The Hen Harrier

  Counts at most winter roosts in Europe show an excess of ringtails over adult males. The most striking exception was the Irish roost where, in October, all the birds—up to ten—were adult males, although ringtails occasionally appeared at the roost but drifted away. King considers that the numbers of adult males in south-west Ireland decline, later in the year, perhaps due to emigration, while many ringtails remain all winter. In Belgian Lorraine Mois found that males exceeded females by two or three to one between October and the beginning of March, but during March and April there was usually an excess of females. A similar pattern of change in the proportions of males and females was found at roosts in Luxembourg by Wassenich and in Eifel, Germany, by Jakobs.

  Elsewhere, the only region where adult males sometimes equal or exceed ringtails at roosts is south-west Scotland, most often in the autumn or early winter. The highest total numbers at roosts in south-west Scotland were recorded between November and January, but in the Netherlands and West Germany peaks were reached in February. In Belgian Lorraine the highest total was only 11 birds, but over seven winters a fairly regular pattern of fluctuations emerged. Peak counts occurred in mid November, late January–early February and in mid–late March.

  Communal roosting in south-west scotland

  When, in November 1959, I saw a ringtail harrier gliding down at dusk above a forest slope, to drop silently into a wide flat expanse of boggy grassland, the possibility that this might be a site for a communal roost did not occur to me. At that time the winter population in the region was extremely small and, anyway, I was not aware that Hen Harriers had ever formed large roosts in the British Isles.

  Over the next few years, I sometimes paused on an evening above the big flat, or flow, and obtained a fleeting view of two or three harriers as they flapped and glided over the orange and ochre grasses. I saw them hesitate and drop out of sight in the tall cover, sometimes rising again to seek a different spot. Once, happening to approach in the hour before sunrise, with the moon still bright in the sky, I was listening to the calls of Fieldfares and Redwings crossing the forest, when I spotted a cock harrier on his first morning flight away from the roost. Yet it was not until 1966, after receiving a postcard from Alan Paterson, announcing that he had seen ten Hen Harriers in the vicinity of the flow (it seemed an incredible number at the time), that I decided to watch the site carefully.

  At first I was disappointed, not to say disheartened. Much of January 1966 was bitterly cold, with severe frosts; sometimes wreaths of white mist settled at ground level and made observation futile. Also, some of my first watches were too distant from the more favoured roosting sites and it would have been particularly easy to miss the light-coloured males as they flickered against a background of pale grass. Later I discovered that this roost was never occupied by many harriers in still, frosty weather, for reasons which are still a subject for speculation, so it was not really surprising that my highest score after the first four watches was still only four birds. Then, on 19 March 1967, a day of gusty north-west wind tossing the old shreds of purple moorgrass about the sky, we found a good vantage point and waited without much expectation. It happened to be one of the rather rare nights, after occupation of breeding grounds has already begun, when turbulent weather seems to force a late resumption of communal roosting. From 50 minutes before, until 20 minutes after sunset we made a cautious estimate of ten to twelve ringtails and two adult males.

  Against the lingering light of a spring evening, even the grey males were darkly silhouetted and the ringtails looked black and menacing. We stayed, fascinated, till all the birds seemed to have settled and the light had become too murky for further watching. We would return, above all, to enjoy the mysterious spectacle of these gatherings in a quiet and lonely landscape, but as time went on it became compulsive to observe each facet of their behaviour and attempt to understand the motivation that lay behind it. Unknown to me at the time, R. C. Dickson had discovered and begun to watch another roost since September 1966. Later we pooled our experience and wrote a joint account of Hen Harriers roosting in south-west Scotland (Watson and Dickson, 1972). In the present account, these two major roosts are referred to as Roosts I and 2, respectively, as explained in the Introductory Note to the chapters on the Hen Harrier in south-west Scotland.

  Over the years a number of friends have shared my vigils and others have watched independently. Many have contributed information. The interest and fieldcraft of Louis Urquhart, with whom I have shared so many watches, have been unflagging and Jeffrey Watson, apart from taking part in a number of watches, has especially stimulated discussion on roosting behaviour. Records have been kept, including counts of roosting harriers, time spanned by arrival or departure, behaviour, and weather, for nearly 170 evening watches and ten in the morning at Roost 1. Evening watches have generally been from about 45 minutes before sunset until it was almost dark; in the mornings, the period from about 40 minutes before, till half an hour, or more, after sunrise was covered. R. C. Dickson has maintained a fairly similar series of watches at Roost 2, where David L. Irving has also made a number of counts. R. H. Hogg has made several counts at both roosts. In 1970–71, R. C. Dickson found another roost near Roost 2 and obtained some evidence that individual harriers moved between these two. I occasionally found very small numbers occupying a fourth roost, intermediate between Roosts 1 and 2.

  Not the least memorable aspect of these frequently damp and chilly occasions was the variety of wildlife incidentally observed. Sometimes, however, when harriers were slow to appear, the swiftly changing spectacle of a Galloway hill landscape beneath a wild evening sky was sufficient compensation in itself.

  Description of roosts

  The site of Roost 1 is a long inland valley, skirted on the north by a stream above which the ground rises gradually to form a range of hills heavily afforested with conifers. To the south the hills are open and grassy, with patches of short heather and bracken and many outcrops of rock.

  The valley is about 75 metres above sea level, with an area of virtually flat boggy ground extending over some 125 hectares, but the usual roosting sites are in a section of about 16 hectares. The vegetation is very rank, consisting of purple moorgrass, sedges, rushes, cotton grass and sphagnum mosses, with some tall stands of reeds in the wettest parts and an occasional willow bush. The ground cover is mostly 75 centimetres or more in height. Several dry rocky hillocks are grassy or sparsely covered with heather. Apart from these, and a narrow stretch of firm ground beside the stream, almost the whole area is extremely boggy and difficult for a man to traverse unless it is frostbound. Sheep, however, are able to forage over parts of it.

  The area is thoroughly water-logged by normal winter rainfall and, at first, I supposed that the harriers must sometimes roost with their feet in standing water because they continued to occupy the roost in even the wettest weather. However, after making daytime visits in search of their precise roosting places, I concluded that matted vegetation, trampled into ‘beds’ by the birds themselves, may always have provided a platform to support them. These ‘beds’, or forms, were mostly found among a strong growth of rushes; some had a base of sphagnum. From observation of birds settling, it was clear that several birds often roosted in close proximity, though a few were sometimes widely dispersed. Examination of the ground revealed many ‘beds’ separated by a metre or less, and some more or less contiguous, but only a proportion of them could have been occupied on any one night. None showed signs of long continuous occupation since most contained only one pellet. A few had two, three or even five pellets, and some none at all. In recent years, the main concentration of roosting birds sometimes shifted by several hundred metres, but these changes were generally temporary and the most favoured ground is still the same as ten years ago. At times, from December onwards, a few birds roosted among the tall reeds; presumably winter weather flattened or broke down sufficient stems to make some sort of platform, even there.

  The roosting area of Roost
2 has been described by Dickson as: ‘a low-lying coastal moor, separated from the sea by a series of sandhills planted with a belt of young conifers. The roost proper is an area of about 23 hectares, with dense heather dominant and, in the area favoured by the roosting harriers, up to a metre tall, with a field-layer of bog moss and cotton grass; in winter the ground is sometimes waterlogged. The area is ungrazed by domestic animals and natural regeneration by Scots pine is taking place in some parts of the moor.’ Heather gives way to rushes, purple moorgrass and bracken on the fringes of the moor, with clumps of willow, rowan, birch and gorse, and on one side a spread of rhododendrons. Dickson found, however, that the harriers roosted in the long heather ‘creating trampled platforms, or forms, half a metre to a metre in diameter, of dead, bent grey heather stalks’. Some forms showed signs of repeated use, sometimes on consecutive nights. He commented that Short-eared Owls probably used the forms by day since their pellets were also found there. The Craigheads found that a Marsh Hawk roost was also occupied by Short-eared Owls in daytime, some owls arriving in the mornings before the Marsh Hawks had left.

  The satellite Roost 3 which Dickson found, 2.4 kilometres from Roost 2, is also on marshy ground, interspersed with dry banks of heather and bracken, and there is an extensive area of thick willows. The harriers roosted between the trees among the heather, bracken and grasses. As at Roost 1, they generally concentrated within 15–20 hectares although the entire area covers 150–200 hectares. This roost is only ‘200–300 metres from high-water mark’. Roost 4, never much used by the harriers, is a fairly wide expanse of level, rough grassland, partly heathery and rushy, in the vicinity of a conifer plantation. It is also on low ground near the coast.

  The main features of the roosts may be summarised as follows. All are on flat ground and the three which have been occupied regularly (Roosts 1-3) are boggy, even heavily waterlogged at times. These three roosts are also virtually free from human intrusion, partly because access is either difficult or discouraged, and partly because of the boggy nature of the ground. Roost 4, on the other hand, was drier, hence more accessible, and was liable to intrusion by shooters or other people. Almost certainly this disturbance prevented it being used regularly by the harriers. Nevertheless at Roost 1 and 2 it was clear that the presence of human beings near the perimeters of the roosting areas, providing they did not enter them while the birds were in occupation, did not discourage the harriers noticeably.

  At Roost 1, occasional flight-shooting of duck and geese from the stream bank, less than 50 metres from the edge of the roost, caused temporary disturbance but the harriers re-occupied their favourite sites on the ground nearest to where shooting occurred, either later the same night or within a night or two after shooting ceased. It was, however, quite likely that harriers were sometimes shot on these occasions. At Roost 2, traffic on a public road passed the perimeter of the roost and had no effect on the birds, but the possibility that birdwatchers who left their cars to obtain better views of the harriers might draw unwelcome attention to the birds, in an area where harriers are known to be disliked by keepers, could not be ruled out.

  Although the roosting grounds were generally open enough to give the birds uninterrupted views of any predators from the air, their vision from the ground must have been limited by the tall vegetation, especially among the willow bushes at Roost 3. They probably relied much on hearing to warn them of danger while they were on the ground, and there is little doubt that the tendency to gather in close-knit groups facilitated mutual warning. It is likely that the choice of marshy ground for roosting is a deterrent to foxes, the most probable danger apart from man; yet, as stated earlier, large roosts on dry ground are known elsewhere. It seems to me more likely that lack of disturbance by man on the actual roosting ground, and the need for shelter and concealment, are the principal factors in roost site selection. In most parts of mainland Britain it cannot be too easy to find extensive areas fulfilling these conditions and they are nowadays most likely to be found in bogland.

  I have often wondered why Hen Harriers appear never to roost in trees. Presumably in their traditional habitats of steppe, moor and marsh, roosting on the ground was generally obligatory. Even so, the conifer forests which adjoin the main roosting grounds in south-west Scotland seem to offer an alternative which might be safer; and it would be interesting to know how the habit of roosting above ground, in a hedge, has arisen in Ireland.

  Arrival at the roost

  The evening arrival of harriers at Roost 1 was generally spread over a period of 40 minutes to an hour or more. Arrivals of more than ten birds were occasionally completed in as little as half an hour; the longest period was more than 83 minutes (16 birds). Once, in November, some 30 harriers reached the roost in only 40 minutes, but most of the highest totals were spread over more than an hour. Nothing resembling the mass arrival ‘in one flock’, described by the Craigheads for the Marsh Hawk, was ever seen, although up to five birds very occasionally appeared together. Usually the birds came in singly, sometimes with gaps of ten to fifteen minutes without new arrivals.

  Harriers approached the roost from all directions, but the majority came from the east or north and very few from the south-west. This pattern fitted well with the diurnal evidence of preferred hunting grounds and was confirmed by morning observation of flight-lines away from the roost. Habitat between the south-west and north-west was largely mountainous with extensive afforestation, compared with a mixture of farmland, moor, marsh and forest or woodland to the east and north. Most adult males arrived from points between south-east and north, whereas ringtails predominated, strikingly, among those which arrived from westerly directions.

  Clearly the pattern of arrival depended greatly on weather conditions during the day and at roosting time. On wet and stormy nights with early darkness, some or all came to roost earlier than on fine evenings. It was also noticeable that when the weather cleared towards the end of a very wet day, birds were slow to reach the roost and presumably took advantage of the clearance for late hunting. The subsequent manner of arrival and behaviour was also much influenced by weather conditions.

  One or two birds sometimes hunted the near vicinity of the roost during the day, and these were almost certainly the individuals which were occasionally seen over the roost as much as an hour and a half before sunset. The earliest record of a bird settling to roost was about 70 minutes before sunset. On the very wet November night, when some 30 birds arrived in 40 minutes, all settled before sunset; but generally the first arrivals were between 15 and 30 minutes before sunset and the majority appeared during the next half hour. The last birds reached the roost up to about 45 minutes past sunset. These late arrivals nearly always dropped swiftly into cover. It may be asked how it was known that no more came in after darkness prevented observation. The long series of watches, usually maintained until five or ten minutes after activity seemed to have ceased, made this an unlikely possibility, nor was this belief disproved by any of the morning counts. No obvious segregation of the sexes, such as Weis noted in Montagu’s Harriers, was observed at any of the roosts.

  At our favourite observation place, some protection from the elements was provided by thicket stage conifers from among which we could command a long view of all except the northerly approaches to the roost. From that side, the first view of a bird was often when it appeared close overhead, not much above tree-top height. Some of the most memorable evenings were those when wind and rain swept across from the west and many birds came labouring up the valley from the east, in low-level flapping flight. On such occasions a procession of three or four well-spaced birds might be in view at once. On dry nights they tended to come in high; and so many were unseen until they appeared, almost miraculously, right over the roosting area, that we surmised that their initial approach may have been at a great height. Fine nights with a fresh or strong breeze were occasions for aerobatics, as many as five or six birds circling and playfully diving at each other, obviously in no hurry t
o settle.

  Unless the weather was either very windy or wet, some of the earlier arrivals always perched for a while, sometimes for 15 or 20 minutes, on posts, hillocks, rocks or a stone dyke overlooking the roosting sites. This was particularly noticeable on the evening of 23 November 1975 when the weather suddenly cleared after a very wet day; four ringtail harriers remained perched and preening on adjacent posts, and one on a nearby rock, for the best part of 20 minutes. This may have been no more than a ‘drying out’ exercise but, since they represented part of a sudden increase in numbers of ringtails, they could well have been tired birds which had just completed a migrational flight. The weather pattern was almost exactly similar on 3 November 1974 when, again, four ringtails had perched close together on the perimeter of the roost for over ten minutes before flying in.

  Schipper (pers. comm.) had the impression, at a communal roost in Flevoland, that birds from hunting ranges with low numbers of voles arrived late at the roost. At Roost 1, early arrivals in wet and stormy weather suggested that it was unprofitable to continue hunting late in such conditions. It seemed likely that the spread of arrival time depended to some extent on the distances to hunting grounds. On 28 October and 6 November 1975, the same number of birds (six) and the same proportion of males and ringtails came to the roost. They arrived in the following sequences:

  It seems probable that the same individuals were concerned on both evenings, and the very similar intervals of time might have been due to the distances they had to travel from hunting grounds. Both evenings were fine, with light winds.