaviour and distribution of wild geese in south-east Scotland

I. NEWTON, V. M. THOM a nd W. BROTHERSTON

The distribution of wintering Greylag numbered about 60,000 individuals, and Anser anser and Pink-footed Geese A. Pinkfeet about 70,000 (Boyd & Ogilvie, brachyrhynchus in Britain is governed by 1969, 1972). the presence of suitable roosting sites, such as estuaries and lakes, from which the birds fly to nearby farmland to feed. The same M ethods roosts and associated feeding areas are used year after year. This paper examines the The first detailed information on goose dis­ dispersal of geese in south-east Scotland tribution in Scotland was built up over and, in particular, their choice of roosting several years from extensive ground and feeding sites. It forms part of a general searches and aerial surveys by Wildfowl study, more of which has been published Trust staff, helped by local observers. This elsewhere (Newton & Campbell, 1970, and information, summarized by Atkinson- 1973; Newton, Campbell & Allison, in Willes (1963), formed the starting-point press). The area involved, covering 24,000 for our survey. From 1960, V.M.T. studied km2, extended from Pitlochry in the north goose distribution around Perth. This was to beyond Biggar and Kelso in the south, done partly at weekends, but also on other and from the east coast to beyond Stirling days, because her job took her to many in the west (Figure 1). It included a repre­ farms in the study area, and provided sentative selection of roosting and feeding further opportunity to find geese and sites, and held the bulk of the immigrant question farmers (Thom & Murray, 1964). population of both species. In recent In 1952, W.B. (in Edinburgh) began autumns, Icelandic Greylag in Britain have watches at certain roosts, which came to

Figure 1. Map showing position of study area, open circles = Pinkfoot; split circles = both and main roosting sites. Filled circles = Greylag; species. Inset shows area covered by Figure 2.

I l l 112 I. Newton, V. M. Thom and W. Brotherston

occupy nearly every winter weekend for 20 years, and from 1955 also organized Behaviour cooperative counts at all known roosts in the study area south of the Forth (Brothers­ Most Pinkfeet arrived in Britain in late Sep­ ton, 1964). He too noted feeding areas, but tember and early October and left during did not cover the ground systematically as April and early May. Most Greylag arrived did V.M.T., north of the Forth. In 1967-70 during October and left from mid-March to I.N. also searched systematically for feed­ mid-April. Thus the average Pinkfoot ing geese, and checked waters not already spent 30 weeks each year in Britain and the known to be used as roosts. This paper average Greylag 26. The difference was synthesizes all this information. Records presumably because the Pinkfeet came accumulated for 20 years, but all roosting direct from the highlands of Iceland, where and feeding areas mentioned here were the summer season was shorter than in the used regularly during 1966-70. Possibly lowlands inhabited by Greylag. some minor roosting and feeding areas Within Britain, different roosts were not were missed, but not enough to affect the self-contained, and a continual shifting of overall conclusions. birds from roost to roost around the country was confirmed by observations, counts and ringing. Recoveries of birds Population ringed near particular roosts came, within the next 3 months, from all parts of the An estimated 37,000^4-6,000 Greylag were winter range, but more often from near present in the study area in different than far (Boyd, 1955, 1957, 1959). Move­ Novembers 1966-70, with an average of ment took place on any date, not just in 41,000, and an estimated 49,000-63,000 response to hard weather or food shortage. Pinkfeet, with an average of 58,000 (Boyd In general, geese concentrated in predo­ & Ogilvie, 1969, 1972). In different years, minantly grain-growing areas in autumn, this represented 59-73% of the total im­ and in predominantly grass-growing areas migrant Greylag in the country, and in winter. As expected, day-to-day inter­ 64-93% of the total Pinkfeet. The large change was greatest between adjacent differences in numbers between autumns roosts, and in any one area certain pairs or resulted partly from variations in the size of groups of waters functioned as one, so that the total population, and also from annual large numbers of geese at one site meant variations in the regional distribution of small numbers or none at the others. North geese within Britain. In the study area of the Forth, two groups of waters in north of the Forth, both species were rela­ Strathmore acted more or less as two units tively numerous in 1968 and 1970, when for Greylag; and to the south Harperig/ much grain was shed before harvest, and Threipmuir and Gladhouse/Portmore were less so in other years when little grain was each linked for Greylag, as were Fala/ shed. South of the Forth, grain and geese Gladhouse and Baddinsgill/Westwater for were most plentiful in 1969. In all autumns, Pinkfeet. the two species were about equally numer­ The number of geese that used any one ous north of the Forth, but to the south water was not normally restricted by its Pinkfeet greatly predominated (Table 1). surface area. The small pond at Fala,

Table 1. Numbers (in thousands) of Greylag and Pink-footed Geese in study area, each November, 1966-70, based on Wildfowl Trust Counts

Greylag Pinkfoot

N orth of South of Total in N orth of South of Total in Forth Forth study area Forth Forth study area

1966 39-3 1-7 41-0 (68%) 35-0 13-7 48-7 (64%) 1967 35-9 2-6 38-5 (73%) 44-0 16-3 60-3 (93%) 1968 41-8 2-6 44-4 (73%) 44-6 10-0 54-6 (84%) 1969 30-1 6-4 36-5 (59%) 44-6 16-8 61-2 (84%) 1970 41-5 4-8 46-3 (71%) 50-3 13-1 63-4 (88%) Mean 37-7 3-6 41-3 43-7 14-0 57-6

The figures in brackets show the percent of the country’s total population present in the study area each November. Wild geese in Scotland 113

which covered about 1 hectare, held for the first flocks were in, later ones flew in much of each season less than 2,000 geese, without the hesitation often shown by first but occasionally up to 6,000-an average of arrivals. There was usually a great clamour, one goose per 2-5 m2. On large roosts geese redoubled whenever a fresh detachment did not spread evenly over the water sur­ came in. Then some time after the last flock face, but remained as a flock in the lee of a had arrived, the noise steadily fell to a low shore or island, their position changing level, which was maintained until about an from night to night, according to wind hour before the dawn flight when it in­ direction. They preferred sheltered areas creased again. (This was confirmed by the near gentle open shores, where they could ‘water keepers’ at several reservoirs, and walk out during darkness. In estuaries and by M. A. Ogilvie who spent many nights coastal bays they usually stood on banks of beside goose roosts.) sand or mud that were flooded for only a During the lighter half of the moon cycle short time. They drifted on the water some geese fed at night, more so in winter at high tide, and returned to the banks when than in autumn or spring (Newton & Camp­ the water dropped again. On waters re­ bell, in press). When the moon rose early, gularly used by both species, the two kept in the birds simply remained in the fields, but separate flocks. when it rose late, they roosted as normal As a rule both species roosted by night and flew out again at night. After such light and fed by day. They left their roosts earlier nights, any geese on the roosts at daybreak with respect to sunrise, and returned later were often slow to leave. Some Greylag with respect to sunset, in mid-winter, when stayed till around noon, while Pinkfeet days were shortest, than in autumn or often flew to a resting area and did not feed spring. On mixed roosts, Pinkfeet left for an hour or more. earlier and returned later, on average, than did Greylag, though there was much over­ lap. In both, it often took more than an Influence of shooting hour for all the flocks to leave a roost in the morning, and for them to assemble Shooting and other disturbance had less again in the evening. effect on geese using large waters than on After reaching a feeding area in the those using small ones. On the large Loch morning, the first party often circled once Leven, no consistent trends in numbers were or twice before alighting, but the build-up detected after shoots, but small waters were was then rapid, with successive parties frequently deserted for several days after­ settling in without hesitation. After the wards. Pinkfeet more often deserted a roost dawn flight, the roosting assemblage usually after shooting than did Greylag. Shooting split into several feeding flocks at varying also affected the time geese arrived at a distances and directions 'from the roost. roost and, when heavily disturbed, both During the day, groups of birds moved species delayed their arrival till well after between flocks, and also between feeding dark. In areas where shooting was especially and roosting places. Birds fed least around heavy, geese also made maximum use of the mid-day and, if water was not available in moon for feeding. The limited disturbance the fields, sometimes returned to the roost on feeding areas was less important, because to drink and bathe. Greylags often accumu­ almost always the birds found alternative lated steadily on their roosts from around places nearby, and did not leave the dis­ mid-day and, in general, spent much more trict. time on water than did Pinkfeet. To large and undisturbed waters, like Loch Leven, Pinkfeet occasionally returned to drink and Influence of snow and ice bathe around mid-day, but otherwise they drank from pools on fields (if at all) and Seldom, if ever, were all localities fre­ rested on special areas (see later), so that quented by geese under snow at the same roosts remained deserted through the day. time and, while many birds concentrated Both species arrived at roosts in flocks, temporarily in snow-free areas, others often from several different directions. On stayed where they were. Shallow snow windy nights their arrival was often more seemed not to influence the foraging of prolonged and fragmented than on calm either species, but deep snow restricted the ones. The geese approached in level flight, choice of areas available. Both species fed sometimes rising when nearing the roost, from large grass tussocks at such times, on then gliding or ‘whiffling’ (side-slipping and fresh shoots and rootlets, and Greylag also tumbling) down to the water surface. After ate protruding turnips. Both species spent 114 I. Newton, V. M. Thom and W. Brotherston

more time than normal resting during snow low flying aircraft, shooting, human pre­ periods and less in trying to feed. When a sence, and unfamiliar and prominent ob­ roost froze over, but food remained avail­ jects like oil drums and scarecrows. Geese able locally, geese usually rested on the ice, soon learned to associate danger with either on their feet or bellies. It was often particular places, and the same birds were possible to see how many had spent the pre­ more wary at one place than at another. vious night on a site from the castles of The main roosts of the two species in the droppings left behind. study area are given in Table 2, and all the sites at which geese were seen to roost in Appendix 1. Roosting sites

Several factors combined to make a site safe (a) Coastal and estuarine sites and acceptable to geese for roosting. These included not only the degree of actual dis­ Some favoured roosting sites were on es­ turbance, but also various intrinsic features tuaries and coastal mud and sand flats, which conferred safety, like situation, area, which were flooded for a minimum period openness, distance from centres of human each day, yet where no enemy could ap­ activity, and so on. The extent to which any proach undetected. A total of nine large particular roost was used seemed to depend mud or sand flats was available in the area, in mainly on the balance between its intrinsic or near estuaries; at seven, Pinkfeet pre­ attractiveness and the level of .actual dis­ dominated, and at two Greylag (Table 3). turbance. The extensive mud flats of Mon­ The smallest mud flat used by geese for trose Basin, for example, were highly roosting covered about 3 km2, but the area attractive to geese, yet little used because of a site was less important than the dis­ of great disturbance by wildfowlers. On the tance it spread from shore. Greylag, in their other hand, Dupplin Loch was basically favourite sites, could get up to 1 km off unattractive because of its small size and shore, but Pinkfeet liked to get even further woodland setting, but was so undisturbed out, and on Abertay Sands and Dog Bank that at times it held more geese than any in the Tay regularly stood up to 3 km off other site in Britain. To judge from the shore. The only smaller areas of coastal reactions of geese, the most important mud which geese frequently used were cut disturbances were, in order of importance, off on all sideby water. Both species roosted

Table 2. Main roosts of Greylag and Pink-footed Geese in the study area Greylag Pinkfoot

Strathmore Lochs Rescobie, Balgavies, Forfar Montrose Basin, Lochs Forfar Kinnordy, Lintrathen, Clunie, and Rescobie Marlee and Stormont, Monks Myre, ‘Bloody Inches’. Firth of Tay Mugdrum island Three sites in the estuary Strathearn Drummond Pond, Pitcairnie Loch, Dupplin Loch three sites on river Strathallan Carsebreck Carsebreck, Loch Mahaick Ochil Hills Glenfarg Reservoir Glenfarg Reservoir Kinross Plain Loch Leven Loch Leven Eden area Edenmouth, Cameron Reservoir Forth Valley Flanders Moss, Lake of Menteith, Flanders Moss, Lake of Inches near Alloa, Grangemouth Menteith, Grangemouth, Aberlady Pentland Hills Harperig and Threipmuir Reservoirs Cobbinshaw, Baddinsgill and Westwater Reservoirs Moorfoot Hills Gladhouse Reservoir, Portmore Loch Gladhouse Reservoir, Fala Moor Lammermuir Hills Watch Reservoir Hule Moss Lanark Hills Upper Cowgill and Culter Reservoirs Lower Tweed Hoselaw Loch Valley Wild geese in Scotland 115

Table 3. Use of coastal and estuarine sites for roosting by geese in the study area Approximate area (km2) exposed at low tide Used by Montrose basin* 7 Pinkfeet, rarely Greylag Tay (1) Abertay sands 3 Pinkfeet (2) Dog/Carthegina banks 26 Pinkfeet, some Greylag (3) Mugdrum and nearby 1-5 Greylag, some Pinkfeet banks Edenmouth* 6 Pinkfeet Forth (1) Tyninghame* 3-5 Pinkfeet (2) Aberlady Bay 4 Pinkfeet, rarely Greylag (3) Grangem outh Flats* 10 Pinkfeet, some Greylag (4) Inches at Alloa* 0-7 Greylag, some Pinkfeet

*Those thus marked are much disturbed by wildfowlers, are now used less than at the turn of the century, and mainly after the end of the shooting season. Edenmouth is also disturbed by low-flying aircraft. on Mugdrum Island and its associated sand pools at the edge. The river at such points banks in the Tay, and on certain islands in was also fairly wide, with low banks devoid the Forth near Alloa. of trees. All three rivers in the area which In both Tay and Forth, several sites were offered these facilities were used. Favoured used, but Pinkfeet predominated at the roosts on the Earn included the flood pools safer seaward ones, and Greylag further at Innerdunning, Dalreoch, and Netherfor- inland. Thus Abertay Sands near Taymouth dun; on the Tay the Inches near Meikleour were used only by Pinkfeet, Dog and and the islets south of Pitlochry; and on the Carthagina Banks further upstream mainly Clyde the Haughs near Quothquan by Pinkfeet, and Mugdrum and its asso­ (Lanark). Pinkfeet were seen at no such ciated sand banks yet further up mainly by sites, but used the extensive sheets of water Greylag. Likewise in the Forth, Tyning- which occasionally resulted from flooding hame Sands and Aberlady Bay were regular­ on the Clyde south of Libberton. This was ly used only by Pinkfeet, Grangemouth especially true when nearby still waters mainly by Pinkfeet, and the upstream were frozen. sites around Alloa mainly by Greylag. Furthermore, at Grangemouth, where a harbour divided the flats, Pinkfeet mostly (c) Lakes and reservoirs used the larger eastern sector, and Greylag the smaller western one. The only suitable Ordnance Survey Maps (scale 1/63,360) site on the Eden (at the mouth) was used show a total of 165 ponds, lochs and re­ almost exclusively by Pinkfeet, as was servoirs in the study area, excluding pools Montrose Basin on the Angus coast. on mosses discussed below. Many such sites were too disturbed, in too narrow and steep­ sided valleys, surrounded by trees, too far (b) Rivers from feeding areas, so high that they were often frozen, or otherwise unsuitable as Further up river, Greylag occasionally roosts. For the remainder, the surface area roosted on bare islands or on shingle banks of water had an obvious influence on in mid-stream, or on wet ground and flood whether a site was used. Only 22% of

Table 4. Use of ponds, lakes and reservoirs for roosting by geese in the study area (mossland pools excluded—See Appendix 2) Used by Area of water Total Used by Used by Used by neither (km2) available Greylag Pinkfeet both species species < 01 99 20 2 0 77 0-11-0-20 29 18 4 2 9 0-21-1-00 31 23 13 8 3 > 10 6 5 5 4 0 Totals 165 66 24 14 89 116 I. Newton, V. M. Thom and W. Brotherston

ninety-nine waters less than 10 ha in area lative degrees of security they offered, and were used by geese, but this proportion the extent of recent disturbance at each. In increased with increasing area, and all practice, the numbers of geese at many waters extending over more than 100 ha sites fluctuated greatly from night to night, were used. The two species differed in the especially during the shooting season. Ap­ water areas they accepted. Greylag used the pendix 1 gives some idea of the maxima for largest number, including both small and the different sites recorded during our large waters, but Pinkfeet used mainly study, but adding them together would give waters more than 20 ha in surface area a figure far in excess of the total geese in the (Table 4). Thus, of 128 waters of less than area at any one time. 20 ha in area, thirty-eight were used by Greylag and only six by Pinkfeet, but of Rest stations thirty-seven larger waters, twenty-eight were used by Greylag and eighteen by When feeding several kilometres from a Pinkfeet (P < 0-01). Further, only twelve of roost, or when the roost offered only a the thirty-seven large waters in the area small sheet of water, geese usually adopted were used by both species together, the areas of moor or rough grass as ‘rest sta­ remaining twenty-five by one or other, so tions’ from which they commuted to feed­ here again the species tended to separate. ing areas nearby. These rest stations were constant from year to year, but not neces­ (d) Moorland pools sarily used throughout a season. Their sit­ Many roosts were centred on small pools on ing appeared to result partly from the need remote, damp, moors (or mosses). Because to be near feeding areas and partly from the of the nature of the ground, such sites were need for safety. Sites were often centred on seldom disturbed and all provided a wide a wide stretch of damp moorland or near view. Geese roosted on the pools themselves the summit of a rounded hill in farmland, or on ground nearby. The map showed but always provided a wide view over nineteen such mosses, with suitable pools, surrounding land. The farmland ones in our area. One of these was used by Grey­ were usually as far as possible from a road. lag alone for roosting, four by Pinkfeet One could not approach geese on such sites alone, and another two by both species without being seen. Of twenty-three regular (Appendix 2). Three of the most important Pinkfoot rest stations found, twelve were Pinkfoot roosts in the area were on east on mosses, six on large grassy fields on hill­ Flanders Moss (Perthshire), Fala Flow sides, four on offshore sand banks or (Midlothian) and Hule Moss (Berwick­ islands, and one in a damp field. Of twelve shire). For daytime resting, Greylag used at Greylag rest stations found, three were on least another two such sites, and Pinkfeet mosses, two on islands, and seven on damp another four. Also, Pinkfeet rested by day riverside fields (Appendix 3). on at least eight other mosses, which lacked Rest stations were used much more by permanent water, and Greylag roosted on Pinkfeet than by Greylag and possibly Cranley Moss (Lanark) when flooding helped Pinkfeet to exploit successfully feed­ created pools. ing areas distant from roosts. After leaving In conclusion, the two species differed a roost in the morning, Pinkfeet sometimes in their favourite roosting places. Pinkfeet flew to a rest station and then, using it as a preferred extensive estuarine mud flats, base, flew to and from the fields throughout large lochs and reservoirs, and remote the day. At evening, too, Pinkfeet often mosses; Greylag used these sites to some assembled on a rest station before going to extent, but also used smaller waters and roost. They then approached the roost in rivers. In general, sites used by Pinkfeet much poorer (and possibly safer) light than offered greater security and freedom from if they had flown there direct. This be­ disturbance than did many of the sites used haviour was especially prevalent when the by Greylag. This does not explain why roost offered only a small expanse of water, Greylag avoided many of the safer sites, like the pond on Fala Moor. As a rule each unless to avoid Pinkfeet already there. On Pinkfoot roost had one or more rest sta­ shared roosts, the two species normally tions associated with it. kept to different areas, and flighted inde­ pendently. Feeding areas The extent to which any particular roost was used seemed to depend partly on the Geese fed on only a small part of the farm­ number of alternatives available, on the re­ land in the study area. Usually each roost Wild geese in Scotland 117

had several associated feeding areas, which mented (Baxter & Rintoul, 1953; Harvey- varied from a single field to tracts of seve­ Brown, 1906; Millais, 1901; Muirhead, ral square kilometres (Figure 2). (A feeding 1895; Nash, 1935). Both species seem to area was considered separate when it was have been increasing in Britain at least since more than 0-5 km from the next, otherwise 1880. The evidence up to 1930, mostly the two were classed as one.) Adjacent un­ based on memory records, was given by suitable terrain limited the extent of most Berry (1939), but since 1950 counts or­ feeding areas. In general, geese fed most in ganized by the Wildfowl Trust have con­ extensive fiat or slightly undulating country, firmed the trend (Boyd & Ogilvie, 1969, with few trees and hedges, and avoided 1972). Throughout, moreover, the increase hummocky or well timbered terrain, with has been most marked in Scotland. small fields and tall hedges, which restricted Not only did numbers rise at estab­ their view. They also avoided areas which, lished roosts, but new sites were also oc­ though topographically suitable, were cupied, mainly inland. The spread inland much disturbed. Pinkfeet were more ex­ was probably encouraged by; (1) increased treme in their preferences than Greylag, disturbance at coastal sites resulting from and often flew further from their roost to greater public wildfowling, aircraft and feed. We have tried to quantify this last military activities (Berry, 1939); (2) con­ difference :n the study area by calculating struction of reservoirs (= roosts) in areas the proportion of the total feeding area of lacking natural lakes; (3) removal of trees the two species which lay at various dis­ and hedges, creating open spaces attractive tances from the nearest roost (Table 5). to geese in inland areas formerly unsuitable; Analysis was restricted to the area north of and (4) more autumn ploughing near the the Forth where feeding areas were best coast rendering much ground useless to known. In this area 90% of the Greylag’s geese thereafter, and improved manage­ total feeding ground was within 5 km of a ment of grass in the uplands providing better roost, and only 2% more than 10 km away feeding there than formerly (Brotherston, compared with 66% and 15% for the Pink­ 1964). foot. About 1% of the Pinkfoot’s feeding The main coastal sites occupied in areas were more than 20 km from a roost. 1875-1900 were Montrose Basin (Pinkfeet), (The difference between the species is signi­ Inner and Outer Tay (both species), Eden­ ficant at the 0-1% level.) No account was mouth (both species), Grangemouth Flats taken of the extent to which the two species (Pinkfeet), Aberlady Bay (both species) and used feeding areas at different distances, Tyninghame (Pinkfeet). The main inland nor the fact that geese did not invariably sites were Loch Leven (both species), fly to the nearest roost from a particular Cobbinshaw Reservoir (Pinkfeet), Hule feeding area. Moss (Pinkfeet), Coldingham Moss (Pink­ feet), Flanders Moss (both species), two unspecified mosses in Lanarkshire (Pink­ Spatial separation of the two species feet), and Fala Flow (Pinkfeet). The Pink­ foot was thus most widespread, roosting Differences in roosting sites and flighting at fourteen sites (eight inland), compared distance tended to separate the two species with the Greylag’s five (two inland). At that and reduced the extent to which they fed on time, the Greylag shared all its roosts with the same ground. Also Greylag generally the Pinkfoot, which always outnumbered preferred to be near rivers and Pinkfeet on it. Today the Pinkfoot has twenty-five extensive open areas, however far from major roosts (nineteen inland) in the same water. North of the Forth, geese of one area, and the Greylag twenty-nine (twenty- species or the other fed over about 351 km2 eight inland); north of the Forth, both spe­ of land, Greylag alone over 151 km2, Pink­ cies are about equally numerous, but to the feet alone over 153 km2, and the two species south the Pinkfoot still predominates, less together over only 47 km2. Hence only so than in 1950. Now the Greylag shares 13% of the total goose country was used by only nine of its twenty-nine main roosts both species. with the Pinkfoot, so the two species are also more segregated than formerly. The disappearance this century of Pinkfeet Changes this century in the distribution of from Coldingham can be attributed to the geese in the study area destruction of the moss by ploughing, but Greylag have recently adopted the ponds Because of the sporting value of geese, their rem aining. distribution over the years was well docu­ Several roosts in the study area have . etn V M To ad . Brotherston W. and Thom M. V. Newton, I.

Figure 2. Part of the study area, showing the separation of feeding areas of Greylag and Pink-footed Geese. Wild geese in Scotland 119

Table 5. Extent of total feeding area of geese lying at different distances from roosts Distance from nearest roost (km) 0-2 2-5 5-10 10-20 20-30 % Greylag feeding area 49 41 8 2 0 % Pinkfeet feeding area 24 42 19 14 1

*Greylag alone fed over 151 km2 of the study area north of the Forth, Pinkfeet alone over 153 km-, and both species together over 47 km2. experienced fluctuating usage by geese over geese because they were the most extensive the years associated with changes in dis­ and open, while inland, large waters were turbance. Coastal sites, being public, more attractive than small ones. For feed­ were generally more disturbed than inland ing, open, flat or slightly undulating ones, the more so in recent years because of country, with a minimum of trees and increased mobility of wildfowlers. At the hedges, was preferred to hummocky, well Tay and Forth Estuaries military operations timbered country, with small fields. Second, were also involved (Berry, 1939; Brothers­ actual disturbance, caused by shooting, ton, 1964). Some coastal sites, with public human presence and other factors, super­ access, such as Montrose, Eden, Grange­ imposed its own pattern on that set by mouth and Tyninghame, which were for­ topography. The extent to which a particu­ merly important, are now little used until lar roost or feeding area was used seemed after the shooting season each year. Like­ to depend on the balance between its in­ wise the use of Cobbinshawby Pinkfeet has trinsic attractiveness and the degree of declined since shooting increased there, actual disturbance. Lastly, while geese while Harperig was largely deserted by clearly linked disturbance with particular Greylag for a time in favour of Threipmuir sites, the degree of overall shooting pro­ under the same conditions. On the other bably also affected the minimum require­ hand, the use of Aberlady Bay increased ments for roosting and resting areas, a again after it became a nature reserve in heavily shot population using only the safer 1952 (Brotherston, 1964). These changes of a range of sites and a lightly shot provide circumstantial evidence for the population accepting others. importance of disturbance in influencing Differences in roosting, feeding and goose-distribution. flighting habits between the species could One last point worthy of comment is the be attributed largely to the greater wariness increased importance to both species of of the Pinkfoot and its stronger reaction to reservoirs and other artificial lakes. Now­ disturbance. For roosting, it used places adays eight of the twenty-nine main Grey­ which were especially safe, either because lag roosts in the study area, and nine of the of their intrinsic characteristics, or because twenty-five Pinkfoot ones, were on man- they were remote and otherwise free from made waters. disturbance. The avoidance of some of these sites by Greylag is puzzling, unless to avoid Pinkfeet. On shared roosts, the two Conclusions species normally kept apart, and several times this century Greylag segregated com­ pletely from Pinkfeet when an alternative Potential limits to the distribution of geese roost in an area became available (Brother­ in the study area were set by the location of ston, 1964). Last century all the Greylag suitable roosting sites in or near farmland, roosts were shared by Pinkfeet, but now where all feeding was done. Within this that more waters are occupied, less than framework, the distribution of the birds one-third of the Greylag roosts are shared was then influenced mainly by disturbance, (A ppendix 1). which banished or reduced them in certain areas and led to their increase in others. The degree of safety offered by an area was Aknowledgments influenced by two types of factors. First were those imposed by the environment. We are grateful to A. Allison & C. R. G. For roosting, coastal and estuarine sites Campbell for allowing us to incorporate their were intrinsically the most attractive to observations of geese, and to H. Boyd, Dr. D. 120 I. Newton, V. M. Thom and W. Brotherston

Jenkins, Professor G. V. T. Matthews, M. A. Greylag Geese wintering in Britain in 1960— Ogilvie & L. Young for their comments on the 1971. Wildfowl, 23:64-82. manuscript. Brotherston, W. 1964. The numbers & beha­ viour of geese in the Lothians and Berwick­ shire. Wildfowl Trust. Ann. Rep. 15:57-70. Summary Harvie-Brown, J. A. 1906. A Fauna o f the Tay Basin and Strathmore. Edinburgh: David In southeast Scotland immigrant Greylag and Douglas. Pink-footed Geese fed entirely on farmland Millais, J. G. 1901. The Wildfowler in Scotland. (including grassland), where their distribution London. was governed by the location of suitable roosts, Muirhead, G. 1895. The Birds of Berwickshire. from which they flew to restricted feeding areas Edinburgh: David Douglas. nearby. Continual movement took place be­ Nash, J. K. 1935. The Birds o f Midlothian. tween different roosts, and peak numbers oc­ London: Witherby. curred in different months in different districts. Newton, I. & Campbell, C. R. G. 1970. Goose Pinkfeet preferred the safer of a range of studies at Loch Leven in 1967/68. Scot. sites for roosting, including estuaries, large Birds 6:5-18. lakes and reservoirs, and remote moorland Newton, I. & Campbell, C. R. G. (1973) pools, while Greylag also used less safe sites, Feeding of geese on farmland in east-central including small ponds and rivers. Of forty-five Scotland. J. appi. Ecol. 10: (in press). major goose roosts in the study area, only nine Newton, I., Campbell, C. R. G. & Allison, A. were used regularly by both species, the rest by In press. The food of Greylag & Pinkfooted one or other. On shared roosts the two species Geese in east-central Scotland. Bird Study, kept apart. Thom, V. M. & Murray, C. 1966. Grey geese Pinkfeet were also more particular in their in Perthshire. Trans. Proc. Perth. Soc. Nat. choice of feeding areas, and often foraged further Sci. 11:38—42. from their roosts than did Greylag. Greylag rarely flew more than 5 km to feed, but Pinkfeet Dr I. Newton, The Nature Conservancy, regularly more than 10 km, and occasionally 12 Hope Terrace, Edinburgh, EH9 2AS. more than 20 km. Mss V. M. Thom, 19 Braeside Gardens, Perth. Differences in roosting and flighting habits W. Brotherston, 22 Rutland Square, Edinburgh led to spatial separation of the two species on E H I2BB farmland and, of the total goose country in the study area, only 13% was occupied by both species together, the rest by one or other. Appendix 1. Sites on which geese were known Within limits set by the location of suitable to roost in the study area, 1966— roosting and feeding areas, the distribution of 70. A: usual maximum number geese, was influenced mainly by disturbance, more than 2,000; B: 1,000- especially shooting. 2,000; C: 500-1,000; D: less than 500. Greylag References O.S. map 49: Haughs ofTay south of Pitlochry (D), Dowally Loch (D), Craiglush Atkinson-Willes, G. L. 1963. Wildfowl in Great Loch (D), Loch of the Lowes (D), Britain. London: HMSO. Butterstone Loch (D), Loch Baxter, E. V. & Rintoul, L. J. 1953. The Birds of Benachally (D), Loch Clunie (A), Scotland. London & Edinburgh: Oliver & Kings Myre (D), Broomhill Pond Boyd. (A), Bloody Inches (A), Marlee Berry, J. 1939. The status & distribution of wild Loch (B), Dykeside Moss (D), Fing- geese & wild duck in Scotland. Int. Wildf. ask Pond (D), Stormont Loch (A), Inquiry, Vol. 2. Cambridge: Univ. Press. Hare Myre (D), Monks Myre (A), Boyd, H. 1955. The role of tradition in deter­ Lintrathen Loch (C), Long Loch mining the winter distribution of Pinkfeet in (D), Redmyre (D), Airntully Britain. Wildfowl Trust Ann. Rep. 7:107-22. Pools (D). Boyd, H. 1957. Recoveries of British-ringed O.S. map 50: Kinnordy (C), Forfar Loch (B), Greylag Geese. Wildfowl Trust Ann. Rep. Rescobie/Balgavies Loch (B), 8:51^1. Monikie Reservoir (D), Duns Boyd, H. 1959. Greylag Geese in Britain in Dish (C). Winter. Wildfowl Trust Ann. Rep. 10:43-58. O.S. map 54: Lake of Menteith (B), Loch Rusky Boyd, H. & Ogilvie, M. A. 1969. Changes in (D), Flanders Moss (A), Muir the British wintering population of the Dam (D), Airthrey Loch (formerly Pink-footed Goose from 1950 to 1975. D), Braco Pond (D), Loch Wildfowl, 20:33-46. Monzievaird (D), Loch of Balloch Boyd, H. & Ogilvie, M. A. 1972. Icelandic (D), Cowden Loch (D). Wild geese in Scotland 121

O.S. map 55: Drummond Pond (A), Carsebreck O.S. map 62: Westwater Reservoir (A), Ponds (A), Glendevon Reser­ Baddinsgill Reservoir (A), voirs (D), Meallbroden Loch (D), Gladhouse Reservoir (A), Fala Gartmorn Dam (D), Peppermill Flow (A), Aberlady Bay (A). Dam (D), Pitcairnie Pond (A), O.S. map 63: Hopes Reservoir (D), Tyninghame Craigluscar Reservoir (D), Bertha (D), Hule Moss (A). Loch (C), Loch Glow (D), Glen­ O.S. map 68: Loch Lyock (D), Floods on Clyde, farg (B), Loch Leven (A), Loch Quothquharn (B), Upper Cowgill Fitty (D), Loch Ore (D), Earn- Reservoir (C), Culter Reservoir Nether Fordun (C), Earn-Dal- (D). reoch (B), Earn-Inverdunning (C), Tay-Mugdrum Island (A), Tay- Appendix 2. Mosses and moors used for Dog/Carthegina Banks (C) roosting and resting by geese. Harperleas Reservoir (D), Ballo G: Greylag; P: Pinkfoot. Reservoir (D), Holl Reservoir (D), Areas with ponds used for roosting: Bowmuir Forth-Alloa Inches (C)j (P), Crane Loch (P), Dunshelt (G) Fala Flow O.S. map 56: Loch Lindores (C), Dunshelt Moss (P), Flanders Moss (PG), Hule Moss (P), (C), Carriston Reservoir (D), Redmyre (PG). Clatto Reservoir (D), Kilconquhar Areas with ponds used for resting: Auchterhead Loch (D), Carabee Pond (D). Muir (P), Clevage (P), Cranley Moss (G) O.S. map 61: Springfield Reservoir (D), Grange­ Dykeside (G), Kippen Muir (P), Rossie mouth Flats (C), Crosswood Moor (P). Reservoir (D), Crane Loch (D). Areas without ponds used for resting: O.S. map 62: Harperig Reservoir (D), Threip- Auchencorth Moss (P), Esperson Moss (P), muir (C), Portmore Loch (D), Toxside Hill Moss (P), Cocksmuir (P), Gladhouse Reservoir (B), Allan- Middleton Muir (P), Methven Moss (P), Muir shaws Reservoir (D). of Orchill (P), Sherriffmuir (P). O.S. map 63: Stobsheil Reservoir (D), Watch Reservoir (C), Coldingham Ponds Appendix 3. Rest stations of geese in study (D), Hirsel Lake (D). area. G: Greylag; P: Pinkfeet. O.S. map 64: Hoselaw Loch (C). Strathmore: Rossie Moor (P), Redmyre (GP), O.S. map 68: Cranley Moss (D). Wester Essendy Farm (G). O.S. map 70: Yetholm Loch (D). Strathtay: Tentsmuir Point (P), Abertay Sands (P), Mugdrum Island (GP). Pinkfoot Strathearn: Methven Moss (P), Bachilton O.S. map 49: Redmyre Moss (D). Farm (P), Pow Water (G), East O.S. map 50: Forfar Loch (A), Rescobie Loch Fordun Farm (G), Bogtonley (A), Montrose Basin (C). Farm (P), Denmarkfield River O.S. map 54: Lake of Menteith (B), Flanders Shingle (G), Milton of Forteviot Moss (A), Loch Mahaick (A). Farm (G), Lauchie Farm (P), O.S. map 55: Carsebreck Ponds (A), Glendevon Kilspindie Farm (GP), East Reservoirs (D), Peppermill Dam Forden Farm (G). (D), Dupplin Loch (A), Clevage Kinross Moor (D), Glenfarg (C), Loch Plain: St Serfs Island, Loch Leven (P). Leven (A), Loch Fitty (D), Loch Eden Area: Dykeside Moss (G). Ore (D), Tay-Mugdrum Island Forth (A), Tay-Dog/Carthegina Banks Valley: Kippen Muir (P), Flanders Moss (A), Tay-Abertay Sands (A), (GP), Sherriffmuir (P), Muir of Harperleas Reservoir (D), Ballo Orchill (P). Reservoir (D), Forth-Inches near Moorfoot Alloa (D). Hills: Toxsidehill (P), Esperton Moss O.S. map 56: Edenmouth (C), Cameron (P), Middleton Moss (P), Reservoir (A). Cakemuir Hill (P), Auchencorth O.S. map 61: Springfield Reservoir (D), Moss (P), Halflow Kiln Farm (P), Grangemouth Flats (C), Cocksmuir (P). Cobbinshaw Reservoir (C), Crane Lanark Loch (D), Bowmuir (D). Hills: Cranley Moss (G). é t

E. E. Jackson

Plate IV. Diving ducks on land. Above: a male Canvasback Aythya vallisneria shows that it can maintain a horizontal posture despite its far-back legs. Below: a pair of Ring-necked Aythya collaris unusually have erected crests (part of the courtship) while ashore.

Philippa Scott