The Hen Harrier

  Fig. 11 Hunting grounds in the breeding season,

  within 7 km of centres of Areas C, D and E

  * After this was written I found a young rabbit as prey, in July 1976, when young had fledged from a forest nest. The rabbit had been almost entirely eaten, except for the forelegs and decapitated head, and the remains had been left with other prey remains in a ride near the nest. Other prey identified at this site consisted of two young Black Grouse and a Chaffinch. There were also feathers of other passerine species and a larger skeleton, probably of a young Pheasant.

  * Lack of grazing or burning on the higher ground, above forest plantations, made this area much less open and so less suitable for many of the birds preyed upon by harriers. Cock harriers, particularly (Schipper, 1975), may have greater difficulty in locating prey in the uniformly dense ground cover. Although the ground became more open again at the highest elevations, above 500 metres, prey was obviously scarce there.

  CHAPTER SEVENTEEN

  Food and Hunting Grounds: September–April

  Prey identifications, out of the breeding season, were made from pellets collected at a winter roost (Roost 1) adjacent to Area C, and the results are given in Tables 27–28. Birds (70%), were the most important items but the proportion of mammalian prey (30%), was notably higher than in pellets found in the breeding season (beetles and moth cocoons are excluded from these percentages). A strikingly different feature of the winter pellets was the large amount of rabbit, hare or mountain hare which they contained (52% of mammal items). Some young rabbits and mountain hares were probably available in the period, but it is likely that some of this prey was taken as carrion. The importance of carrion food, generally, has been discussed in Chapter 4. Only one local observation, by Louis Urquhart in September 1975, of a female Hen Harrier at an old rabbit carcase, gave presumptive evidence of carrion feeding. Rabbits incapacitated by myxomatosis or caught in snares might also have been eaten. Field and water voles and a pigmy shrew made up 29% of the mammalian items, and unidentifiable fur was present in a few pellets. The highest proportion of Red Grouse occurred in 1968, a year of high grouse population in the district. All grouse taken in the period must, of course, have been full-grown, but they may have included birds which were already dead from other causes. The proportion of Red Grouse among avian prey (11%) was precisely the same as in summer pellets. It was very difficult to make specific identifications of small bird prey from pellets as harriers, unlike owls, generally reduce skeletal parts to very small fragments. Nevertheless, small, mainly passerine, birds (72% of bird prey) were clearly very important in winter, as in summer.

  R. C. Dickson (1970) has published findings from an analysis of 32 pellets collected in December 1966, January 1967 and September 1968, from the other major roost in the region, on low ground near the coast (Roost 2). These showed an even higher proportion of birds (89.8%) than at Roost 1. Bird prey identified was mainly passerine, including finches, buntings, pipits, Skylarks, Starlings and a Wren; in addition there were two waders (one Redshank, one unidentified) and only one grouse (species unknown). Mammals (10.3%) consisted of three mice or voles and one shrew. Two small beetles have been excluded from these percentages. More recently, Dickson has collected a much larger number of pellets from Roost 2. The full analysis of these is not yet complete but he has kindly given me the results for 21 pellets. These contained 81% birds (mostly pipits, Skylarks, finches and Starlings) and 19% mammals (all rodents, the largest of water vole size). These later findings are therefore very similar to Dickson’s published evidence on winter prey from pellets at this roost.

  Fig. 12 (a–b) Sightings of hunting Hen Harriers within

  24 km of Roost 1, Sept.–March

  Winter hunting grounds covered a much wider range, with a greater variety of habitat, than in summer. Of 253 sightings of hunting Hen Harriers between September and March, within a radius of 24 kilometres of Roost 1 (see Figure 12), 68 were on moorland above 150 metres*, 69 on grassland and cultivation below 150 metres, 66 on marshland below 150 metres and 50 were over woodland (mainly young conifer forests, but occasionally hardwood scrub). As such these figures are difficult to interpret owing to differences in the relative extent of these habitat types and to a further possible bias due to uneven observer cover.

  In the Netherlands, Schipper found that males showed a hunting preference for less structured vegetation, and took more agile passerine birds, than females. In south-west Scotland ringtails slightly exceeded adult males on moorland over 150 metres (often with long heather), woodland and marsh, but sightings of each on pasture and cultivation below 150 metres were nearly equal (Fig. 12). In general, the latter habitat had the barest ground and undoubtedly held the largest numbers of passerine birds. The tendency for males to hunt the most open ground is further emphasised by my analysis of 34 winter sightings of Hen Harriers in a particularly open area of salt-marsh and coastal fields, near the estuary of the River Nith; these consisted of 20 grey males (59%) and 14 ringtails (41%). It must be remembered that all totals of ringtails also included a number of brown, first year males. It is interesting that the proportions of males to ringtails hunting either moorland above 150 metres, or marsh with tall vegetation at lower elevations, were so similar.

  Fig. 12 (c–d) Sightings of hunting Hen Harriers within

  24 km of Roost 1, Sept.–March

  There was some seasonal change in preferred hunting habitats during the winter. There was an indication (see Fig. 12) that higher moorland was hunted mainly in September–October and March, lowland marshes mostly in November–December, but the largest number of sightings on grassland and cultivation below 150 metres was obtained in December and February. There was a peak in woodland sightings, with some suggestion of a move to lower elevation woodland, in January. In general, there was an indication that Hen Harriers moved to hunting areas on lower ground between November and February, which may reflect changes in the distribution of available prey. In the early winter there are many Meadow Pipits, Skylarks, Fieldfares and Redwings on upland moors, while there is certainly a decline in small birds in the marshes as the vegetation dies down in mid-winter (a reed-bed Starling roost, for example, was usually deserted about December). The peak numbers of sightings on farmland, in December and February, may well be associated with the distribution of winter flocks of passerines such as finches and buntings. On several occasions harriers were seen hunting over fields of kale or weedy stubble where flocks of passerines gathered to feed, and a harrier was sometimes seen attempting to make a kill there. The same fields provided feeding for finch flocks for many weeks at a time and I gained the impression that harriers made regular hunting visits to such sites. It is possible that the distances travelled from roost to hunting grounds were more restricted on the shortest winter days, or when the weather was particularly wet or stormy; some evidence on distances travelled is given in the following section on winter roosts.

  Although duck or geese rose in alarm from fields or marshes as a harrier approached, the prey sought usually appeared to be smaller birds or mammals. It was common to see a harrier hunting assiduously over the tall grasses or reeds in the river valleys, when it would pounce into the vegetation at intervals and sometimes remain hidden for several minutes. Occasionally, two harriers were seen hunting in close proximity and once, in autumn, three were seen together in association with larger numbers of Kestrels, Short-eared Owls and one or two Buzzards, over young conifer forest and grass moor where voles were abundant. A similar association of predators, including two Hen Harriers, was seen hunting a field of un-harvested oats in January. Here too the prey sought probably consisted of small mammals. As noted in Chapter 4, R. C. Dickson has observed up to five harriers at one time hunting a large kale field in winter. Harriers which hunted young conifer forests or upland moors in mid-winter were most likely to be preying on voles or grouse, as there appeared to be very little other prey available at that season in those habitats. They may, however, have found some pass
erine birds such as Bullfinches (present in winter pellets) in the forests. In March, I have watched a female Hen Harrier over 15 year old conifer forest, shadowing flocks of Starlings as they flew to and fro before roosting. It made no attempt to attack them at this stage but may have caught one later, as I saw it reappear in the dusk with its crop bulging. Possibly it was in the habit of raiding the Starling roost as the birds settled. Some of the invertebrate prey, such as the great water beetles Dytiscus marginalis, could have been taken on the marshy roosting grounds; water voles may also have been taken there. Weighing 120–180 grams in winter (Southern), a water vole would provide a substantial meal for a harrier.

  There was evidently some variation from year to year in the proportions of wintering adult males and ringtails. Table 29 gives the number of diurnal sightings of each, within 24 kilometres of Roost 1, in three winters for which records were most complete. Since in these years, particularly, there was fairly consistent observer cover for all the main habitats, the numbers probably give a good indication of the true proportions. The counts at Roost 1 (Fig. 13) support the impression given by Table 26 that males were more numerous than ringtails during much of the winter 1973/74, and that they were strikingly less numerous than ringtails in the winter 1975/76. This latter impression was further supported by observations at the neighbouring coastal area of Caerlaverock, where 1975/76 was the first winter in five when ringtails were seen more often than males. In general, throughout south-west Scotland, the proportion of adult males, in winter, is much higher than in districts such as Orkney and the Scottish Highlands. As I have mentioned in Chapter 7, there is some indication that adult males are also relatively plentiful in winter in other western parts of Scotland.

  The total numbers of diurnal sightings of hunting harriers within 24 kilometres of Roost 1, varied little between October and February, but were less than half as many in September and March (Table 30). These figures and the roost counts (Fig. 13) suggest that many of the harriers were winter immigrants to the area (see also discussion of counts at Roosts 1 and 2, Chapter 18).

  Harriers dispersing to hunt from Roost 2, in a coastal area, frequently hunted lowland fields and marginal land. Such habitats were easily accessible to the roost and covered a wide area. Small areas of moorland and young conifer forest were available near the roost but the harriers would have had to travel at least 8–10 kilometres to find large areas of moorland and the pellet analyses from this roost suggest that they did not regularly hunt these areas. Extensive young conifer forests were considerably more distant.

  * Less than half of these were above 225 metres, most in September and March.

  CHAPTER EIGHTEEN

  Communal Roosting in Winter

  Introduction

  I have referred in Chapter 1 to the habit of communal roosting among harriers and noted that it occurs in at least six species.

  The earliest reference to this habit in the Hen Harrier appears to belong to Selby (1831) who wrote: ‘It roosts upon the ground in very long heath or ling and generally in companies of five or six together, males and females indiscriminately’. There is, however, no certainty that Selby’s observations preceded those of Sir William Jardine or, indeed, that he discovered the behaviour independently, as the two men corresponded and made expeditions together. Jardine (1834–38) gave a fuller account and there is every reason to believe that this referred to his native Dumfriesshire in Scotland. He noted that after the young were ‘perfectly grown they, with the old birds, left the high country and returned . . . to hunt fields of grain with great regularity.’ He continued: ‘At night they seem to have general roosting places, either among whins or very long heath, and always on some spot of open ground. On a moor of considerable extent I have seen seven in one acre. They began to approach the sleeping ground about sunset; and, before going to roost, hunted the whole moor, crossing each other, often three or four in view at a time, gliding along in the same manner as described by Dr Richardson of the C. Americanus [the Marsh Hawk]. When they approached the roost they skim three or four times over it, to see there is no interruption and then at once drop into the spot.’

  Although it was to be expected that with the decline of the Hen Harrier during the nineteenth century the opportunities for observing communal roosting would become fewer, it is perhaps surprising that during all the rest of that century and, indeed, well into the twentieth, no further references to the habit can be found. Macpherson (1892) however, gave the following account of a ‘pair’ roosting together in the winter of 1886. ‘They were adult and had appeared in the same locality [Weddholm and Bowness Flows, Cumberland] in previous winters but had never remained to breed. Smith noticed the spot in which they roosted in tall heather and, after watching them for several weeks, killed one and then the other.’ Walpole-Bond (1914), also, said that ‘in Orkney a pair in winter patronise a special roosting site in very long ling’ and the gamekeeper naturalist, Dugald Macintyre (1947), wrote: ‘the keeper saw Circus and his less perfectly plumaged son meet together at dusk, after each had consumed a Snipe, and he noticed the direction in which the couple flew to roost. Their roosting place was on a bank of very tall heather and there were piles of castings on the ground.’ In March, he watched a pair fly in company from their home glen each morning. ‘They parted, on reaching the first marsh, to hunt in different directions but, the hunt over, they met and flew home together, to roost side by side in the heather.’

  It is remarkable that the next British reference to communal roosting should come from Kent, a district where the Hen Harrier’s status is that of a far from plentiful visitor from autumn till spring. This account of a roost of up to three males and two ringtails in Walland Marsh during November and December 1953, by F. J. Walker, appeared in the Kent Bird Report for that year and was later mentioned in British Birds 49 (1956). The birds roosted in tall grass growing in 15–22 centimetres of water. These accounts seem to have passed almost unnoticed by most ornithologists.

  Meanwhile, communal roosting of the Marsh Hawk had been observed in Florida, by Stoddard (1931). He wrote: ‘This species has the unhawklike habit of roosting on the ground, frequenting the same spot night after night. If numerous the hawks form a loose roosting group numbering from two to three, to as many as thirty. A large field grown up to heavy broom sedge and preferably upon a hilltop is chosen as a roosting site. Each bird has a beaten-down spot in the sedge, well “limed” with droppings.’ Further accounts of Marsh Hawk roosts, in Pennsylvania, were given by Randall (1940); in Missouri by Weller (1955), in Michigan by the Craigheads (1956), and in Indiana by Mumford (1975). All these roosts were on the ground, with vegetation tall enough to provide cover for the birds, in otherwise open country. Sites were either wet, with vegetation growing in several centimetres of standing water, or dry among tall prairie grass or a luxuriant growth of weeds on stubble or wasteland. The Craigheads noted a maximum of 48 birds at a roost in late January–early February and Mumford counted at least 60 on 24 January.

  Brown and Amadon suggest that communal roosts may be more common in the Marsh Hawk than in the Hen Harrier, but a number of recent European accounts indicate that they occur in most or all of the regions where the winter population of Hen Harriers is considerable.

  Roosts in Germany have been described by Baron Geyr and Haas (1957), by Andris (1970) and by Jakobs (1971). Wassenich (1968) studied a roost in the Grand Duchy of Luxembourg and Mois (1975) has given an account of roosting at two sites in Belgian Lorraine, where detailed observations were made from 1968 to 1974. Several roosts in the Netherlands have been closely studied by Schipper and others since 1969–70, in which season a maximum count of 54 birds was noted at one roost. In correspondence, Schipper has told me that more than 60 birds were seen at a roost in Bulgaria by Johnson and Biber.

  Later in this chapter the study of communal roosts in south-west Scotland, between 1966 and 1976, is described. Maxima of at least 30 birds have been seen at two different roosts in this region. This accou
nt adds new material to, and modifies, some conclusions in an earlier publication (Watson and Dickson, 1972).

  The late Ernest Blezard told me of a small roost in the southern Highlands of Scotland, discovered by E. Blake in 1960, while from 1971 onwards Balfour and Green observed several roosts in Orkney. The latter were generally small but at one roost up to 20 birds were noted. Other roosts—none very large—have been found within recent years on Deeside by N. Picozzi, in Speyside by D. N. Weir, in the Inner Hebrides by R. H. Hogg, Allan Goodin and others, and in south-west Ireland by Frank King.

  Descriptions of these roosting grounds generally show a similarity with those of the Marsh Hawk, but in Europe by far the majority of sites is in damp or marshy ground. In Flevoland, Netherlands, the birds roost close together in small reed-beds but are well scattered in larger tracts of reeds (Schipper, pers. comm.). All the Orkney sites are on marshy ground, especially among reeds, but in autumn and early winter when reed beds are still dense the birds tend to favour rank, mixed vegetation, with plants such as meadow sweet (Filipendula ulmaria). At the first site discovered in Belgian Lorraine the ground was marshy but flooding apparently caused the harriers to desert this site. Thereafter they roosted in drier conditions on an extensive, sloping heather moor, three kilometres from the first site.

  The only sites on perfectly dry ground have been reported, by Andris, on wasteland in the Upper Rhine valley. King, however, found the most exceptional roosting site of all, in Kerry, Ireland; the birds spent the night in a low ragged hawthorn hedge, on neglected farmland which had reverted to moor. His observations at this site were made only in October, over two years, when the hedge was still leafy, and he doubted if it would provide sufficient cover later in the winter.