Observations on the Breeding of the Golden Plover in Great Britain

Transcription

1 Bird Study ISSN: (Print) (Online) Journal homepage: Observations on the Breeding of the Golden Plover in Great Britain Derek A. Ratchliffe To cite this article: Derek A. Ratchliffe (1976) Observations on the Breeding of the Golden Plover in Great Britain, Bird Study, 23:2, , DOI: / To link to this article: Published online: 23 Jun Submit your article to this journal Article views: 643 View related articles Citing articles: 27 View citing articles Full Terms & Conditions of access and use can be found at

2 Observations on the Breeding of the Golden Plover in Great Britain CONTENTS by D. A. Ratcliffe Introduction 63 Breeding distribution in Britain 66 Nesting habitat.. 69 Nest-finding 76 Laying, clutch size and hatching 80 Food and feeding grounds Variations in breeding density 87 Breeding dispersion 93 Population changes 98 Possible causes of decline 102 Relationships with other waders 110 Conservation ' ' 112 Acknowledgements 113 Summary References ' Vignettes by Kevin Baker INTRODUCTION THE GOLDEN PLOVER Pluvialis apricaria IS ONE OF THOSE FASCINATING BIRDS which are in some ways familiar and in others mysterious. Whilst too numerous and widespread to have excited interest as a rarity, the species is too elusive to have attracted much serious attention in research. In spring and early summer the Golden Plover shares with the Red Grouse Lagopus lagopus scoticus the distinction of being par excellence the bird of the British moorlands, where its voice seems to epitomise the character of these wild places. Yet this upland association is only partial and the Golden Plover spends more than half the year on the farmland of the lowlands, or the saltings of the estuaries. The majority of invertebratefeeding birds of the uplands leave the high ground to winter elsewhere, but the Golden Plover departs earlier than most, and the first successful breeders with their progeny flock together and desert their breeding grounds for the low country even by the beginning of July. 115

3 64 BIRD STUDY Figure 1. Distribution of suitable breeding habitat of Golden Plovers in Great Britain The British bird is thus a partial migrant, occupying two main and contrasting habitats. Only as a breeding bird is it a northern and submontane to montane type, and even then in many areas the Golden Plover feeds to a considerable extent on the improved and enclosed grassland where the uppermost hill farms and crofts pass into the open sheep walks, grouse moors and deer forests. The young, nevertheless, have to feed on the ground where they are hatched, at least until they are big enough to move away. It is capable of exploiting some of the most barren habitats in Britain for nesting, yet in other respects shows a close dependence on ground of fairly high fertility. There are few accounts of the breeding biology of the Golden Plover in Britain, that by D. and C. Nethersole-Thompson in Bannerman (1961) being by far the most detailed and informative. Behaviour has been dealt with in detail by these

4 BREEDING OF THE GOLDEN PLOVER A glta "PS 3 r inn _..,.. : 2:::.. '.....,. ', :es.:::%:::: : I ::::. : 11.1:..9ii ci`r.: iii _...,,.: ' hil III \VA:: I; k II in ,.: : 11 MINI 113 iii,,-,, s,, 1 i, ,, ----,,:*,,d. M...* Ill Figure 2. Golden Plover distribution by 10-km squares. (Large dots = proved breeding; medium dots = probable breeding; small dots = presence in breeding season.) From The Atlas of Breeding Birds of Britain and Ireland, compiled by Dr J. T. R. Sharrock for BTO/IWC. writers and by Williamson (1948), whilst Sauer (1962) has given an exhaustive account for the Asiatic Golden Plover Pluvialis dominica fulva. I shall therefore make only limited reference to this topic. The present contribution aims to amplify certain aspects of breeding biology, especially in regard to distribution, status and population regulation. My observations are of a limited and patchy kind, and in no sense represent an attempt at comprehensive study of even a restricted population. If anything they have identified more problems than they have solved, but may thereby direct attention to aspects requiring critical study. The majority of the observations are from northern England and Scotland, and I have no experience of the Golden Plover in Ireland. Certain statistics on breeding are supplemented by data from the British Trust for Ornithology Nest Record Cards, 214 in number, submitted from inclusive. Though mostly for single visits, these records are valuable in representing

5 66 BIRD STUDY a more random sample than my own. Their only marked geographical bias appears to be in the rather high proportion from the Pennines, and in the absence of any from Ireland. BREEDING DISTRIBUTION IN BRITAIN The Golden Plover is confined as a breeding species to certain kinds of moorland and mountain terrain. The nesting habitats consist of a wide range of hill grasslands, dwarf shrub heaths, moss and lichen heaths, and peat bogs. On rare occasions, Golden Plovers have nested in anomalous lowland habitats in England (e.g. on stony ground adjoining a Lincolnshire airfield in 1945, P. Wormell, pers. comm.; Lundy Island, 1925; Oswestry race-course, Shropshire, 1914), but the habit does not persist. There are no breeding records from the heather Calluna vulgaris heaths and valley bogs of southern and eastern England, which have affinities with some of the lower-level northern heather moors, and closely resemble heathland in Germany and Denmark where the species nested until recently. Although the Golden Plover breeds in virtually all the main mountain and moorland areas of Great Britain from Dartmoor northwards, its numbers vary greatly according to terrain. Nesting is confined to flat or gently sloping ground, and few nests are placed on slopes exceeding Sharply contoured mountain systems, with prevalence of steep slopes, narrow watersheds and small, peaked summits, thus tend to be unsuitable; and the bird is sparse or locally absent on many of the more precipitous hill ranges, especially in western Britain. In the Lake District it is scarce or absent on the rugged hills of the central Scafell area, but occurs sparingly on the more rounded tops of the Skiddaw group and the High Street range to the north and east. There is a general tendency for the extent of moorland and peat bog to increase with distance northwards and westwards in Britain (Figure 1). This is partly a reflection of the distribution of high land, but is also associated with major gradients of climate. Mean temperature decreases from south to north, and oceanicity increases from east to west. The result has been to produce huge areas of barren moorland and blanket bog, not only on the hills, but also in the lowlands of northern and western districts such as Caithness, Sutherland, the Outer Hebrides, Orkney and Shetland. An altitudinal descent of 'life zones' of vegetation and animal communities towards the extreme north and west is involved. As an example, blanket bog, a favourite nesting habitat, is restricted to elevations above m at its southern limits, on Dartmoor, whereas ;n western Scotland, from Argyll northwards, it descends almost to sea level. The breeding distribution of the Golden Plover shows parallel trends, with both an increasingly continuous extent (Figure 2) and an associated downwards expansion (Table I) with distance north and, to a lesser extent, west. The geographical pattern is closely matched by other birds such as the Merlin Falco columbarius and Ring Ousel Turdus torquatus, and by plants such as the cowberry Vaccinium vitis-idaea and crowberry Empetrum nigrum. Towards its southern limits, though, the Golden Plover is clearly not restricted in either lateral or vertical distribution entirely by lack of suitable habitat. In Wales, and even southwest England, there is a good deal of suitable habitat over a wide range of elevation which is not used for nesting. Some other factor is evidently partly responsible for the species'

6 TABLE I. LATITUDINAL District Dartmoor, SW England South Wales North Wales Peak District North York Moors Northern Pennines Lake District Scottish Borders Eastern Southern Uplands Dumfries and Galloway BREEDING OF THE GOLDEN PLOVER 67 DESCENT OF LOWER LIMITS OF BREEDING OF GREAT BRITAIN Lowest elevation for breeding Mainly over 500 m Formerly occasional at 305 m, now seldom below 452 m Nowadays probably 365 m on Denbigh Moors Formerly common at 275 m, now mainly above 395 m Regularly at 213 m, occasionally lower Small numbers down to 245 m, but mainly above 365 m Formerly down to 245 m around Devoke Water, Eskdale Regularly at 92 m in north Cumberland up to 1955, now mainly above 245 m Down to 200 m in the Lammermuirs but mainly above 305 m Once occasional on coastal raised bogs near sea level, small numbers regular from 92 m but mainly above 150 m GOLDEN PLOVER IN Observer A. Archer-Lock H. Money Salmon W. M. Condry E. K. Allin J. M. Harrop W. Gibbs D. W. Yalden A. Barnard D. A. Ratcliffe M. Garnett D. A. Ratcliffe D. A. Goode H. Gladstone A. D. Watson D. A. Goode In the Scottish Highlands and Islands nesting is frequent below 92 m and even down to near sea level, where suitable habitat occurs at these low elevations, i.e. mainly in the west and north. Even here, suitable habitat is local and limited in extent below 150 m, especially in many coastal areas, so that low level breeding is somewhat patchy. In the east, nesting occurred almost at sea level on the Tentsmuir dunes, Fifeshire, until 1938; but south of the Moray Firth, nesting below 150 m is now unusual, there being little suitable ground. In the west of Ireland, Golden Plovers occur almost at sea level at their southern limits, on the great blanket bogs of Galway and Mayo. This district is ecologically similar to the Western Highlands, with marked altitudinal depression of 'life-zones' under an extreme oceanic climate. generally sparse distribution and restriction to higher levels in these two regions. Variations in breeding density show a still less clear-cut geographical pattern. Only from the Peak District northwards does the bird regularly occur in good numbers and at a more than moderate density. There are, nevertheless, large areas of the Scottish Highlands and Islands where it is found only at low densities, despite a great extent of apparently suitable nesting habitat. The highest breeding densities known at present are in parts of the Pennines, the eastern Southern Uplands, and the eastern Grampians, although the total populations in some other districts of the Highlands and Islands are considerable. The high densities are mostly at intermediate elevations of m, though breeding occurs widely below and above these levels. There are many pairs below 150 m in northern and western Scotland, and on suitable terrain breeding occurs down almost to sea level, as in the Hebrides and Shetland. Golden Plovers nest regularly at 1,040 m and occasionally still higher in the Cairngorms, and there are, in total, considerable numbers nesting in montane habitats of the kind favoured by the Dotterel Eudromias morinellus on high plateaux, ridges and spurs at 760-1,000 m.

7 68 BIRD STUDY Analysis of the BTO Nest Record Cards into altitudinal groups gives the following picture (n = 186 for records with elevation noted): Altitudinal range (m) Number of nests This probably accords reasonably well with the spectrum of altitudinal distribution of the Great Britain population as a whole (there are no Irish Nest Record Cards). The sharp cut-off above 610 m reflects the typical maximum elevation of upland plateaux favoured by the species; suitable breeding grounds at higher altitudes cover a relatively small area. It has been claimed that some of the Golden Plovers which breed on the high plateaux of the Grampians and Cairngorms are representatives of the Northern race altifrons. Black-faced and black-fronted birds of Northern appearance can occur almost anywhere in the British breeding range, but it seems fairly clear that there is a cline in which their frequency increases with latitude and, to some extent, altitude. Dark birds thus appear most frequently on the high Grampians and Cairngorms, and there is here a greater chance of their being mated to another dark bird. While there is a continuous range of variation in plumage from Southern to Northern birds, the geographical gradient in the proportions of the various forms appears to be uneven. Few of the dark birds which remain to breed in Britain are as black-fronted, with clear white margins, as the true Northern Golden Plovers; the latter pass through this country, heading for their Arctic and Subarctic nesting haunts, when many of the British breeders already have eggs. Wynne-Edwards (1957) has contested the validity of trinominal designation in this type of polymorphism (see also Gordon et al. 1957). Voous (1960) suggests that Pluvialis apricaria is 'a not very successful northeast Atlantic isolation product of the circumpolarly distributed group of Golden Plovers'. This is by reference to the smaller but more widely distributed P. dominica, which replaces the other through the Arctic regions of eastern Siberia and North America. The viewpoint seems arguable, but the Golden Plovers as a whole could be regarded as a successful species-group. The British birds would appear to be at one end of the spectrum, as a divergent insular form adapted to an oceanic climate, with a diminished migratory urge, and reduced seasonal dimorphism in plumage. Within this country it is an eminently successful species, and in no sense relict or 'fringe', e.g. by comparison with the Dotterel. I have attempted to estimate the present total breeding population for the British Isles by applying known densities for particular habitats and areas to all the 10 km squares in which nesting was certain or probable during the Atlas survey (Figure 2). The estimates for the various districts, in number of pairs, are as follows: Southwest England 10 Wales 900 Northern England 7,550 Southern Uplands 3,850 Highlands south of Great Glen 8,725 Highlands north of Great Glen 6,820 Hebrides, Orkney and Shetland 2,000 Ireland 900 British Isles Total 30,755 pairs

8 BREEDING OF THE GOLDEN PLOVER 69 NESTING HABITAT Golden Plovers formerly nested occasionally on the boggy tracts and dry heaths which form clearings, sparsely grown with small trees, amongst the Scots pine cover of the foothills on the Spey valley side of the Cairngorms (D. and C. Nethersole-Thompson 1961(a)). They have evidently disappeared from such places, which represent their boreal forest bog habitats in Fenno-Scandia, and virtually all the present British breeding haunts are on treeless moorlands and bogs. Throughout its total northern hemisphere breeding range, the Golden Plover is primarily a bird of treeless moors and tundra. A great deal of the British nesting habitat below the potential tree limit has, however, been created by human destruction of the earlier forest cover (see p. 98). Perhaps the most typical nesting habitat is grouse moor at m, in areas such as the Pennines, North York Moors, Cheviots, eastern Southern Uplands and eastern Grampians. These moors are characterised by prevalence of ling heather Calluna vulgaris, which shows a gradient from complete dominance on dry ground to codominance with cotton-grass Eriophorum vaginatum on wet ground where blanket bog has developed. Rotational moor burning is an integral part of management for grouse, and benefits the Golden Plover by keeping the vegetation short on some of the ground at any one time; for this bird usually avoids vegetation more than 15 cm tall for nesting, preferring ground where the growth is particularly short. Exceptions to this rule have repeatedly been found in the Moorfoot Hills, where nine out of 110 nests (8%) were in heather and cottongrass from cm tall, a few being almost completely hidden. The majority of nests on grouse moors are nevertheless placed on ground burned within the previous eight years. Nesting in short vegetation has presumably evolved as a defence against predators, giving the sitting bird optimum visibility. This may be related to the Golden Plover's frequent habit of leaving the nest when intruders are far distant; yet the Greenshank Tringa nebularia, which nests in equally short vegetation, almost invariably sits tight as intruders approach. It may also be related to the newly hatched chicks' need to make their way through and find food amongst the vegetation, both activities being more difficult in a rank compared with a short growth. On dry heather moors, the bare ground strewn with charred or bleached stems of heather after recent burning is favoured nesting habitat, especially in eastern

9 70 BIRD STUDY districts. Such ground may become partially grown with mosses, crustaceous lichens or grasses before the heather becomes dense again, and sometimes regeneration may take a good many years. Freshly burned ground is occasionally used but the later stages in the burning cycle match the plovers' cryptic plumage pattern more closely and are more favoured; and they often nest where grasses are mixed with short heather. On the blanket bogs, the most favoured of all nesting habitats, other dwarf shrubs such as Empetrum nigrum and Erica tetralix may locally be more abundant than Calluna, and Eriophorum vaginatum is sometimes subordinate to E. angustifolium and Scirpus caespitosus. At higher levels, the cloudberry Rubus chamaemorus is often a constant in the Calluna-Eriophortim bogs. In many parts of the Pennines, repeated burning and heavy grazing have eradicated Calluna from the blanket bogs, producing fairly pure stands of cotton-grass, but in some areas, such as the Peak District, Empetrum nigrum and Vaccinium myrtillus have taken the place of heather. On many of the higher moorland plateaux, the blanket bogs have become severely dissected ('hagged) by deep ramifying gullies which cut back gradually from the edges to the centre and may, in extreme cases, link up so completely that sheet erosion of the peat eventually occurs. Species such as Eriophorum angustifolium, E. vaginatum, Nardus stricta and luncus squarrosus often recolonise the eroded ground where the peat has wasted away, almost down to the underlying mineral soil and stones. By contrast, undisturbed bogs often have a high cover of bog-moss Sphagnum and frequent wet hollows or open pools. In Caithness and east Sutherland in particular, such blanket bogs cover vast areas of flat or gently undulating moorland at low elevations ( m), and similar llowes' were formerly extensive in Galloway and the Borders. In western districts they have a good deal of Molinia caerulea, Scirpus and the moss Rhacomitrium lanuginosum, and in the Highlands the last species is locally abundant in high level bogs. Littledisturbed blanket bog also locally has a high cover of 'reindeer moss' lichens such as Cladonia arbuscula, C. impexa and C. uncialis, but where there has been burning and drying, especially on shallower peat, Parmelia physodes and hypnaceous mosses are often abundant. In western Scotland, the lower hill slopes of the deer forests and sheep walks have a great deal of damp ground with shallower peat than real blanket bog, and a vegetation best termed 'wet heath' in which heather is mixed with Scirpus and/or Molinia. Here, too, excessive burning and grazing have destroyed the heather in some areas, and produced pure stands of Scirpus and Molinia. The repeatedly burned, drier moraine ground has a characteristic open community with sparse growths of Erica cinerea, Potentilla erecta, Narthecium ossifra gum and Rhacomitrium lanuginosum. In the uplands of Wales, northern England and southern Scotland, management for sheep on ground of intermediate wetness has caused extensive replacement of damp heather communities by types with dominance of Nardus stricta and/or Juncus squarrosus. Dense, pure stands of the Juncus are largely avoided by nesting Golden Plovers, probably because the birds cannot easily scrape a hollow amongst its tough basal rosettes. The mixed communities and especially the less dense

10 BREEDING OF THE GOLDEN PLOVER 71 growths of Nardus with finer grasses or moss are, however, quite favoured nesting habitat on these sheep walks. On still better-drained ground, the dry heather communities are converted by heavy burning and grazing to short grasslands of Festuca ovina, Agrostis tenuis, A. canina and Deschampsia flexuosa, though Vaccinium myrtillus often becomes dominant as a transitional stage. These types are usually on ground too steep to have nesting plovers, but where the grassland occurs on more level ground it provides sites. On eastern moorlands bracken Pteridium aquilinum, which often invades degenerating heather ground in dense beds, is occasionally used for nesting, early in the year when the old growth lies dead and flattened. The above vegetation types belong essentially to a submontane complex widespread below the potential forest limits. Some of the described communities extend with only slight variation to much higher levels, as do the Festuca, Agrostis, Deschampsia and Nardus grasslands, Juncus squarrosus heaths and Eriophorum bogs. There are, however, gradual transitions to more distinctively montane types on the windswept higher spurs and summits. These include prostrate Calluna heaths often rich in other dwarf shrubs and the lichens Cladonia arbuscula, C. ran giferina and C. uncialis; dwarf Vaccinium myrtillus V. vitis-idaea and Empetrum nigrum E. hermaphroditum heaths also with lichen-rich facies; alpine Nardus grasslands; and Rhacomitrium lanuginosum carpets with varying amounts of Carex bigelowii. This montane vegetation complex is especially well represented on the massive, rounded summits and dissected plateaux of the central Highlands at 760-1,050 m. The terrain here is often subject to frost-heaving and downslope movement, producing patterned ground, though shallow blanket bog with Eriophorum vaginatum extends to around 915 m or even higher on stable ground. These are the characteristic haunts of the Dotterel, and the Golden Plovers breeding here are sometimes in fairly close company with this species. Such montane habitats correspond to the tundra and fell-field habitats occupied by the two races of Lesser Golden Plover Pluvialis dominica in Arctic regions (Sauer 1962, Parmelee et al, 1967). Virtually all the vegetation described so far belongs to acidic soils and peats of low fertility, which predominate in most British uplands as a result of the cool oceanic climate and general prevalence of hard, base-poor rocks. In a few districts, outcrops of calcareous rock give soils of much higher base-status and fertility, but these are extensive only on the Carboniferous limestone of the Pennines, and even here they occur mainly on steep slopes and rocky ground. Only in a few places, between Alston Moor in the north and Malham in the south, are there appreciable areas of short limestone turf over fertile brown earths on level or only gently sloping ground. The vegetation is a close-grazed species-rich grass sward in which Festuca ovina, F. rubra and Agrostis spp. are mixed with Anthoxanthum odoratum, Sesleria caerulea, Briza media, Sieglingia decumbens, Carex caryophyllea, Luzula cam pestris, the tiny shrub Thymus drucei, and numerous small forbs, e.g. Plantago lanceolata, Achillea millefolium, Viola riviniana and Campanula rotundifolia. The hummocks are often capped with a cushion of 'reindeer moss' lichens, and the sward grades into more acidic, species-poor types. These areas hardly exceed 120 ha in any one continuous stand, and occur mainly within the narrow range of m. They are thus submontane, though

11 72 BIRD STUDY on some of them frost effects have thrown the ground into extensive systems of hummocks. The Golden Plover is so usually associated with the acidic moorlands in Britain that these limestone grasslands are atypical habitat, and the Lapwing Vanellus vanellus is the most characteristic wader. They are, nevertheless, important Golden Plover breeding places in terms of ecological interest. The over-riding feature of the Golden Plover's nesting place is that it is on unenclosed land. There appears to be a general inhibition to nesting on suitablelooking upland fields enclosed by walls or fences, though occasional nests are reported from such places, e.g. near High Force, Upper Teesdale (A. W. Colling, pers. comm.). Breeding was at one time regular on unenclosed ground close to hill crofts (D. and C. Nethersole-Thompson 1961(a)), but fields are normally used only for feeding. The nesting habitats of the Golden Plover may be grouped into the following main and intergrading classes of vegetation: (a) Submontane (b) Montane (i) Heather moor (i) Dwarf shrub heath (ii) Wet heath (ii) Moss and lichen heath (iii) Blanket bog (iii) Blanket bog (iv) Acidic grassland (iv) Acidic grassland (v) Calcareous grassland Table II gives details of nest site selection in relation to vegetation type. These do not show a true reflection of country-wide preferences, for the sample is far from random in terms of total distribution and numbers in Britain. The preponderance of nests referring to the three main study areas (cols. 1, 2 and 6) heavily biases the overall picture for selection of site according to vegetation. Even within these three areas, the samples of nests cannot be related to accurate measures of the total availability of different plant communities, as there are no vegetation maps detailed enough to give this information. For the other areas, the nests found represent a more random sample, but the numbers are mostly too small to show if any marked preferences exist for particular vegetation types. It is clear, however, that certain types of vegetation are avoided, and that within the broad classes which are favoured, a wide range of plant communities is acceptable. Provided that the general character of the ground is suitable, Golden Plovers do not show any notable selectivity for particular plant communities, and frequency of nesting in any type is largely a reflection of its relative extent in the breeding area. Certainly there are no preferences sufficiently obvious to be an aid to nest-finding. It would hardly be advantageous to the species if there were, for predators would quickly learn any such tendencies. The converse seems to be the case: tilt Golden Plovers avoid any conspicuous features in placing their nests, and often choose a particularly nondescript piece of ground or mixed plant community where the lack of pattern or uniformity in the vegetation renders the eggs even less easy to see. The only possible exception seems to be that nests are occasionally placed on old mine-spoil on the moor, where this is vegetated or partially so. D. Nethersole-Thompson (in lilt.) has found that in the Cairngorm foothills Golden Plovers sometimes use Greenshank-type nest sites, against fallen pieces of tree trunk or branches, and once actually used an old Greenshank nestbowl thus situated.

12 BREEDING OF THE GOLDEN PLOVER 73 About 10% of the nests were situated in communities which were mixtures in terms of the types listed. Dry heather moor is under-represented as nesting habitat in Table II, simply because it is not extensive in the areas examined. Probably the most favoured habitat in Britain nowadays is damp, peaty ground with Eriophorum vaginatum abundant to dominant. Yalden (1974) has noted a strong association with this plant in the Peak District, and the same is certainly true of other parts of the Pennines, Cheviots and eastern Southern Uplands. Calluna vulgaris is the other most characteristic plant of the nesting places, though it is often mixed with other species. In some areas, numbers have declined on dry heather moor with dominant Calluna, though this is still a favourite habitat in the eastern Pennines, the North York Moors and the eastern Highlands. The BTO Nest Record Cards for the species include 192 with description of the nest site vegetation during the period While still not a true random sample, these records probably give a more representative picture of habitat selection than my own, though they are mostly weak in the details of botanical description. Of the 192, 47 described as being in `heather' and 20 others in 'burnt heather' probably included a proportion in vegetation where Calluna was mixed with other moorland species, since ling heather is by far the most familiar moorland plant to most people. Only ten nests were recorded in cotton-grass (two with heather also), so that the majority of 45 nest sites described as 'grass' were likely to have been in Eriophorum vaginatum. Similarly, 25 further nests in 'heather and grass' were probably mostly in Calluna-Eriophorum mixtures. All but one of 19 nests placed in short limestone turf were on the Mallerstang area (Table II, col. 1), so that this habitat is probably over-represented in the list. Heather deer sedge (6), montane heaths and grasslands (6), Sphagnum (5), dead bracken (4), mat grass (2), old mine waste (2) and heath rush lichen (1) accounted for the remaining records. The scrape, about cm in diameter and 3-6 cm deep is placed where it will not collect water during wet weather, and so is often on a tussock or hummock. On hummocky limestone turf most nests are placed on the crests, as are those of Lapwings; sheep walk in the intervening hollows, but these are also damper and give less good visibility than the hummocks. On stony ground a good many nests are placed against or between the smaller bedded blocks or slabs, and in the Pennines some are amongst the more fragmentary types of limestone pavement. On the high tops some sites are in exactly the kind of places chosen by Dotterels, on bare, windswept ground, often amongst stones, or on fringe moss hummocks. Scrapes containing eggs are invariably lined, typically with the broken-up dead leaves of the most abundant surrounding monocotyledonous plants. Out of 179 nests placed in Eriophorum vaginatum communities, 172 contained the leaves of this plant, and 64 had no other lining material of note; E. angustifolium leaves also frequently appear as lining material. Nardus leaves are used as lining in most nests placed in vegetation with abundance of this species. If Calluna is beside the nest, there are usually dead shoots in the lining, but these could fall into the scrape, in addition to being placed there. On recently burned heather ground, the scrape usually contains small Calluna twigs. Scrapes on limestone turf often have small tufts of the basal parts of Festuca ovina, but seldom forming the dense flat pad characteristic of Lapwing nests. The other favourite lining material

13 74 BIRD STUDY TABLE II. NEST SITES OF GOLDEN PLOVERS ACCORDING TO PLANT COMMUNITY BREEDING AREA a.) t a. L. t.) a a...f. Vegetation Plant community dominants, with distinctive facies a z,,.., I'l z &- _ 0, type at L.-- a '' --..4) 4))..1 a E 4), a,.> ) z "Eii E..!:2 13 a, -F.. E Zi -., (Z) 0 a, I* 2* 3 4 { Calluna vulgaris + Eriophorum vaginatum (a) species-poor (b) moss-rich 2 (c) lichen-rich 17 3 Blanket bogs (d) Sphagnum-rich Eriophorum vaginatum 13 1 (a) dense, sub-montane (b) sparse, montane 3 (c) + Empetrum nigrum 5 (d) Sphagnum-rich 2 1 Eriophorum angustifolium 1 2 Wet Calluna + Scirpus caespitosus 1 Heath Calluna + Molinia caerulea { Calluna + Nardus stricta 1 1 Calluna vulgaris Heather (a) recently burned, 1 Moor mainly bare (b) short, but closed Nardus stricta 12 2 Acidic Nardus + Juncus squarrosus 2 3 Juncus squarrosus grassland 1 1 Festuca ovina + Calcare- Agrostis spp. (poor) 4 2 ous grassland Festuca ovina + Vaccinium Festuca ovina + rubra (rich) spp. 2 Festuca + Rhacomitrium lanuginosum 1 Montane grassland and heaths Deschampsia flexuosa Calluna + Rhacomitrium lanuginosum Vaccimum spp + Empetrum spp. +Nardus Carex bigelowii Rhacomitrium lanuginosum Total number of nests Narns : Data are for 302 nests examined by D. A. Ratcliffe in the Peak District, northern pennines, Lakeland, Cheviots, Southern Uplands and Scottish Highlands during One nest was on an old lead mine spoil-heap.

14 NOTES TO TABLE Ii: Details of breeding areas : Those marked * are the main study areas. BREEDING OF THE GOLDEN PLOVER Mallerstang limestone. Northern Pennines at m, 120 ha, mainly of closegrazed Festuca-Agrostis sward rich in forbs and lichens, thrown into regular hummocks, on which the majority nest. Some nests are on rocky ground where the turf is broken by fragmentary limestone pavement. Grades into acidic Festuca- Agrostis turf and Nardus-luncus squarrosus grassland. 2. Alston Moor. Northern Pennines at m. 1,100 ha, mainly of blanket bog varying from Sphagnum-dominated flowe with pools to areas of sheet erosion. Limited areas of Nardus and Juncus squarrosus communities, and Festuca-Agrostis limestone sward. 3. Hallam and Derwent Moors, Peak District, at m. The eastern moorlands of the High Peak, with predominance of dry heather moor on lower ground and blanket bog on the higher plateaux. Much of the blanket bog has lost Calluna and is now dominated by Eriophorum vaginatum, with local abundance of Empetrum nigrum and Vaccinium myrtillus. Areas severely affected by sheet erosion show some recolonisation by Eriophorum angustifolium. 4. Other Northern Pennines, at m. These extend from Great Shunner Fell in the south to Tindale Fells in the north. Blanket bogs of the Alston Moor type are extensive and form the most favoured habitat, but there is a wide range of both acidic and basic grasslands. Montane habitats are represented on the Crossfell range and on the fells around the head of Swaledale. 5. Lakeland. Scattered breeding haunts, ranging from dry heather moor at 305 m above Caldbeck to blanket bog and damp acidic grassland at over 610 m in the Skiddaw group. 6. Moorfoot Hills. Eastern Southern Uplands at m. 1,000 ha, mainly extensive blanket bog on ridges and plateaux, varying from Sphagnum flowe to a more widespread drier facies, but with relatively little erosion. Beyond the study area, the lower and drier margins of the massif have extensive dry Calluna moor and acidic grasslands. 7. Other Southern Uplands and Borders, at m. Ranges from the transitional raised blanket bogs of north Cumberland and Wigtownshire to dry heather moor and upland blanket bog in the Langholm Hills and Cheviots. Montane habitats are represented on the high tops of Tweedsmuir. 8. Sutherland and Harris. These two districts represent the northwest Highlands and Islands, at m. The nesting habitats are chiefly the dry to wet heaths of morainic foothill country, and the short Calluna heaths of oceanic type which, at 450 m in the far northwest, belong to the montane zone. 9. Eastern Highlands. True montane heaths at 825-1,000 m on the Caenlochan hills, the Monadhliath and Ben Wyvis. is lichen, and if this is available in the immediate vicinity, the nest will usually contain at least a small amount. Of 273 nests in which the lining material was recorded, 175 contained at least some lichen; usually there was vascular plant material as well, but in 34 nests only lichen was present. The 'reindeer moss' Cladonia lichens are much used, and sometimes form a thick pad under the eggs. In 107 Cladonia uncialis was present, and in 100 either C. arbuscula or the closely related C. impexa. As in Dotterel nests, C. uncialis is often an important component of the lining, presumably because its relatively large, tubular thalli form an excellent insulating material. The crustaceous Cladonia lichens are much less used. Scraps of moss are often present, but these plants do not figure significantly in nest linings.

15 76 BIRD STUDY Many Golden Plovers build up the layer of lining material as incubation proceeds, but individuals vary, and a few nests contain little lining even at hatching. Sauer (1962), in his studies of the Asiatic Golden Plover on St Lawrence Island in the Bering Sea, emphasised the relatively elaborate structure of the nest and found that the lichen Thamnolia vermicularis was the preferred lining material. This species, which is montane and somewhat sparse in occurrence in Britain, also has little-branched tubular thalli with good insulating properties. NEST-FINDING Although this is a widespread and locally plentiful species on the northern uplands, its nest has always been regarded as among the more difficult to find. Observations on actual nests are nevertheless basic to a study of breeding biology, and only the finding of all nests within selected areas can give an exact measure of breeding population size or density. Attention has therefore been given to nestfinding techniques, with sufficient success to merit some description here. If possible, it is best to gain an impression of numbers and density by walking over the nesting grounds just before laying time, when the birds are dispersed in pairs on their territories; but even then they may come and go according to weather or even time of day, and an accurate count is sometimes difficult. When there are eggs, Golden Plovers are often relatively silent and undemonstrative, even after leaving the nest, and only one of a pair will usually be present at any one time. When the majority of pairs have eggs one may thus under-estimate the population by a substantial margin. After the young have hatched, behaviour changes markedly and the parents are then typically noisy and conspicuous, and usually both will be present; they tend to come to meet and then follow human intruders, and where they are numerous, each pair becomes mixed up with its neighbours. Sometimes a little party representing three or four pairs will form up and fly around calling, and under these circumstances it is easy to overestimate numbers. Yet again, on moors where predation on eggs is heavy, different parts of the population will be out of phase in their nesting and this can be confusing. Gilbert and Brook (1924) make the intriguing comment that despite the difficulties of nest-finding, the species has a 'chink in its armour' which can be exploited to simplify the process. Apparently this referred to the bird's alleged Plate 1 Golden Plover Pluvialis apricaria of 'northern' plumage type. Photograph by courtesy of Eric Hosking, F.R.P.S. Plate 2 Typical nesting habitat: heather moor, acidic grassland and blanket bog in Swaledale, Pennines, at 610 m (2,000 ft). Plate 3 Unusual nesting habitat on short, hummocky limestone grassland in Mallerstang valley, Pennines, at 534 m (1,750 ft). Photographs by D. A. Ratcliffe. Plate 4 Golden Plover at the nest. Photograph by courtesy of Eric Hosking, F.R.P.S. Plate 5 Golden Plover approaching the nest. Plate 6 Golden Plover accompanying chicks. Photographs by courtesy of Eric Hosking, F.R.P.S.

16

17

18

19

20 BREEDING OF THE GOLDEN PLOVER 77 inability to count, so that if a party of three or four people hid one of their number and went away, the hidden observer could fairly readily watch the bird back to its nest (D. Nethersole-Thompson, pers. comm.). I have not tried this technique, but have found a small number of nests simply as a lone and relatively unconcealed watcher lying at some distance. This is not, in my experience, a good technique. Few Golden Plovers will return to their nests if a single watcher is closer than 400 m. With a telescope or good binoculars it is sometimes fairly easy to watch a bird to its nest at this distance, and under favourable conditions I have done it successfully at 700 m. But so often conditions are not favourable, and they are frequently impossible. Presence of dead ground or convex slopes may make watching-back out of the question; the light is often poor, whilst the combination of wind and cold may sorely test the watcher's patience, and also make it unjustifiable to keep the bird off its eggs. Only under favourable circumstances, or when all else has failed, is this method recommended. R. Parr (pers. comm.) has found watching from a Landrover successful, though this is limited to places approachable by vehicle. Much depends on the nature of the terrain, and on whether the nesting plovers tend to be 'sitters' or 'fliers'. For there is a curious difference that whilst the majority of Golden Plovers fly direct from their nests as a human being comes within sight, in certain areas or under certain conditions or at certain times, nearly all the birds sit close and flush only if the intruder chances to walk within about 3-10 m of the nest. In areas where the birds are predominantly 'fliers' I have found that careful watching whilst walking is by far the best method of finding nests. There is a special technique which has to be diligently cultivated; the impulse to look at the ground at one's feet has to be resisted, and the gaze concentrated as far ahead as possible, not in one place, but scanning continuously over a wide arc from one side to the other and back. This has to be kept up whilst walking over the moor, and it is astonishing how often a bird chooses to fly at the moment the gaze falters or the mind wanders. Ideally, one tries to pick up the first flick of movement as the bird raises its wings and exposes its light underparts, for once airborne, it is extremely difficult to see against the background of moor. A few run from their eggs, but flight is much more usual. A plover just off the nest has a distinctive flight, skimming low over the ground in an unhurried way as though trying to appear casual. A bird which brushes the vegetation or flutters across it with half-trailing wings in distraction display is virtually certain to be just off a nest or chicks. Usually the flight is to one side or the other, i.e. more or less at right angles to the intruder's approach, but some birds fly away along the line of approach, and a few (usually when the nest is a long way off) come towards their disturber. A flushed bird is nearly always silent until it has pitched, usually after flying at least m when it begins to give the alarm call, but occasionally one calls in its first flight if flushed at close hand. Calls usually mean that the nest has been left, but sometimes, if the nonsitter of the pair is present and calling, its mate will continue to sit for some time before moving and may even allow a very close approach. Individuals are very variable in these respects, but most 'fliers' usually leave when they first see an intruder has come within 50 m before being seen. Some individuals are cunning distance may sit very tight the next time the intruder approaches, and the same bird may behave differently on different days, or according to weather. The

21 78 BIRD STUDY different birds of the pair may also vary in their behaviour. On uneven ground it is often possible to approach within 100 m without being seen, but even a bird normally a 'flier' will sometimes continue to sit if it is taken by surprise and the intruder has come within 50 m before being seen. Some individuals are cunning and will let a man walk past and then fly away behind when his back is turned; and if there is cover, such as a peat gully, this will be used by the departing bird. Calling also varies greatly; some birds do not call for 5-15 minutes after they have left the nest, and a few remain silent altogether. Paradoxically, on the Pennine area with a particularly dense Golden Plover population, it is very difficult to spot the birds flying from their nests, for each pair takes warning from the calls of its neighbours, and there is also a large population of Lapwings which raise the alarm, even before one has reached the nesting ground. The pale grassland of the area is also a more difficult background against which to see a flying plover than the darker colour of typical grouse moor. Nearly all the plovers here slip off their nests long before they can be seen, and cold (or lukewarm!) searching is almost the only method of finding them. However, on areas with a dense nesting population, most of the nests are spaced out with a high degree of regularity, and once the average spacing is known, and a few nests are found, it becomes easier to find more nests, or even the whole number, because the areas of 'no man's land' become apparent and the search is progressively narrowed (p. 93 seq. and Figure 3). On most moors, however, cold searching is a waste of time and energy if a Golden Plover is off its nest and calling. Except on the Pennine grassland where nests are unusually close together, I have found exceedingly few nests by searching in the vicinity of a calling bird which had left its nest before being detected. Moreover, birds which have left their nests often remain silent and, if not detected, their next flight may mislead the observer into thinking that he is seeing the bird fly away from the nest; fruitless searching then follows before the truth dawns. A calling plover will inevitably be a fair distance from its nest and it is far better to go away and let the bird return to the eggs, and then come back for another attempt. Sometimes it pays to try a different approach route; but some pairs or birds are very difficult subjects, and several attempts and approaches may have to be made before the nest is found. Birds which have lost their eggs, have not laid, or have incomplete clutches can be especially puzzling; they may be sitting around calling as the intruder approaches, or they may fly silently and low in the manner of birds leaving nests. Birds with incomplete clutches may, however, stand by or actually cover their eggs, and fly in the usual way, though I have also known a bird covering an incomplete clutch to sit until I was a few metres away. Birds which come to an area to feed, especially in the evening, can call a good deal and give the wrong impression that they have a nest. When Golden Plovers are predominantly 'sitters', the problems of nest-finding can be considerable. The odd nest can be found by cold searching but no more than by random walking over the moor. Once in a while one comes across an uncovered nest of eggs during random walking, but for every nest found this way a large number of miles has to be covered. I have only four times found nests by spotting tight-sitting plovers whilst walking, for the incubating bird is extremely well camouflaged, the gold-spangled plumage of the back matching exactly the appearance of a moist hummock of woolly fringe moss Rhacornitrium

22 BREEDING OF THE GOLDEN PLOVER 79 lanuginosum or other mosses which form a characteristic part of the moorland vegetation. For 'sitters', the rope-towing technique (sweeping the ground with a rope dragged between two or more people) can sometimes be effective, but it is a tiring business with a rope of reasonable length and very tedious on rough or uneven ground. I have no experience of nest-finding with dogs, but a trained dog is probably the best way of finding 'sitter' nests, and an animal which can 'point' a nest whether the bird is on or off is best of all. Yet even when the plovers tend to be 'sitters', there may be other solutions which observation will reveal. On Alston Moor, the ground appears to be almost devoid of birds during the late morning and afternoon, and a walk over the area at this time would give a totally false impression of numbers. But towards evening, around hours (BST), there will be bursts of activity, especially of territorial and greeting song, apparently as the non-sitters return from feeding grounds elsewhere. This may be a change-over time for many pairs. Whatever the case, a walk over the moors after this time will usually give a different response, for there are now 'fliers' which can be worked in the usual way. Often too, the pair will be present in a territory and the non-sitter then gives the alarm, which also tends to make the sitter fly. Nest-hunting in such areas is obviously best left until the evening. Yet, to emphasise the risks in sweeping statements, I have known the usual situation in this locality to be completely reversed, with birds all over the place during the day, and silence in the evening. And conversely, on the Moorfoot Hills, where the Golden Plovers are mainly 'fliers', on occasional days the majority will sit closely. Much depends on weather conditions. Bright sunlight is best for spotting `fliers', especially when the angle is low, for the light flash of the underparts can be more readily picked up. 'Back-lighting', i.e. walking towards the sun, is often best of all for revealing rising or flying birds. Mist is hopeless because the bird sees the human intruder first and flies before it can be seen, and rain also gives it an advantage. Very strong wind, especially with rain, can cause many birds to sit tightly and become more vulnerable to the rope technique, but such conditions are unpleasant to work in, and there is a greater risk to eggs exposed during searches. It is seldom possible to pick up other clues during a search. The off-duty bird often has a conspicuous hummock where it stands, and leaves feathers and droppings, but this is invariably well away from the nest. Scrapes are commonly made all over the territory, though the number made by any pair varies, and the finding of these is no help at all. Golden Plovers have a way of avoiding any piece of ground which draws attention, such as a small, isolated patch of recently burned ground or unusually tall hummocks (see p. 72). On the other hand, there are certain types of vegetation or habitat which it is safe to assume will not contain a nest. Areas of dense heather more than 25 cm high, tall tussocky growths of cottongrass, dense stands of bilberry and heath rush, and excessively wet situations such as flush bogs and spongy Sphagnum channels or hollows are avoided. On ground where much of the vegetation is tall and dense, it is always best to search first where the growth is short and sparse. It is instructive to re-examine known nests after a period of prolonged and heavy rain; much of the nesting habitat which previously seemed relatively dry will be found to have been saturated with running or standing water, but the nests will still be dry and virtually unaffected.

23 80 BIRD STUDY Areas where predation is heavy are difficult to work, for in the early part of the season some pairs will have lost their eggs, and even if some first 'repeats' escape, there will be a good chance that losses will again occur, to be followed by second, or even third 'repeats'. The breeding of the population will then become staggered, with only a certain proportion having eggs at any one time. Some pairs may disappear, or come and go, from the breeding area until they have re-laid. It is then extremely difficult to make an accurate census, unless the individual pairs can be followed, and this can be done only by several counts spread over a period of at least two months. In any area where nests have been found in a previous year, it is always worthwhile examining the old scrapes, because these persist, and are not uncommonly used in subsequent years. The majority of nests are placed on relatively featureless ground and accurate bearings taken with a prismatic compass are by far the best method of relocating them. The chicks of the Golden Plover are usually more difficult to find than the eggs, for they leave the nest at a day or so after hatching, and the parents seldom give any help in locating them. If, however, the nesting grounds are accessible by motor vehicle, it is usually quite easy to find some chicks, for they do not respond to the vehicle by crouching, but continue to run or walk about, despite the parental alarms. They hide only when a human emerges from the car. By choosing suitably placed roads and tracks it would thus be possible to ring young more readily than by walking over the moors, unless nests were already known and could be visited exactly at hatching. LAYING, HATCHING AND CLUTCH-SIZE During a mild winter, Golden Plovers reappear on many nesting grounds in February. A cold spell, especially with snow, sends them down to the feeding places on the upper farm pastures, and there may be several comings and goings between the high and the low grounds before nesting begins. The onset of nesting varies by up to two weeks or more between years according to the severity of the weather, but in a typical year laying begins in the last days of March. The majority lay their first eggs between 10 April and 7 May, on breeding grounds up

24 BREEDING OF THE GOLDEN PLOVER 81 to about 610 m. There appears to be a delay in breeding with increasing altitude, though data are inadequate to show the exact relationship Pairs nesting in the bleak montane environment of the high tops at 900 m or more are up to two or three weeks later than those of the lower breeding grounds (D. and C. Nethersole- Thompson 1961(a)). Table III provides data on the time spread of breeding. In contrast to Golden Plovers in high latitudes, British birds typically show a long spread of nesting, over the four months April-July, associated with their wide altitudinal range, and more especially the high failure rate of early nests in many areas and the laying of repeat clutches. In some years, severe weather after the onset of laying is a major cause of early nest failure. The snow-storms of April 1966 resulted in many nests being snowed-over and abandoned, and most pairs had fresh clutches by May after conditions had improved. In 1975, the blizzards and icy gales on 1-3 June, with a fall of 14 cm of snow, caught the Golden Plovers at Moor House in the northern Pennines at the critical hatching period, and J. Parkin (pers. comm.) noted later small parties of adults together, most of the nests or chicks presumably having been lost. Failure at this stage probably spells an end to nesting for the season, and when a large proportion of the population is affected may mean a disastrous breeding success for certain areas. Predation is an important cause of egg-loss in many areas, but although a more consistent and continuous adverse factor than bad weather, it is seldom as catastrophic at any one time. Loss of clutches is usually followed by the laying of repeats, unless the eggs were far advanced in incubation; and pairs will produce at least two, and sometimes three, replacement clutches if there are further losses (D. and C. Nethersole- Thompson 1961 (a)). These writers found that repeat clutches took days to produce, but in the Pennines one pair took 24 days (D. A. Ratcliffe). In areas where nest predation is heavy, pairs with fresh eggs will thus be found right through May and well into June, and a few will still be incubating eggs up to mid- July. I have not seen eggs in August, but occasional birds are to be found on the moors then, behaving as though they still have eggs or small young. There is no evidence whatever of double broods. The eggs are normally laid at intervals of 48 hours, but an interval of 60 hours is frequent, and sometimes the gap is several days, especially between the third and fourth eggs. Clutches thus take at least a week to complete. The incubation period is quite long, from 27.5 to 34 days, with a mean of 30.7 days (D. and C. Nethersole-Thompson 1961(a)). These authors found a long chipping period of usually hours, whilst Williamson (1948) found 100 hours to be normal in the Faeroes. The fledging period is around five weeks, so that when first nests are successful, flying young begin to appear from about 10 June. Golden Plovers do not linger unnecessarily on the moors; successful breeders and their young may begin to flock even before the end of June, and early July sees the first parties of resident birds back in the lowlands, already in moult. The bulk have left by 1 August, though small parties may be seen on many moors at intervals right through the winter, at least during milder weather. A typical nidifugous species, the Golden Plover is one of the British waders which consistently has a clutch of four eggs, and there is no evidence of significant departure from this norm anywhere within this country. Out of 231 full clutches

25 82 BIRD STUDY 00 -, Ks) 2"?..?.' 4; 24 0 at 0 crl.r.,:g cq = > ;:74 a.).0 Q ;. 0 4,,?, 58HuV V C) -. el cl 0 w -.,..-, g -.50 ( en et 0 "0 CV V,) 0,,., a.) 0 i7) ea ' rn o c.) ).. en.1-0 a.) cel cn.-. ;5, rtr-0 Ter1 ci t) b0e. - t"- -, , 0 ul 1-1 c%".5.:t CI NZI v., cn 0 cd o cz. o, Er.,;,.,, 0 E en '- 0 4 cl ZI.4 0 tel t--. oo ''-'.0 t...1- VI en c4.-. X-. of,,.0 0 m..4-,-,, 0 g 0,../,-, W,-, t.),....,:i CO cn C..).". 0 " _ v:, > Q.-. a) rn.. FjP 0.2 0,..0 t) t) c),..., 0. >.0 t - v-1 c4 t) --a _ acoc'd,...8,-. CO.-.2 K o 0 0 CO rn3.5"-"m rn.4- ON ,.0 CO. en 12.-0,5 w c.; el - el CD 1.) u, '5 a) -' 0.0 cn " a) 0...,.4, C ',..1 c.) '-',b ,.1- ;..,..., 0.-. CU rn 4-1, el 3 u 8 ' 8 0.'4, v.) 1--. oo N 0 0 i'.!.4 '1')bi5-0 0 a) t;. " , > r, -.-A 6. o - m...,.... cl.,_. u en.1- en el lel E '-' `g a) 2, el -,,--.,--, co) E CO - v) a..c) 4'a -s.'-'..-,f1..,r 1..." Lel,r1.1..fD I:12, r3 0 el CO,,,, u E134 5 ""ii , o,., cl -. ''',_, c) OJ 44 ON , - en 4, 0 S ;-1,-1 cn to 0 0 1, a-, g X, 00 0 cn at 0 Ts' 4' o 0 0 S,L), o 0 a-..a...,, el I en... >- cn =1., cu0 OD ' -... el -0.,:a 00E 4 o 0 - I.....o *?;1-( 0 4.). l V::, 40 w 'I el 0 pq cd O o o -zr.--, o q-, Ea, c,,,,,..) r) 0., `-.) c') bp.., 0.1,.. N tn 0, a) , - N --' 2.?,"'IM.OE - 5 4, c1-05u vl.. vo 0.,... -,c, 0.2 -o sa, 0 4 _...4 E44 "' 'l c 0.9_, cn... c.) t.... S '0, cd (L) 4-, 1_,.-t.-4 ern 0 ' '.72,'),.g 4.4; g (a co C.) a).--. n, i- ---E-1 -o 0 f.,:"o" 0 a al.,_ N " 0,-.:. '-) "" ir. 0.'" 0 o t-t "..., N I en..._,-, el '. > (1) 0 '.0,...1 CI, C 0 '-' _, v.,c, 6,, d.3 (4 "C) 0. O 0 1_, o 0 In, 0 eal A-, cd 0 ".2..., 2 WI,4 cu en Q C) 0 1.) COQ.-... C),.., 0 1;) '14,.), ca , 0 O,,., 0.),,,.., 8 3 g ' 1 tm --. el en - el en 0 0.c.,2, v.2.. 'Fel.. E:0 ',5, 0 > II.., I--, I. 0 7.). u , I-, > el 8.,0.,,,...5., > '0 eel a.) ONIAV-1 DNIIIDIVH Z <1=4 '7).5 tf).0 c 4 A 0

26 BREEDING OF THE GOLDEN PLOVER 83 seen by the writer, 207 were C/4, 15 were C/3 and 9 were C/2, giving a mean of 3.86 eggs per clutch. Clutches of less than C/4 were only included in this total if they showed signs of incubation. It is, however, evident that many clutches of less than four eggs have been depleted; three additional C/2 and one C/1 had remains of other eggs in the nest or close to it. Moreover, two of the 207 full clutches had single eggs lying outside the scrape and thus effectively lost. When these six known depletions are included, mean clutch size drops to 3.81 eggs. Predation of the first egg or two is probably the most usual cause of clutch depletion, for many birds then continue to lay the rest of the clutch, but no more. During the hard weather so often prevailing on the moors at laying time, Golden Plovers sometimes drop eggs prematurely; these are typically thin-shelled, pale blue and finely spotted instead of blotched. This would probably be another factor leading to small clutches. Eventual clutch size is thus slightly below the four eggs which the great majority of females probably lay. Hatching success in clutches which are incubated to full term is probably subject to only slight further reduction. As in most waders, the incidence of addled eggs is small. Cracked and dehydrated eggs are occasional, and probably usually represent first eggs which were unprotected against night frost. Only six unhatched eggs were found out of 135 in 35 clutches known to have hatched (4.4%). When this is applied to the figure of 3.81 eggs per clutch, it indicates that the mean brood size at hatching is likely to be around 3.64 chicks. An insufficient number of newly hatched broods has been seen to establish the figure directly, and 3.64 is probably slightly high; the mean clutch size on which it is based includes many fresh clutches, and the chances of clutch depletion are higher in those incubated to full term. Analysis of the BTO Nest Record Cards gave a mean clutch size of 3.82 eggs for clutches at various stages of incubation (n = 163), and a mean brood size at hatching to three days old of 3.45 chicks (n = 46). Only four unhatched eggs were found out of 105 in 29 clutches known to have hatched (3.8%), but in two more clutches (both C/4) incubated to full term all the eggs proved infertile. The BTO records also showed that in 48 nests with full clutches which were revisited after a gap of four or more days, nine had lost one egg and three had lost two eggs; 25% thus showed clutch depletion. Of 50 nests followed up to determine success, 31 hatched at least one egg (62%), and 19 failed completely (38%). Of the failures, 14 were predated, two were deserted, two were completely infertile, and one was trodden in, evidently by a sheep. Although the method of analysis may give an upwards bias to success, this figure of 62% hatching success for nests may be valid as a general indication of hatching performance in British Golden Plovers. At the moment it is the only figure available, and would lead to an expectation of an average of about 2.2 chicks hatched per breeding pair. There is even less information about the survival of chicks between hatching and fledging, for this is extremely difficult to obtain. The chicks leave the nest at an early age and usually separate when danger threatens. They are also often divided between the two parents. Their cryptic coloration makes them exceedingly elusive much more so usually than nests with eggs and they can hide unless the vegetation is extremely short. Even if chicks are found, counting the brood with certainty is often difficult. There are therefore no data on the average breeding success, in terms of output of fledged young per breeding pair, for any district.

27 84 BIRD STUDY FOOD AND FEEDING GROUNDS The food of the Golden Plover appears to be largely invertebrate. There is little direct information on food during the breeding season, but two detailed studies have given data for other times of the year. Ernest Blezard (unpubl.) examined the stomachs of birds collected mainly during autumn and winter in their lowland haunts, and found as follows: (a) 14 birds (7 cp 79) from grass fields at Blackwell, Carlisle, Cumberland, October February, 1935 and Food item No. of individuals taken No. of birds containing item Earthworms Never more than traces 13 Beetles Carabidae Amara plebeja 2 2 Amara sp. 2 1 Clivina fossor 1 1 Bembidion sp. 1 1 Staphylinidae Philonthus varius 1 1 Philonthus sp. 1 1 Unidentified adult 1 1 Unidentified larva 1 1 Histeridae Hister sp. 1 1 Elateridae Agriotes sp. 4 3 Dryopidae Dryops sp. 5 1 Scarabaeidae Aphodius fimetarius 2 2 A. punctatosulcatus 5 2 A. prodromus 1 1 Curculionidae Phytonomus austriacus 37 9 Sitona lepidus 9 3 Moth: unidentified larva 1 1 Diptera: unidentified larva 3 2 Slug: Agriolimax sp. 5 3 Univalve molluscs Vitrea sp. 2 2 Cochlicopa lubrica 20 5 Grass leaves and roots 14 Grit fragments 13 (b) 2 birds from Solway shore: cr juvenile, Brow Scar, Moricambe, Cumberland , very small shore crab Carcinus maenas; small bivalve molluscs Tellina sp. in fragments; 3 small univalve molluscs-2 Hybrobia sp., 1 Littorina sp.; minute beetle remains; sand. cr Redkirk, Dumfriesshire, , weevil, Sitona sp.; fragments of common mussel Mytilus edulis; grit; grass leaves. (c) 2 birds from breeding habitat, Cronkley Fell, Upper Teesdale, Yorkshire: a , 2 larvae of Broom Moth Ceramica pisi; Carabid beetle Feronia sp.; grit; rootlets and fibrous plant shreds. a , Elaterid beetle Agriotes sp.; weevil Otiorhynchus sp.; grit and mass of grass with leaves of Thymus drucei. All beetle remains were named or confirmed by the late F. H. Day, senr. The stomachs of 16 out of 17 birds contained vegetable remains, usually tightly packed wads of grass leaves and roots. The material could be ingested incidentally as the birds pick up their invertebrate prey, but some at least might be taken deliberately; in either case it could well have nutritional value. Burton (1974) has published detailed analyses of the stomach contents for 28 birds feeding mainly on farmland year Bradwell-on-Sea, Essex. In summary, these correspond quite closely to the data just presented, in showing preponderance of earthworms, beetles and Diptera, with a lower frequency of Lepidoptera, spiders and molluscs. The majority of the 28 stomachs also contained grass and small stones or shell fragments.

28 BREEDING OF THE GOLDEN PLOVER 85 These observations suggest strongly that the food of the Golden Plover during the breeding season consists mainly of a wide spectrum of the common invertebrates of upland heaths, grasslands and bogs which can be taken from the ground surface or the litter layer. This bird has a short bill and is unlikely to be able to probe more than one or two centimetres below the soil surface. Feeding birds certainly appear to pick up their food from the surface, and Burton (1974) has provided details of feeding behaviour. The actual feeding grounds show considerable diversity. Although the majority of Golden Plovers breed on barren uplands, where the soils are deficient in major nutrients, an association with richer terrain for feeding is noticeable during the breeding season in many areas. Gilbert and Brook (1924) pointed out that Golden Plovers feed but do not breed on the improved pastures which persist around abandoned shielings on the otherwise infertile moors where the nests are placed. Moreover, many nesting grounds are within easy flight (5-8 km) of the enclosed and improved grasslands of the upper farms which adjoin grouse moors and sheep walks. These permanent pastures have usually been subjected to heavy grazing and manuring over a long period, and many have received applications of artificial fertilisers. In consequence, their soils typically have much higher levels of nitrogen, phosphorus and potassium, as well as calcium, than those of the unenclosed moors and hill pastures beyond. Their invertebrate populations, especially of earthworms, are much higher than in the infertile podsolic soils and blanket peats of the typical nesting grounds. Golden Plovers commonly resort to these richer pastures to feed, and, until the young have hatched, evidently do so in preference to working the poorer ground around the nest. As in wintering areas, certain fields and 'intakes' are preferred, and will be used year after year. Birds are to be seen there regularly throughout the breeding season, and whilst some may be non-breeders, others come and go, and are evidently off-duty nesting birds from the moors above. D. Nethersole- Thompson (in lilt.) has seen brooding birds visit enclosed farm pastures in Speyside, and R. Parr (pers. comm ) has found regular flights from nests on a moorland fringe to adjoining grass fields at Kerloch Moor in Kincardineshire. In east Sutherland, D. A. Goode (pers. comm.) found a regular flight over 2-5 km between nesting places on the Knockfin Heights and farm pastures at Achentoul, Kinbrace. D. Humphrey (pers. comm.) also noticed fighting in and out of grass fields along the edge of the moorland, especially during late afternoon and evening, at Forsinain and Melvich, Strath Halladale, east Sutherland. Flocks of c. 100 and birds were involved. The variable pattern of incubation sharing between the sexes (D. and C. Nethersole-Thompson 1961(a)) may be related to the distances which the off-duty birds have to fly to feed. Golden Plovers breeding in high montane terrain may depend on uncultivated ground during the nesting season, as do their northern tundra relatives, but evidently have preferred feeding places, and show some tendency to feed in the flush bog and spring complexes which are well developed where seepage water emerges at the ground surface, especially in association with late snow patches. They seldom nest on such ground, and may fly some distance from their nests to reach these places. Sauer (1962), in his study of the Asiastic Golden Plover, also found that pairs had preferred feeding areas outside the breeding territory. The species' complete territorial system included an inner territory with the nest,

29 86 BIRD STUDY surrounding feeding areas, and 'extra-territorial' feeding grounds; and he suggested that this complex structure has developed as the result of food shortage within the breeding territory. The availability of food in the vicinity of the nest is likely to be more limited in both variety and quantity on many British breeding areas than on the preferred feeding grounds described, though subject to marked differences according to the type of habitat. It is therefore not surprising that Golden Plovers favour, and actually nest on, unenclosed rich grasslands on the hills, especially where these occur in substantial areas as they do on the Pennine limestone formations. I am much indebted to Dr J. C. Coulson, who has also noted a marked preference of Golden Plovers for base-rich grassland, for detailed information on the differences in invertebrate populations of different soil/vegetation types, based on his work and that of colleagues at Moor House in the northern Pennines (Coulson 1959 and 1962; Coulson and Whittaker, in press; Houston 1971). These studies show that the biomass, averaged over the year, of soil invertebrates is greater on limestone and alluvial grassland with their base-rich brown loam soils than on blanket bog with highly acidic and base-poor peat. Up to the time when Golden Plovers lay, there is only low availability of invertebrate food on both limestone grassland and blanket bog, with little difference between these habitats in their above-ground invertebrates. However, there is marked difference in both number of species and abundance of beetles and earthworms, in favour of the limestone grasslands. Earthworms are typically scarce or absent from blanket bog and podsolie soils of dry heather moor. Two species of cranefly Tipula subnodicornis and Molophilus ater have a highly synchronised emergence in late May and early June on blanket bog, and give an above-ground biomass greatly exceeding that on mineral soils for this short period. By contrast, the above-ground biomass of invertebrates on mineral soils shows a much greater spread, at an overall higher level, throughout the spring, summer and autumn. The difference between a short seasonal eruption dependent on only two species, and a more spread-out availability of at least 15 species (albeit each at lower densities) is likely to give substantial differences in the feeding value of these two contrasting habitats. Dr Coulson considers that the marked preference of sheep for limestone grasslands may have two relevant effects. First, the presence of sheep droppings greatly increases the numbers of earthworms, dung beetles Aphodius, and Diptera associated with dung such as Scopeuma and Borboridae all potential or known food for Golden Plovers. Secondly, grazing keeps the bid short and this seems greatly to assist movement and searching, so that feeding may be more efficient on limestone grassland than on the longer Calluna-Eriophorum community of blanket bog. It is interesting that although many of the parent birds are able to take advantage, through flight, of richer feeding grounds, the chicks are limited to the food they can pick up in the vicinity of the nests; for the bulk of the population this means that the young grow up in a particularly unproductive environment. Once the chicks are hatched, both parents usually tend them continuously and so are also dependent at this stage on the food supply of the nesting grounds. Some pairs with a choice of habitat will move with their chicks to ground away from

30 BREEDING OF THE GOLDEN PLOVER 87 the nest-site, and may then be found on steeper or wetter places than those normally used for nesting. The relatively large size of the Golden Plover's egg can be regarded as an adaptation to breeding on these barren and climatically inhospitable uplands, giving the chick a good start in life. The Lapwing is a larger and heavier bird (mean weight, in breeding condition, 217 gm against 191 gm for the Golden Plover), but lays a much smaller egg, for it nests largely on the richer hill grasslands and its chicks immediately have access to a better food supply. The average measurements of 45 fresh Lapwings' eggs at m in the Pennines were length 46.7 mm by breadth 33.5 mm and weight 25.1 gm, compared with 52.0 mm by 36.2 mm and 34.9 gm for 102 fresh eggs of the Golden Plover at a similar elevation. The longer laying and incubation periods of the Golden Plover accord with this difference. Golden Plover eggs are laid at intervals of hours, and the average incubation period is 31 days; in the Lapwing the times are hours and 26 days (D. and C. Nethersole-Thompson 1961 (a) and (b)). With a full clutch of four eggs representing about 73% of the bird's body weight, it is hardly surprising that the Golden Plover requires good feeding areas to build up the necessary reserves, or that the replacement period is lengthy (12-24 days). For the Lapwing, a full clutch represents only 46% of body weight. Murton and Westwood (1974) have shown that in East Anglia the mean weight of Lapwings' eggs varies according to soil fertility, being higher on arable farmland (27.8 gm) than on uncultivated 'brecks' (25.3 gm). They interpret this as an effect of difference in feeding value of habitat, the richer soils allowing the parent to build up larger reserves of food prior to laying. In the Pennines, no significant difference in size or weight has been found between the eggs of Golden Plovers on the fertile limestone and those on acidic moorland. This is hardly surprising, since the bulk of the food reserves which go into the eggs are built up through feeding elsewhere than on the actual nesting ground, i.e. on the richer pastures of the lowlands or upper hill farms. It is noteworthy, however, that the mean weight of Lapwings' eggs from Pennine limestone grassland approximates to that of eggs from the more infertile East Anglia habitats studied by Murton and Westwood. VARIATIONS IN BREEDING DENSITY This section examines some of the geographical differences in breeding densities of the Golden Plover, but discounts variations which arise from limiting availability of suitable habitat. Breeding density is most readily measured as the

31 88 BIRD STUDY average distance between adjacent pairs, estimated as the observer crosses the nesting grounds, though complete counts of the numbers on particular areas are preferable. The following categories have been recognised: Density Very high High Moderate Low Very low Average spacing (m) < > 1650 Average area (pairs I ktn 2) > <0.5 Only in a few areas has breeding density been measured accurately by counting the nesting pairs on the basis of actual nests found. Estimates of density based on counts of calling birds seen during the incubation period are unreliable; only when the eggs have hatched are the birds certain to be calling and conspicuous, but they may then be over-estimated (p. 76 seq.). Counts made on some parts of a district before hatching and on other parts after hatching may thus give a false idea of variations in density. Table IV gives an extremely patchy and incomplete indication of density differences for recent years, and much more census work is needed. The variations summarised in Table IV show some correlation with differences in soil fertility of the areas concerned, though this factor does not provide a total explanation. The low numbers and density in southwest England and Wales cannot be explained in this way, and present variations within the Peak District are somewhat pluzling. However, farther north it remains generally true that high to very high densities are associated with the presence of substantial areas of baserich soil within the feeding range, and that, conversely, low density is characteristic of large, continuous tracts of acidic, infertile terrain. The Pennines from the Craven District to the Tyne Gap are one of the main strongholds of the Golden Plover in Britain, with not only a large extent of suitable ground, but also a good average breeding density. Their Carboniferous formations have large tracts of acidic gritstone giving extremely poor soils and extensive blanket bogs, but they also have much limestone ground with rich pastures. The limestone is sufficiently exposed or present in drift along most of the main Pennine dales to have encouraged farming well up the hill slopes, so that there are here rich feeding areas. One area of unenclosed limestone grassland, Mallerstang, supports the highest density known at present (Table II, col. 1 and Table IV), but there are also considerable variations in density from place to place, even within quite small areas of the Pennines. In the Southern Uplands, high density is confined to certain massifs composed of Ordovician or Silurian greywackes and shales which give relatively fertile soils, at least around the edges of the nesting grounds. By contrast, the acidic granite hills of Galloway have almost uniformly low to very low densities. This contrast is even more pronounced in the western and northern Highlands and Islands where, over vast tracts of suitable wet heath and blanket bog on the acidic Lewisian, Torridonian, Cambrian and Moinian formations, density is between very low (or even absent) to moderate, but becomes noticeably higher on exposures of the Durness limestone at Knockan and Inchnadamph in Sutherland, or close to crofting land with improved grassland. In both east and west Suther-

32 BREEDING OF THE GOLDEN PLOVER 89 TABLE IV. GEOGRAPHICAL VARIATIONS IN BREEDING DENSITY OF GOLDEN Breeding density Period Evidently low District and Area SOUTHWEST ENGLAND (Dartmoor) WALES PEAK DISTRICT NORTHERN PENNINES Mallerstang limestone plateau (Table 2, area 1) Alston Moor (part Table 2, area 2) Widdybank Fell and Cronkley Mickle Fells, Upper Teesdale Great Shunner Fell Fair Hill (Mallerstang Edge) Melmerby Fell Skirwith Fell (Cross Fell) Summer Lodge Moor Ravenseat Moor Nine Standards Rigg Bowes Moor Baugh and Wildboar Fells Tindale Geltsdale Fells Moor House NNR NE Slopes of Crossfell LAKE DISTRICT Skiddaw Group SCOTTISH BORDERS Bewcastle--Gilsland Moors Carter Fell, Cheviots SOUTHERN UPLANDS Moorfoot Hills (Table 2, area 6) Spango Water, NW Dumfriesshire Cairn Table Hills N of Gatehouse of Fleet Tarras Water, E Dumfriesshire Tweedsmuir Hills Only low moderate on most remaining breeding areas but numbers said to fluctuate, and one area in W Montgomeryshire sometimes has high density, with c. 40 pairs on 2,100 ha (5.3 pairs /km2) Varies from low to high but with overall average of 1.93 pairs/krn2, corresponding to moderate density Varies from low to very high Very high, with 18 pairs on ha (16.4 pairs /km 2) High, with 36 pairs on ha (5.0 pairs /km, 2) Evidently high, at least in some parts Evidently high on most of the ridges at over 610 m High, at over 650 m Moderate Moderate, perhaps low Moderate Moderate Moderate Moderate Moderate Low moderate Low moderate Moderate & Varies from very low to high Generally high, with 20 pairs on sample area of 372 ha (5.4 pairs /km2) Evidently high locally 1975 Evidently high 1960 Moderate high Moderate, perhaps only low in places Moderate on ground over m PLOVER IN BRITAIN Observer A. Archer-Lock H. Morrey Salmon E. K. Allin W. M. Condry J. M. Harrop R. Lovegrove D. W. Yalden D. A. Ratcliffe D. A. Ratcliffe I. Findlay D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe A. I. Kerr & J. Martin A. D. Watson A. D. Watson D. A. Ratcliffe D. A. Ratcliffe

33 90 District and area Moffat Hills Merrick, Kells, Lamachan and Cairnsmore of Fleet hills SCOTTISH HIGHLANDS SOUTH AND SOUTH- WEST Callander Braco Moors Loch Lomond Trossachs area Head of Glen Almond Breadalbane Hills Rannoch Moor MID-WEST Morvem and Sunart CENTRAL Dalnacardoch Gaick- Glen Feshie Fealar Glen Ey West Drumochter Hills East Drumochter Hills Monadhliath Inshriach Forest Invereshie Forest (low ground) Reyvack Moor Rothiemurchus Forest (low ground) Abernethy Forest Dorback Moor West Cairngorms (Braeriach, Cam Ban, Moine Mor) Central Cairngorms (Cairngorm Ben MacDhui) East Cairngorms (Beinn a'bhuird, Ben Avon) EAST Clova Caenlochan Lochnagar Glen Esk Crathie Upper-Strathdon- Glen Livet Cabrach Kerloch NORTHEAST Ben Wyvis BIRD STUDY Breeding density Period Observer Low on ground over 700 m Mostly very low, but a few pockets with moderate density Varies from very low to high High locally Low Low, with high patches D. A. Ratcliffe A. D. Watson D. A. Ratcliffe Moderate, perhaps high E. Blake Low J. Mitchell 1965 J. Mitchell D. A. Ratcliffe D. A. Ratcliffe C. J. Mead Very low, locally absent 1961 D. A. Ratcliffe Moderate with pockets of A. Watson high density D. Nethersole- Thompson Moderate, perhaps high A. Watson locally Low D. Nethersole- Thompson Low to moderate, with some D. Nethersolehigh patches Thompson A. Watson Moderate 1969 D. A. Ratcliffe Low D. Nethersole- None recorded j Thompson Low, with moderate patches D. Nethersole- Thompson Moderate D. Nethersole- Thompson Low D. Nethersole- Thompson Low D. Nethersole- Thompson Moderate to high D. Nethersole- Low to moderate f Thompson Low to moderate D. Nethersole- Thompson Occasional only D. Nethersole- Thompson Low A. Watson Moderate Moderate, perhaps high locally Moderate generally, high locally Moderate high D. A. Ratcliffe D Jenkins A. Watson A. Watson R. Parr Low to moderate 1959 D. A. Ratcliffe

34 District and area Struie Inveroykell Braelangwell Dalreavock Scriberscross Morvich Dunrobin Ben Armine Forest Badanloch Forest Strath Helmsdale Kildonan Lairg Forest Strathmore Mybster flows NORTHWEST Loch Maree Torridon- Flowerdale Hills Beinn Dearg massif and Fannich Forest Altnaharra and Ben Klibreck Laxford Scourie Moors Gualinn Foinaven- L. Stack, L. Eriboll Parphe A'Mhoine ISLANDS Skye Rhum Orkney, Mainland Unst, Shetland Fetlar, Shetland BREEDING OF THE GOLDEN PLOVER 91 Breeding density Very low Low Low Moderate Low to moderate Moderate, high locally Moderate, high locally Low, moderate locally Low, moderate locally Moderate to high locally Low Low, moderate locally Very low, absent locally Low Moderate to high locally Very low to low Very low to moderate Low to moderate locally Low to moderate locally Very low to low, absent locally Low to moderate Low (8 pairs on 1,098 ha- 0.7 pairs /km) Low high, but variable (58 pairs in pairs / km 2 overall) Low high, but variable (67 pairs in 3,450 ha- 1.9 pairs/ km 2 overall) Period Observer B. & P. Thompson P. & R. Thompson B & P. Thompson D. Nethersole- Thompson D. Nethersole- Thompson D. Nethersole- Thompson D. Nethersole- Thompson D. Nethersole- Thompson D. Nethersole- Thompson I. D. Pennie D. Nethersole- Thompson 1969 D. A. Ratcliffe D. A. Ratcliffe D. A. Ratcliffe I. D. Pennie A. Watson E. G. Holt D. A. Ratcliffe D. Nethersole- Thompson D. A. Ratcliffe A. Watson A. Watson (Sen.) D. Nethersole- Thompson D. A. Ratcliffe A. Watson A. Watson (Sen.) D. A. Ratcliffe H. J. B. Birks P. Wonmell 1975 D. Lea and W. R. P. Bourne G. Bundy 1975 G. Bundy J. F. Hunt NOTES: These data have been chosen to avoid the complications introduced by the long-term population declines which have occurred in many areas; these are dealt with later. Some observers claim marked fluctuations from year to year in some areas, but I have been more impressed by the recent population stability in areas studied closely, i.e. Mallerstang, Alston Moor and the Moorfoot Hills.

35 92 BIRD STUDY land, Orkney, and the Spey valley, D. Nethersole-Thompson (pers. comm ) has found density to be markedly higher close to surrounding farmland or crofting enclaves, both worked and abandoned, compared with the barren moor. In Kincardineshire, R. Parr (pers. comm.) has also noted a similar marked increase in density between the centre of a grouse moor and the edges bordering farmland. Although there is much unsuitable terrain or barren moor, density is locally higher on the Hebridean island of Rhum, where there are substantial areas of basic rock and rich grassland, than on many parts of the adjoining mainland. In Shetland, G. Bundy (pers. comm.) found the density high on the Serpentine areas of Unst and Fetlar, while numbers were generally much less on the acidic gneiss and granite moorlands. A pair or two of Golden Plovers breed now and then on the tops of small, steep-sided seabird islands along the western and northern seaboard, such as Hirta of St Kilda, Handa and Fair Isle, where the combination of sheep, birds and sea-spray produces maritime heaths and grassland on quite fertile soils. Soil fertility differences fail to account for some differences in density. In the Southern Uplands there is, for example, no obvious reason why the Moorfoot Hills should have a much higher density than the geologically, topographically and climatically similar group of hills around the Tarras Water (Table IV). The only obvious environmental difference is in management. The Moorfoot Hills have been carefully managed for Red Grouse, and both grouse and Mountain Hare Lepus timidus numbers are high, but the Tarras Moors, which from carried heavy stocks of grouse, have deteriorated in this respect: sheep numbers have built up and heather has decreased or disappeared over large areas where it was once dominant. Grouse and hares have declined markedly on Tarras. There is, in fact, some tendency for densities of Red Grouse, Mountain Hares (where they occur) and Golden Plovers to run parallel. In the Highlands, the best areas for all three are in the east. Grouse management is favoured by the relatively dry climate, congenial to the growth of heather, and by a gently contoured topography of the type favoured by Golden Plovers. Careful rotational burning of the heather and control of sheep numbers aim to give maximum nutritional value of the vegetation for grouse, and incidentally benefit Mountain Hares. The connection with Golden Plovers is less clear, unless systematic rotational moor-burning has parallel value for it in maintaining open and productive feeding areas. Long heather is probably difficult feeding habitat for the plovers, especially the chicks. Grouse numbers also show a correlation with the fertility of the underlying substrata, so that high density of both species may in general reflect plant/soil systems with good nutrient budgets. While the importance of nutrient status for the Golden Plover is presumably through its effect on invertebrate populations, if the vegetable component of the bird's food is deliberately eaten, there could also be a more direct effect. Not all the moors with high densities of Red Grouse have high densities of Golden Plovers, and there are some exceptions at present in the Pennines, such as the Tindale and Geltsdale Fells and Bowes Moor. In any case, there seem to be few places nowadays where high density is maintained over large areas, and in general the species shows a tendency to marked variations in density within relatively small areas (Table IV). In the better areas of the eastern Highlands, Adam Watson (in litt.) commented that they are never distributed evenly in high

36 BREEDING OF THE GOLDEN PLOVER 93 numbers but occur patchily. Accurate counts for a larger number of areas are needed, and over several years in localities where it is said that annual fluctuations occur, before these varying density patterns can be properly quantified and the reasons for them understood. The food advantage of the Mallerstang limestone breeding ground over the best blanket bog areas for Golden Plovers nowadays would appear to be threefold, if differences in breeding density are accountable solely in terms of this factor. This is rather surprising in view of the superabundance of food on the blanket bogs at hatching time, but until there is more precise information about spring and summer food requirements, our understanding of these relationships must remain uncertain. It may well be that both chicks and adults need a good deal more than craneflies in their diet. A further point is that as the breeding season progresses, and especially when there are chicks to care for, both adults of a pair become increasingly dependent on the nesting grounds for their own food. When the chicks are hatched both parents are typically in attendance, whereas during incubation it is more usual for the non-sitter to be elsewhere. In summary, there is a clear correlation in many areas between breeding and soil fertility of the nesting ground : and soil fertility differences appear in turn to produce varying levels of feeding value of habitat which could be causative. This would be entirely consistent with the similar relationships which are well established for a number of herbivorous vertebrates in Britain, including sheep, Red Deer Cervus elaphus, Rabbits Oryctolagus cuniculus, and Red Grouse. The process whereby the Golden Plover responded in its breeding density to carrying capacity is of great interest, and evidently involves the proximate mechanism of spacing behaviour. The study of changes in density in time, as in the long-term decreases which have occurred in many places, could throw light on these relationships. Study of decline may also help to identify other quite different factors which have an important influence on breeding density. There seems to be little information on breeding densities of the species elsewhere in the world. Haviland (1915) said of the Asiatic Golden Plover that "each pair occupied perhaps two furlongs of tundra" around Golchika at the mouth of the River Yenesei in Siberia. This would correspond to high density in Britain. For the same species on St Lawrence Island, Sauer (1962) estimated territory size as from a quarter to half a square kilometre (100 ha), i.e. moderate to high density, on tundra which varied from dry gravel fields (the nesting sites) to marshes. Parmelee et al. (1967) estimated one pair per square mile within five miles of the beaches of Cambridge Bay, Victoria Island. Farther inland their numbers dropped rapidly and Black-bellied Plovers Pluvialis squatarola were more numerous. BREEDING DISPERSION At moderate to very high breeding densities, Golden Plovers typically show a markedly regular dispersion of nests (p. 78). Detailed study of three different populations has provided accurate measurements of this spacing, and Table V shows the small variation in mean distance between adjacent nests in each area. The values are also remarkably constant for each area from year to year. Figure 3 is an example of the regularity of spacing in one area in one year. This regularity of nest spacing and its constancy in time point to a high and stable degree of

37 94 BIRD STUDY Figure 3. Dispersion of Golden Plover nests on Mallerstang limestone grassland in 1974 territoriality between pairs, giving the impression that each breeding area is shared out evenly every year into territories of similar size. The nest itself is assumed to be the central and focal point of a territory. Nest spacing of this kind is characteristic of certain predatory birds e.g. Raven Corvus corax and Peregrine Falco peregrinus (Ratcliffe 1962), Golden Eagle Aquila chryscietos (Brown and Watson 1964) and Sparrowhawk Accipiter nisus (Newton, in press). The actual distances are much greater in most of these species, but the regular and constant pattern is the same. Nest spacing in these predators is affected by the distribution pattern of suitable breeding cliffs and woods, and in some areas this is such as to prevent complete regularity. For the Golden Plover, suitable nesting habitat is typically continuous over large areas, so that there is seldom any such chance limitation to regular spacing. On some breeding areas the topography, vegetation and management (especially moor-burning cycle) influence the placing of nests, so that certain places on the moor tend to be used repeatedly; whereas on others the siting of nests over a period of years appears to be more or less random. There are, however, indications that individual pairs returning in successive years tend to nest closer to the site of the previous year sometimes actually in the identical scrape than do new birds. On Alston Moor, R. Stokoe (pers. comm.) found the same scrape to be used for six successive years, but this

38 BREEDING OF THE GOLDEN PLOVER 95 TABLE V. NEST SPACING IN ERENT BREEDING AREAS Mallerstang limestone Alston Moor Moorfoot Hills grassland blanket bog blanket bog ± ± ± ± ± ± 80 } ± ± ± ± ± ± ± Figures are mean distances in metres between each nest and its nearest neighbour, with standard errors (95% confidence limits). (Distances are estimated in the field by measured paces.) is exceptional. Thus, whereas the predatory birds, resorting typically to traditional nesting sites over a long period, tend to give their breeding territories a certain permanence of identity, the Golden Plover has territories which are essentially ephemeral and may, indeed, best be regarded as entirely hypothetical. It is then pertinent to enquire into the nature of territorialism in the Golden Plover. D. and C. Nethersole-Thompson (in Bannerman 1961) record that Golden Plovers are strongly territorial, frequently showing aggressive defence of the nest site, in addition to the distinctive display flight, evidently performed by the males only. The males also fight each other for mates and continue to defend their hens against rivals after eggs are laid. This overt aggression and advertisement display are evidently the mechanism which spaces out pairs and helps to determine the actual position of nests. Sauer (1962) found the Asiatic Golden Plover to be strongly territorial and describes the visual display and song marking of boundaries between territories. In the previous section, nest spacing was used as a measure of breeding density, and its varying magnitude from one area to another expresses differences in this density. It would seem that underlying these variations are differences in the strength of territorial interaction according to locality. Since variations in breeding density show some correlation with differences in food quality of the nesting grounds, degree of territorialism must also show a similar correlation with this factor. Correlations of a similar kind have been shown in Peregrine and Raven (Ratcliffe 1962), and support the view that one function of territorialism is as a response to carrying capacity, adjusting population density to a level which the particular nesting grounds can support. The argument has been amplified by

39 96 BIRD STUDY Wynne-Edwards (1962) in evidence for his controversial view of the homeostatic regulation of animal populations. Holmes (1970) also concludes that in Alaska the density of breeding Dunlins Calidris alpina is related to the abundance and availability of their food supply and that the main function of territorial behaviour is to disperse the populations in relation to food. In the Peregrine and Raven, an important point in the thesis is the existence each year of a surplus floating population of potential breeders which is excluded by the territory holders. Territorialism thus involves more than just an even spacing of whatever number of birds chances to arrive on the breeding grounds. Moreover, in these predators, whose food supply remains fairly steady from year to year, the constancy of spacing (and hence of breeding population) over a long period is a further point suggesting that territorialism is actually geared to food supply. Whilst I have not confirmed the presence of a non-breeding population of Golden Plovers in any of the main study areas, the birds occurring on certain upper farm pastures may well include such a surplus. Several observers elsewhere have commented on the presence of small non-breeding flocks on lower ground adjacent to the breeding haunts throughout the nesting season. C. and D. Nethersole-Thompson (1961(a)) considered that non-breeding birds in the Cairngorm foothills were sexually immature, but some of the non-breeders elsewhere, during the nesting season, could include surplus birds which were excluded by the territorialism of the breeding population. David Jenkins tells me that the fields at the edge of the Glen Esk grouse moors, Angus, also regularly held a breeding season population of Golden Plovers, evidently composed of off-duty feeding birds from the nesting grounds and non-breeders living permanently on these pastures. At Barrow, Alaska, Holmes (1970) established by removal experiments that there was a floating, surplus population of Dunlins from which birds rapidly obtained breeding territories when vacancies were created. The clear implication was that these birds were previously prevented from breeding by their inability to gain territories. The difficulty about validating the hypothesis that breeding density is adjusted to carrying capacity (i.e. available food supply) by the intermediary device of territorial interaction is that it postulates a mechanism which cannot be directly observed in any meaningful way. The critical approach of studying response after artificially altering the food supply for one or more populations would be difficult to attempt with the Limicolae. If the above interpretation is correct, increase in food value of habitat should be followed by increase in density through reduction in territorialism, and vice versa. There are, however, further obstacles to establishing such an hypothesis. The first is that, up to hatching time, the Golden Plover typically feeds non-territorially on areas well away from the nesting grounds. The second is that in this species territorial behaviour appears to cease when there are chicks, if not before, as Holmes (1970) also found in the Dunlin. The Golden Plover is thus doubly interesting in that it establishes territories on the nesting ground which it does not use to any extent for feeding until the eggs have hatched, by which time territorial behaviour appears to have broken down. Territorialism at the onset of the breeding season limits the overall potentiality of a population for increase, by restricting the number of pairs which can nest On a particular area. Within these limits other factors later determine precisely how many young are reared. Territorialism thus places a basic limitation on the

40 BREEDING OF THE GOLDEN PLOVER 97 demands which the population can make on its food resources, but is there a specific relationship between spacing behaviour and the carrying capacity of the nesting ground? More particularly, could such a relationship involve a balancing of breeding density below the 'ceiling' set by the food supply? Even allowing that unpaired males may establish territories, spacing behaviour in breeding birds is essentially a group phenomenon, involving a minimum of four birds in the simplest possible situation. An individual's spacing behaviour is only expressed fully when it is paired and can, with its mate, react against another pair. In practice, a larger group is almost invariably present. Yet selection for the aggression which must be involved has to be at the individual level. This is feasible, for increase or decrease in aggression of one individual in a basic quartet will modify the response of the group as a whole. If a pair of Golden Plovers rears a full brood of four young, their breeding area has to meet a food requirement which has trebled within a short time. Since the parents are apparently no longer drawing on the richer feeding grounds beyond the moor edge, it would seem that the period after hatching must be the time when the Golden Plover comes closest to a ceiling in the availability of food during an average nesting season. If a population began the season with a density which later gave a level of numbers critically close to the food ceiling, the chances of any pair rearing a brood to fledging and surviving themselves would seem to be less than if numbers had been farther below the food ceiling. And since the survival of the young eventually determines the survival of the population, there would be selection for those individuals whose spacing behaviour led them to begin the breeding season at a density which later imposed less strain on their food resources, and thus increased the survival chances of their young and themselves. This would, in fact, mean selection for the more strongly territorial birds which were less tolerant of other pairs settling close at hand. This response is allied to a peck-order within the population, accounting for the exclusion of low-rank individuals into a non-breeding surplus. The selection pressure and the response are separated in time, through the non-breeding period. Selection is strongest at the climax of the breeding season, but it operates on a behavioural character expressed several weeks earlier, and its effects are not manifest until the beginning of the next breeding season. The delayed effect is puzzling, for instead of appearing as an effect, the resulting behaviour seems to be anticipatory and causal. The pairs space themselves out at a particular density at a time when it is of no immediate advantage for them to do so. The advantage comes later, but once the critical density has been determined, the continuation of territorial behaviour into the post-hatching period might be disadvantageous, through unnecessary use of energy, exposure to predation during display, and restriction on use of especially good localised feeding places on the nesting grounds. The tendency for species to exploit their habitat to the maximum would apply a converse pressure counteracting the tendency to wider dispersion, ensuring that density reached an optimum level, balanced between too close to the food ceiling and too far below it, on any particular breeding ground. This explanation seems to show adequately how breeding population level can become adjusted to carrying capacity of habitat through territorial behaviour, using the orthodox concept of individual selection, and avoiding any recourse to altruistic behaviour or group selection. The exact nature of the territorial response

41 98 BIRD STUDY nevertheless remains uncertain. Territorialism in the Golden Plover appears to involve other behavioural mechanisms besides simple aggression. But if selection is for straightforward behavioural response, then an individual bird's territorialism is likely to be fixed in strength for its life. The relationship between nest spacing and carrying capacity of habitat would then be inflexible, and an individual would have little if any ability to adjust to a habitat of different carrying capacity, or to a time-change in food supply in the same area. If selection were, instead, for an instinctual responsiveness whereby a bird's territorial behaviour was related to its perception of habitat quality, there would be much greater flexibility for adjustment to change in food supply spatially and in time. This hypothesis would not need to invoke any idea of prescience on the part of the bird in reacting to its habitat. Although the context is different and selection pressure more obviously direct, birds in general have a strongly developed capacity for reacting to the visual qualities of their habitat in their choice of nest site. The idea of a territorialism mediated by perception of habitat might be tested in a species more amenable to experimentation, but hardly in the Golden Plover, and it is mentioned here simply as a possibility. The point which remains obscure is how the availability of food to the bird begins to exercise a selection pressure affecting survival chances of individuals during the critical period after the young have hatched. Careful observations on feeding behaviour within and between families might help to throw light on this problem. POPULATION CHANGES Many Golden Plovers breed and feed on ground which was forest earlier in the Post-glacial Period, and became denuded of trees through human activity, mainly during the last 2,000 years. A great deal of hill land was obviously unavailable to the species before the era of widespread deforestation. The potential climatic tree limit varies and shows altitudinal descent northwards and westwards in Britain. In northwest Sutherland, the natural tree line now would probably be at no more than 305 m, with a still lower limit on the most exposed coasts, but in southern districts such as the Breconshire mountains and eastern districts such as the Cairngorms, tiny fragments of scrub woodland still persist at around 610 m. It is therefore probable that human intervention greatly extended the distribution

42 BREEDING OF THE GOLDEN PLOVER 99 and population of Golden Plovers breeding in Britain by creating large areas of suitable new habitat, but we shall never know the densities at which the species occurred on the natural nesting habitats above the former forest Emits, before this expansion occurred. Moreover, many areas of blanket bog below the tree line were probably once too wet to have been wooded, and would be available as nesting areas in these ancient times. The Golden Plover in Britain probably reached its maximum breeding population around when the phase of deforestation was virtually complete, and before other changes reduced its numbers. Declines have occurred in many districts since, but they have differed in scale and the causes appear to be varied. The data for these declines are only approximate, for this was a species seldom counted, and the records are mostly subjective. Parslow (1973) has given a summary of the general decrease in Britain and Ireland. Decline has been most obvious in areas where the nesting habitat has been profoundly modified or destroyed. The most clear-cut instances are on moor extensively afforested with conifers by the Forestry Commission, beginning in 1919, and reversing the expansion of habitat caused by deforestation. On the great expanses of Border moorland drained by the rivers North Tyne, Irthing, Lyne and Liddel, Golden Plovers nested in considerable numbers and moderate to high densities in the 1920s and 1930s (Graham 1932; H.M.S. Blair, pers. comm ) The steady march of conifers has gradually obliterated the nesting grounds, up to 450 m locally, and even where there is still much suitable terrain above the forest limits, this may be rendered less favourable by reduction of feeding areas at lower levels. Of the 50,500 ha of new forest in Northumberland, Cumberland and adjoining areas of the Scottish Border counties, a very large part represents breeding habitat lost to the Golden Plover for the foreseeable future. The extensive afforestation of the Denbighshire moors is likely to have been a major factor in the decline of the species in this part of North Wales, e.g. on the area around Glan Ceirw, and much suitable nesting ground has been planted in other parts of the Cambrian mountains farther south. In Galloway, substantial areas of moorland known to A. D. Watson and the writer as nesting haunts of Golden Plover after 1947 have gradually been covered with young forests; the densities were low in some areas, but high locally to the east of the Ken valley in Kirkcudbrightshire, and on some of the Wigtownshire flowes the total number of birds displaced is probably considerable. Golden Plovers disappear completely from afforested moorland, but in many other districts where environmental change has been outwardly negligible there have been decreases involving mainly a reduction in breeding density, or withdrawal from still suitable habitat. As Parslow (1973) has shown, the general trend to a reduction in numbers during the last hundred years is unmistakable, though the evidence for the areas concerned is mostly qualitative and fragmentary. Though the overall picture of decline is consistent, it is somewhat confused by the assertions of many observers that breeding populations fluctuate markedly from one year to the next. While numbers may well show such fluctuations in certain areas, my own experience is the reverse, with populations showing only slow decline, or else remarkable constancy, over periods of a decade or more. Some known examples of decline will be described, but the possible causes will be discussed in the next section.

43 100 BIRD STUDY The first evidence of such decrease seems to come from the western Highlands and Islands. Many earlier writers on Scottish birds refer somewhat vaguely to the Golden Plover as a numerous breeder on most uplands (Baxter and Rintoul 1953). For instance, Gray (1871) said: "This beautiful plover is extremely abundant over the whole of the western counties of Scotland, where it breeds on all the hills of moderate elevation." Even allowing for possible exaggeration by these earlier writers, the present position in many districts is so different that at least local decline is indicated. By the beginning of the present century, Harvie-Brown and MacPherson (1904) detected a sharp diminution in the northwest Highlands. Harvie-Brown (1906) also noted a decrease in the central Highlands (Perthshire) by this time. Mackenzie (1924) observed during his long life a general and substantial decrease in numbers of moorland birds in the area around Loch Maree, Wester Ross; these included Golden Plover, Ptarmigan Lagopus mutus, Red Grouse, Black Grouse Lyrurus tetrix, Partridge Perdix perdix, Snipe Capella gallinago, Lapwing and Dunlin. Mountain hares also declined. Breeding density of Golden Plovers has remained low through most of the western Highlands and Islands up to the present day. For Shetland, L. S. V. and U. M. Venables (1955) comment: "Saxby, who lived in Shetland from 1859 to 1871, wrote: 'The Golden Plover breeds abundantly in every part of Shetland, even on the small outlying holms.' With the possible exception of Gluss Isle (180 ha), we know of no holms that are now occupied and the smallest island where we know of regular breeding is Uyea, Unst (240 ha). Though it is a widely distributed hill species in the nesting season, we would certainly not describe it as common, either in Unst or elsewhere, at the present day. This view was shared by Harvie-Brown and Evans and Buckley from their observations in 1890, 1891 and 1892." Kearton (1899) also commented that the species was quite scarce in Shetland in In Orkney, Buckley and Hat-vie-Brown (1891) described the Golden Plover as widespread but not abundant, much less numerous than formerly, and still decreasing rapidly here and elsewhere. High densities persisted in some northern Scottish districts for another years or so. Ernest Blezard (pers. comm.) noted a dense population on the flowes of Strathmore, Caithness, in 1921, and D. Nethersole-Thompson (pers. comm ) was impressed by the high numbers where farmland passed into the open moor at Harray on mainland Orkney in 1932; four pairs were nesting within 4 ha in one place. These high densities no longer obtain, and Balfour (1972) indicates that there has been a substantial decline in Orkney since the 1930s. In the northern foothills of the Cairngorms, Inverness-shire, at Dorback, Abernethy and Rothiemurchus, D. Nethersole-Thompson (pers. comm.) found a population reduction of over one half between the 1920s-30s and the 1960s, and still greater decline on Invereshie and Inshriach. Habitats previously holding single pairs lost them from the 1950s onwards, but groups around cultivated fields and meadows maintained their numbers better. Substantial decline was reported during on the moors of Kintyre in the southwest Highlands (N. Paterson, per A. McArthur). On the southern fringe of the Highlands, in Glenartney, Perthshire, John Mitchell found Golden Plovers sparse in 1964, but met a shepherd who said they were once numerous. Marked reductions in breeding density have also occurred in various parts of England and Wales, though the pattern is irregular. As with Scotland, perusal of

44 BREEDING OF THE GOLDEN PLOVER 101 the literature gives the general impression that the bird was once more common in many areas, but the information is mostly too imprecise to be dependable. Kearton (1899) remarked, "... I have within recent years noticed a considerable decrease in its numbers throughout some old breeding haunts both in England and Scotland", and he referred specifically to a considerable decrease during the previous 15 years in the area around Nine Standards at the head of his native Swaledale in the Pennines. Most of the reliable evidence for decline comes from more recent times, after The Golden Plover still nests on Dartmoor but has not been recorded in recent years from Exmoor, Bodmin Moor, the Mendips or Blackdown Hills, where there were once occasional breeding records. Decline in most parts of Wales has been marked. In South Wales, Salmon (in litt.) records a contraction of distribution and reduction in total numbers in the counties of Glamorgan, Pembroke, Carmarthen, Brecon, Monmouth and Radnor, while the lower limits to breeding seem to have risen from under 305 m to 425 m. W. M. Condry (pers. comm.) believes that numbers have decreased in Cardiganshire; on one area where J. H. Salter saw numerous pairs in 1933, he has since 1945 found it sparse, or in some years absent. In North Wales, only occasional pairs are encountered nowadays on the highest tops of the Berwyn Mountains, where Forrest (1907) mentioned four nests found in one day in Forrest indicated that the species was frequent on the lower moorlands around Bala and Corwen in east Merioneth, and also occurred on the higher Aran and Rhinog ranges to the west, up to 1906, but it is sparse or absent today. There was an isolated breeding record on Anglesey in 1959, but nesting seems to have ceased since 1950 in Caernarvonshire to the west of the Migneint, although there is much suitable terrain. In the Peak District of Derbyshire and south Yorkshire, W. Gibbs (pers. comm.) knew very high densities from on Burbage Moor, Totley Moor and Big Moor, a mainly dry heather moor habitat at the low elevation of m. With a keeper friend, he commonly found nests in a single day and, on one exceptional occasion, 14 or 15 nests. This extraordinary score, which could probably not be equalled anywhere in Britain today, points to an exceptionally high density over a large area. The habitat itself is largely on acidic Millstone Grit soils of low fertility, but good feeding grounds on farm pastures lie within two or three kilometres of most of the area, and there is limestone ground within 5 km of Big Moor. Only a few pairs were seen on Burbage Moor by the writer in 1948 and 1949, though there was much suitable ground, and nesting was doubtful from (Yalden 1974). In the same district, Abney Moor was reported to have 20 pairs in 1941 and 12 pairs in 1944, but I saw only a single bird in 1948, and the population was extinct by 1952 (H. C. B. Bowles, in Yalden 1974). Elsewhere in the Peak District, I noted a high density on Birchinlea Moor, in the Derwent valley, in 1949, but only a low density in 1964, though the area seemed unchanged. On the east side of the Derwent valley, however, the low to moderate density observed in 1949 seemed much the same in Yalden (1974) has estimated the total breeding population of the Peak District in the period as around 400 pairs; this evidently represents a substantial decline since 1939, with withdrawal from lower peripheral moors and reduced density even on some higher, central areas.

45 102 BIRD STUDY In the northern Pennines, A. W. Colling (pers. comm.) recorded a seemingly very high density in Upper Teesdale, around Cronkley and Widdybank Fells, from Numbers are clearly lower nowadays, though still high; nesting on the lower pastures around Widdybank Farm was common from but is now infrequent. On the borders of Yorkshire and Westmorland, numbers are clearly much less since 1960 on Wild Boar, Swarth and Baugh Fells, than those found by Sedbergh School ornithologists during the late 1930s (per E. Blezard). At the extreme northern end of the Pennines, Blezard noted a decline on the Geltsdale Moors between , and the writer found numbers reduced on the adjoining Tinsdale Fells in 1974 compared with Farther west, in the Lake District, Golden Plovers have evidently ceased to nest on the Eskdale-Duddon Moors, Derwent Fells, Helvellyn Range, Ullscarf-Armboth Fells and Fauld's Brow, which were nesting haunts with from one to a few pairs regularly up to around Up to at least 1955, a few pairs bred on the low-lying (92 m OD) peat mosses around Hethersgill, just northeast of Carlisle, but these have disappeared. On the Bewcastle and Gilsland Moors, Graham (1932) formerly noted a high density (400 m spacing), but this was evidently reduced by 1945, and numbers have remained low on the unafforested parts of these moors. POSSIBLE CAUSES OF DECLINE Decline in breeding population as distinct from disappearance through wholesale destruction of habitat is not easy to explain, and may well be caused by a combination of separate factors. With a partial migrant such as the Golden Plover, the search for a cause of population decline has to take account of events on both the wintering and the breeding grounds. Outside the breeding season Decline in breeding population is evidently paralleled by the reduced size of winter flocks. Various observers have mentioned that the large flocks of former years are no longer to be seen but, since changes in winter distribution have evidently occurred in the past, it is possible that these apparent reductions may

46 BREEDING OF THE GOLDEN PLOVER 103 simply reflect displacement of some birds to other haunts. It is, in any case, very difficult at present to relate winter populations to breeding populations, either numerically or in ecological/geographical affinities. The ringing evidence suggests strongly that most British breeding Golden Plovers winter in this country, especially in lowland districts adjacent to the nesting grounds (e.g. Bannerman 1961). Winter flocks in any lowland district adjoining hill country thus probably represent mainly the local stock of breeding adults and their progeny, but the degree of exchange between flocks, their discreteness and movements are not known. The winter situation is complicated further by the influx of immigrant Golden Plovers from Arctic and Subarctic areas, for these northerners are indistinguishable from indigenous birds in winter plumage, and little is known about the distribution of their respective flocks, the possible mixing of 'races' in the same flock, and the sizes of their populations. Only an extensive ringing programme combined with accurate censusing of winter flocks over most of the country is likely to resolve these problems. Winter mortality Winter mortality for both adults and juveniles would seem to be largely a matter of food shortage in relation to weather, for Golden Plovers are shot much less than formerly, and natural predation is unlikely to be high. The species shows remarkable fidelity to traditional feeding places, resorting over many years to the same selection of grass and arable fields and ignoring others which appear to be equally suitable. It would, however, be unlikely for this fidelity to over-ride a pressing need to seek other feeding places. Golden Plovers certainly move about according to weather, and rapidly seek alternative quarters during hard frost and snow, notably on the coastal saltings and shores. The species seems so much at home on both ploughed land and pasture in most wintering districts that reduced availability of feeding habitat through large-scale land-use changes is extremely unlikely to be a limiting factor. Severe winter weather could be a limiting factor, through causing heavy mortality. Baxter and Rintoul (1953) comment that: "Golden Plovers seem to be very susceptible to severe cold: they become much emaciated in long frosts. In January 1889, during a severe storm, many were found dead and dying in Islay, and we have several times picked them up dead on the shore during hard frost. In the great storm in the early spring of 1947 many died, and they were recorded as below their usual numbers at many of their Scottish breeding places next season." In Shetland, L. S. V. and U. M. Venables (1955) said: "In the spring of 1946 breeding pairs were thinly distributed all along the six miles of hill east of, and parallel to, Whiteness Voe and the Loch of Strom. This was followed by an abnormally severe winter both in Shetland and over most of Europe, and in 1947 we did not find a single pair along this hill: indeed it was 1950 before the population recovered." Decline certainly took place in some areas in the 1940s, a period with several hard winters. Some observers claimed a decline after the severe winter of January- March 1963, when the species was said to be locally hard-hit in its winter quarters, but in three breeding areas under surveillance by the writer there was no sign of decrease compared with The large population of Lapwings on one of these

47 104 BIRD STUDY areas was severely reduced, but had recovered by The lack of any such cycle in Golden Plover numbers suggests that the long-term declines in this species have resulted from some other cause than winter severity. There are no clues to connect the introduction of the new synthetic pesticides to agriculture since 1945 with the decline of the Golden Plover, either through direct toxic effects or indirectly through an effect on food supply. Decrease had almost certainly begun in some areas before the introduction of DDT and gamma BHC, and was well established when the dieldrin group of insecticides appeared. Four eggs from different nests in the Pennines and Southern Uplands had only small amounts of pp DDE, pp DDT, dieldrin, BHC and heptachlor epoxide, and the Golden Plover shows no significant egg-shell thinning since 1946; egg-shell thinning is regarded as a good indication of more serious pesticidal effects (Ratcliffe 1970). This species did not figure in the large-scale 'kills' of birds attributed to the dieldrin-type seed dressings during , and locally to During the breeding season The main limitation of numbers during the year thus seems most likely to occur during the breeding season. If the total population has declined, as evidenced by size of winter flocks, this could be reflected in the breeding season in withdrawal from certain areas, or wider dispersion over the breeding grounds generally, or any combination of these two effects. In the Peak District there has certainly been a retreat from the lower and perhaps more marginal nesting haunts, and a reduction in density on some others. The same appears to be true in parts of northern England, but the decline on the Tindale and Geltsdale fells, compared with the stability on Alston Moor, km to the south, is puzzling unless the last is better ground being held preferentially. It may be that the maintenance of a breeding population in any district depends on its reproductive ability, and that there is little immigration from other districts. A strong tendency to return to the birthplace to breed would account for substantial geographical isolation of populations. It is also possible that separation into distinct populations is reinforced by the manner of aggregation into winter flocks. Reduction in carrying capacity of habitat Agricultural improvements have caused encroachment on moorland fringes in some areas, producing habitat unacceptable for nesting and thus reducing the population. In Orkney, the marginal ground at the moor edge which carried such a high density in 1932 is now improved farmland quite unsuitable for nesting (D. Nethersole-Thompson, pers. comm.). This form of direct habitat attrition has been a much less important factor than large-scale afforestation in causing overall population decline. More subtle land-use effects may, however, have been responsible for long-term decrease in certain districts. The early decline in the northern and western Highlands and Islands, apparent by 1900, may well be part of a general depletion in biomass of the upland fauna which has occurred through this whole region in response to a sharp reduction in productivity of the land. This deterioration covered the period described by Mackenzie (1924) and is believed to be the result of long-continued extractive exploitation, involving heavy grazing by sheep and

48 BREEDING OF THE GOLDEN PLOVER 105 deer, and moor burning. In a region with predominantly acidic, base-poor rocks and soils, a highly adverse balance of precipitation/evaporation and relatively low temperatures, the effect has been to reduce drastically an already low carrying capacity for vertebrates (McVean and Lockie 1969). A parallel fall in numbers of Peregrines seems also to reflect this trend (Ratcliffe 1963). In some districts well to the south of the Highlands, a post-1945 drop in numbers and density has coincided with increase in sheep stocks and has been paralleled by a decline in Red Grouse, e.g. in the Peak District (D. W. Yalden, pers. comm.). Such effects on vegetarian species might be expected, but it is less easy to understand how they lower carrying capacity for invertebrate feeders such as the Golden Plover, and one can only point to the general correlation. Another factor conceivably of some importance was the much increased destruction of the species by shooting in Victorian times, especially after the introduction of the breech-loading gun around The Golden Plover was an esteemed table bird and numerous 'sportsmen' refer to the numbers which might be killed at a single shot by firing 'into the brown' of a massed flock on the wing. Buckley and Harvie-Brown (1891) considered increased shooting to be a potent cause of the diminution then occurring in Orkney. If this factor had been important, though, one would expect the change to have been synchronous in different regions, but the north and west Highland decline seems to have been markedly out of phase with that elsewhere. Intensive military training during probably first caused decline on the Peak District area with high density noted by W. Gibbs, but there was no appreciable recovery when this activity ceased. Recreational pressures increased and regular management of the moor by rotational burning decreased after 1945, but these factors are unlikely to account wholly for the lack of recovery. Golden Plovers may have declined in a few areas through increased tourist pressure, but this factor is probably unimportant on the whole. In 1975 a nest hatched off only 14 m from a busy section of the Pennine Way track (I. Findlay, in litt.). In many areas, the nesting birds are disturbed a good deal by shepherds, but they do not readily desert their nests and disturbance has to be intense before it has much effect. Vehicles have obviously increased on many nesting grounds during recent years, but whilst on the move they cause little annoyance to the plovers, and I have known a nest hatch successfully only 12 m from a minor tarmac road. In many areas where reduction has occurred, there are no obvious changes in the appearance of the habitat or in human influences, and the causes of decrease must therefore be fairly subtle or inconspicuous. Long-term, slow declines in carrying capacity of breeding grounds could occur without noticeable change in outward character of habitat, and could be out of phase in different parts of the country. They would tend to accentuate the correlation between high density and highly productive habitats, such as those with limestone grassland. If the relationship between carrying capacity and territorialism is understood correctly (p. 95) then decline in the one would be followed by increase in the other, leading to wider dispersion and fall in breeding density. Declining carrying capacity could also have an adverse effect on chick survival and hence overall numbers, and any other factor which reduced either breeding success or adult survival could diminish the breeding population without territorialism being causally involved. The other advantages of territorialism may

49 106 BIRD STUDY cause the species to spread out evenly over the available ground, even when it is in low numbers, giving a wide dispersion either quite unrelated to carrying capacity, or related only indirectly. There are many areas, especially in the Pennines and eastern Scotland where, from subjective impression, breeding density is far lower than the character of the ground would seem to allow; and the marked differences between apparently similar districts are both frequent and puzzling. It is possible that Golden Plovers in many areas are at a much lower density than the ground could carry, and that territorialism is only a real limiting factor to density on habitats of especially good feeding value, where they naturally tend to congregate preferentially. The key to this problem lies in the proof of the existence or otherwise of a surplus of potential breeders in any area, a most difficult factor to establish (p. 96). Predation Although Golden Plovers are frequently taken by Peregrines in the breeding season, this predator is scarce or absent in many breeding areas. Hen Harriers Circus cyaneus sometimes take birds on the ground, but are again very local and sparse overall. The Fox Vulpes vulpes certainly catches sitting or roosting birds but I have noted only a few instances. The total impact of these predators is likely to be small, and since adult birds are virtually unmolested by humans during the nesting season, mortality in the adult population is probably very low. By contrast, in some areas the losses of eggs are heavy and may cause substantial reduction in output of fledged young. Golden Plovers seem especially vulnerable to egg predation and sucked or half-eaten eggs are found with striking frequency, whilst the numbers of parents evidently still with eggs or young on the moors in July is a further symptom. At one time humans commonly took the eggs to eat, but this practice has probably declined to insignificance, and the effects of egg-shell collectors have been even more negligible on this species. Sheep occasionally tread on the nests by accident, but I have known only a few instances. Shepherds' dogs sometimes raid nests, and some Foxes may search for them more systematically, but the most important predators of eggs appear to be other birds. On some moors there are scattered colonies of Black-headed Gulls Larus ridibundus, and more occasionally colonies, groups or pairs of Lesser Blackbacked Gulls Larus fuscus; these, especially the latter, are egg robbers. In some districts Rooks Corvus frugilegus have taken increasingly to foraging on the uplands, and sometimes raid nests. However, the Carrion Crow Corvus c. corone and Hooded Crow Corvus c. cornix appear to be the most serious predators of Golden Plover eggs. Ravens may locally contribute to losses. In some districts a high proportion of first clutches is predated, and though the species will repeat at least twice, predation continues and some pairs fail to hatch young in some years. Interestingly, predation is especially heavy on the Mallerstang limestone area with the very high breeding density. In 1974, at least 17 pairs nested here, but only four hatched clutches from the first round of laying% In 1975 probably 18 pairs laid, but all the first layings were lost. I was not able to follow events to the end of the season in either year, but in 1975 all of nine known repeats failed too. The predators were not seen raiding any nest, but at

50 BREEDING OF THE GOLDEN PLOVER 107 least one pair of Carrion Crows visited the nesting grounds early in the morning, and seemed the most likely culprits, though Foxes also range over the ground. Egg loss also appears to be heavy on the high nesting ground of the Crossfell range and the Tweedsmuir hills, which are within reach of Crows nesting on the lower slopes. In contrast to these sheep walks with their numerous Crows, predation of eggs is much lower on heavily keepered grouse moors, where Crows are eliminated, and a high proportion of Golden Plover pairs has young by the end of May, i.e. from first clutches. There is probably also less predation on remote moors far from the nearest breeding Crows. I have no exact figures, but would estimate that at least 50%-75% of first layings hatch in most years on Alston Moor and the Moorfoot Hills, two good grouse moors. In 1973 a single pair of Crows had escaped the vigilance of the keepers on the Moorfoots study area, and by 7 May at least four out of 20 pairs had lost their eggs; whereas in a typical year there would be no clutch losses by this date. If the stability of a particular population of Golden Plovers is closely related to its own breeding success, the differences in egg predation could be an important cause of geographical variations in numbers and density. It would make sense that high density shows a correlation with productive grouse moors, for these are the areas where predators are most completely suppressed and egg losses are therefore minimised. There would also be an explanation for population decline beginning in 1940, because of the war-time lapse in keepering on many moors a change which resulted in local increase in predators such as the Raven Corvus corax. A logical extension would be to connect population decline in many areas with the reduction in control of predators which accompanies a waning interest in grouse and an increasing interest in sheep, e.g. on the Tarras Water and in Glenartney. Although shepherds detest and persecute Crows Corvus corone, they are usually ineffective in controlling them, and most unkeepered sheep walks have breeding Crows. This interpretation is contradicted by the combination of very high density and very heavy predation in Mallerstang, but it may be that under special circumstances the normal converse relationship is over-ridden. This locality has a small area (110 ha) of especially favourable habitat with high carrying capacity at the edge of a very large and continuous expanse of nesting ground which supports in total a massive Golden Plover population, and includes extensive grouse moors where output of young is likely to be high. Under these circumstances there may be a continual pressure to colonise habitats which are marginal in certain respects e.g. exposure to predators especially when these offer the immediate advantage of good food supply. The effect of egg predation in lowering output of young needs first to be established. Golden Plovers will lay repeat clutches once, twice or even three times, and if young are produced from a reasonable proportion of replacement clutches, earlier losses may be of little consequence. There may even be some advantage to a wide spread of nesting during the season. There are no data on final breeding success for any area, and predation on unfledged young is a completely unknown factor. Crows and other successful egg predators will doubtless continue to search for any chicks which may hatch, and in some areas there could be a substantial lowering of final breeding success, compared with that on many grouse moors.

51 108 BIRD STUDY The ephemeral superabundance of potential food for chicks on blanket bog, beginning in mid-may and lasting only a month (see p. 86) may be of significance here. The hatching peak of successful first layings coincides closely with the onset of the cranefly 'explosion', and thus ensures that chicks can take full advantage of the good food supply. Since repeat clutches would appear to take up to a month, the chicks from them will hatch when the food supply is about to 'crash'. Such a situation could seriously reduce chick survival from first or second repeats, and on infertile moors the laying of replacement clutches may be an ineffective compensation for nest predation. It is an accepted tenet of animal ecology that the numbers of a predator are determined by its food supply, rather than vice-versa, but this may sometimes involve over-simplification, at least for species with a varied diet. A predator whose numbers are determined by the abundance of its principal food may be able to prey selectively on a minor item, to the extent of reducing this substantially without adverse effects to the predators. If Crows are particularly adroit at finding Golden Plover nests, as seems likely, they may be able, over a period, to depress the breeding success and hence eventually the numbers of Golden Plovers, without themselves being affected. A decrease in the number of young reared, year by year, could have a cumulative effect leading to a substantial reduction in population, with withdrawal from the more marginal areas, and thinner dispersion elsewhere. Egg and chick predation and, hence, the land-use changes which promote increase in the predators concerned, may thus be a major factor in the more recent decline of the Golden Plover in some areas. Predation could reinforce decrease in carrying capacity or be completely independent, according to local circumstances. Some areas show anomalies, and numbers have declined on certain well-keepered grouse moors, e.g. the Tindale-Geltsdale fells and Bowes Moor in the northern Pennines. The importance of predation will remain unclear until there are adequate data on breeding success, adult mortality, and the dependence of a local population on its own recruitment capacity. The Arctic Fox Alopex lago pus was found to be an important predator of wader nests on St Lawrence Island (Sauer 1962) and in Jameson Land, East Greenland (de Korte 1975), and the Asiatic Golden Plover and Golden Plover were among the species affected in the respective areas. Skuas Stercoarius sp. were also potentially important nest predators in both places. Climatic Change Climate could be involved in the decline of the Golden Plover. The general increase in temperatures in northern Europe between is well documented, but the trend has been reversed, with spring and summer mean temperatures showing a marked fall since Nethersole-Thompson (1966, 1973) has connected the post-1900 declines in British populations of the Snow Bunting Plectrophenax nivalis and Dotterel with the warming phase, and their subsequent recovery with the reverse trend. The geographical distribution of the Golden Plover in Britain and other countries, and the altitudinal descent of lower breeding limits with distance north in Britain, strongly suggest that this is a coldadapted submontane/montane species which might react in a similar way to

52 BREEDING OF THE GOLDEN PLOVER 109 climatic change. It has certainly shown a tendency to decline in its southernmost breeding areas and to withdraw from marginal haunts, especially those at lower elevations an expected pattern during the slow northward retreat of a species under adverse climatic change. There has, however, been no corresponding readvance since the climate became colder and other northern and boreal species began to expand southward and increase (Sharrock 1974, Williamson 1975). The Golden Plover may be less responsive to climatic change than other northern species. Virtually all the northern species which have shown southward breeding expansion in Britain are strongly migratory, and their spread may reflect a southward displacement of the migration stream, so that larger numbers of birds make contact with the right habitat at the right time in this country. The British Golden Plover is a different case: it has become adapted to our insular conditions and appears to be migratory on only a minor scale, from the hills to nearby lowlands and back. Since the climatic change towards colder conditions after 1945 seems to have been mainly in the breeding season, with winters recently tending to be milder than hitherto, it is difficult to see how displacement southwards could occur. Not enough is known about the mechanisms involved in climatically controlled changes in bird distribution and numbers though Williamson (in press) has suggested that easterly weather occasioned by the spring 'blocking anticyclone' over Fenno-scandia may well be responsible for stronger immigration, leading in some cases to settlement in Britain, of boreal elements from the Continent. There is a possibility that, as with some analogous vegetation changes, effects occur through the response of some competitor, predator or pathogen in an adjacent habitat, but probably a whole complex of factors is involved. Whatever the overall shift in environment, it must presumably be sufficient to cause disadvantage in returning to the same nesting place, or advantage in seeking one farther south or at a lower elevation. The Red Grouse is a still more sedentary species which has evolved to British upland conditions, and has not yet shown a tendency to retreat or advance with recent temperature changes. Probably for such species, the climatic change required to produce significant response has to be of much greater magnitude, and may require first a shift in whole ecosystems, including the vegetation. The dependence of birds on different climatic regimes is probably indirect, through their adaptation to a particular biotope or food supply which is more directly controlled by climate. In Britain, the high degree of environmental modification by man complicates and modifies the influence of climatic change. It could be that one of the more sedentary insular species might retreat under adverse climatic change, and then be unable to expand again if conditions took a reverse turn, because other factors were then inhibitory. De Korte (1975), in discussing the breeding of the Golden Plover in East Greenland, points out that the long incubation and fledging periods place this species at a disadvantage compared with other waders, and may well make it particularly responsive to the changes in length of breeding season which would result from climatic shifts. The slightly smaller body and egg size of Pluvialis dominica, compared with P. apricaria, could be an adaptation to the shorter breeding season of the higher latitudes which it tends to occupy, since they could involve a reduction of a few days in the combined laying, incubation and fledging periods.

53 110 BIRD STUDY RELATIONSHIPS WITH OTHER WADERS The Golden Plover shares its nesting haunts, in one place or another, with Lapwing, Curlew Numenius arquata, Redshank Tringa totanus, Greenshank, Snipe, Dunlin, Common Sandpiper Tringa hypoleucos and Dotterel. All of these waders exploit the uplands as breeding habitat, but mostly disappear by the end of summer, and spend the winter in contrasting types of biotope. All are largely invertebrate feeders, utilising a seasonal availability of food on their nesting grounds, and it is interesting to consider how they relate to each other in adaptation to different food niches or in the overlap between these. Common Sandpiper and Greenshank depend on rivers, lakes and pools for their food, obtaining it from the stony or sandy margins of these open waters. Although Golden Plovers may feed in these habitats at times, they are unlikely to compete significantly for food with the other two species. Common Sandpiper and Greenshank evidently take a great deal of aquatic or semi-aquatic invertebrate food, whereas the Golden Plover feeds mainly on terrestrial organisms. On the lower moorlands of the northern and western Highlands, Golden Plover and Greenshank are characteristic associates, nesting in an identical range of sites on flow bogs and drier, stony moraines. Curlew, Snipe and Redshank are long-billed probers which on the whole seek wetter places on the moors for feeding than do Golden Plovers. The Curlew is another characteristic bird of the northern and western moorlands but, while overlapping considerably with the Golden Plover, it is much more typical of marginal hill land than the latter, and has shown widespread colonisation of many lowland habitats in recent decades. Conversely, it does not normally breed within the montane zone, and has a mean altitudinal range lower than that of the plover. The Snipe has important upland niches, but again few pairs nest above 610 m, and the main populations belong even more to the lowlands. The same may be said of Redshanks, except that they are seldom more than sparse upland nesters and are absent from much of the hill country. Dotterels are very typically associated with Golden Plovers in their breeding haunts, though these are confined to the montane zone and mostly lie above 825 m. Once settled to nesting, Dotterels appear to live largely at these high

54 BREEDING OF THE GOLDEN PLOVER 111 elevations and are seldom seen on lower moorland. This species would seem more likely than any other wader to overlap significantly, if not completely, with the Golden Plover in its food, and thus to be a potential competitor. The numbers of the Dotterel are, however, usually too small to offer appreciable competition at a population level, and high montane habitats may be underutilised by Golden Plovers for other reasons. D. Nethersole-Thompson (in Um) has seen these two species nesting within 9 m of each other. The Dunlin is perhaps the most constant wader associate of the Golden Plover, from the low-level flowes to the high plateaux at 1,000 m or more. Their curious 'Plover's Page' relationship is general and well known, though its meaning is still obscure. Unlike the plover, though, the Dunlin seldom forsakes the uncultivated ground to feed, and it seems to have a particular dependence on pools or the margins of moorland tarns and rivers for food. Most inland Dunlin breeding haunts are blanket bogs with pools or Sphagnum-filled hollows, or high plateaux and ridges with shallower peaty ground also containing pools. From its smaller size, too, the Dunlin would seem to draw on a rather different food spectrum than the Golden Plover. Holmes (1970) found that in Alaska, the breeding density of Dunlins was especially high on sea-level brackish tundra marshes containing numerous pools which supported large populations of chironomids, the principal food. Interestingly, the Golden Plover does not share the Dunlin's other important British breeding habitats, the salt-marshes of the more northern estuaries and the marshy parts of the shell-sand machair of the Outer Hebrides. The Dunlin has declined substantially as a salt-marsh nester during the last 50 years, and perhaps on the machair too, but on the latter it still occurs in favoured areas at far higher density than in any of its acidic moorland and mountain haunts. This is another good example of correlation between wader breeding density and fertility of the nesting/feeding grounds. On some British breeding grounds, the Dunlin appears to show marked fluctuations in numbers from year to year, perhaps connected with periodic drying-out of feeding places in some seasons. Finally, the Lapwing is an associate of the Golden Plover only to a limited extent in the breeding season, though the two are typically in company in their winter flocks. In many respects, the Lapwing is the complete counterpart of the plover, as a bird of a wide range of farmland, both arable and pasture, in the lowlands. In the hill country, it is noticeably a bird of the better pastures, especially favouring those on limestone or other basic rocks, and for the most part it eschews the acidic moorlands beyond the limits of the upper farms, though in a few places small groups nest with Golden Plovers on blanket bog or even on high acidic summits with Dotterel ground. There is a converse overlap in breeding between the two species on the Mallerstang limestone area, which the 18 pairs of Golden Plovers share with about twice that number of Lapwings, nesting in exactly the same habitat and, inevitably, in close proximity. There appears to be some interaction between the two species, which must draw on the same food supply, and may therefore be in competition. They nevertheless continue to co-exist on this area, and the Golden Plovers have maintained the more constant population (p. 103).

55 112 BIRD STUDY Though my observations are inconclusive, the Golden Plover appears to suffer heavier egg predation than the Lapwing on this shared nesting haunt. The plovers and their eggs may be less cryptically matched on the short grass turf hummocks, compared with the typical moorland habitats, and the species has not evolved the non-cryptically coloured Lapwing's aggressive defence of its nest against bird predators. It has a variety of attraction/distraction displays, but these are more likely to be successful against ground predators. The high density of nests probably also makes them relatively easier for predators to find and, combined with the even higher density of Lapwings' nests, may encourage predators to concentrate on the particular area. Egg predation could thus be density-dependent, with the Golden Plover also suffering disadvantage through its association with the Lapwing. This contrasts with the findings of Goransson et al. (1975) who concluded that the aggression of nesting Lapwings and Curlews significantly reduced predation by Hooded Crows and Herring Gulls Larus argentatus on artificial nests in the vicinity, compared with an adjoining area lacking these waders. Although the limestone grassland is optimum habitat for breeding plovers as regards food value, it may therefore be suboptimal in relation to predation/vulnerability pressure, leading to selection against expansion of nesting on this type of habitat. This relationship is interesting, in that both Lapwing and Golden Plover are successful species, and show in certain places a tendency to attempt expansion of their ecological adaptations, in terms of breeding habitat. For both species, these attempts to colonise each other's habitat seem unsuccessful, perhaps not through direct competition as much as through lack of critical adaptations to the new habitat. Their changed habits persist in a few places, but they do not spread and become built into the populations more widely. Sauer (1962) found the Asiatic Golden Plover to be a relatively aggressive species, behaving territorially to its own kind, and to a variety of other bird types, but especially Turnstones Arenaria interpres which shared the same biotope. This species was not tolerated near the nest site, but little hostility was shown towards Dunlins. The Arctic Fox was usually treated to vigorous attraction/ distraction displays, whilst winged predators seemed to produce greater aggression in flight than is shown by British Golden Plovers. Sauer stressed the survival value of the Asiatic species' cryptic patterns and colours, in the adult, the chick, the egg and even the nest. In general, the Golden Plovers breeding in high latitudes are unlikely to have the advantage of repeat nestings to make good predation or other nest losses, because of the shortness of the far northern summer. The British bird, with its extended breeding season, easily accessible wintering grounds, and even local freedom from predation, thus lives in a relatively congenial environment compared with its Arctic relatives. CONSERVATION Although the Golden Plover is less numerous in many districts than formerly, its present status gives no grounds for anxiety. It has proved adaptable in its lowland wintering grounds in the face of agricultural changes involving much conversion of grassland to arable crops, and has so far shown no sign of being adversely affected by pesticides. Dr Norman Moore has, however, drawn my attention to the potential dangers for certain birds, including waders, which feed on arable land, from the use of new and highly toxic carbamate nematocides in granule

56 BREEDING OF THE GOLDEN PLOVER 113 form. There is likely to be further loss of breeding habitat through afforestation and agricultural reclamation of moorland, but a considerable proportion of the Golden Plover's nesting haunts lie on ground which is climatically intractable and is protected against the likelihood of such land-use changes. Even if more grouse moors become converted entirely to sheep walks, and more of the remaining bogs are drained, there will in the foreseeable future continue to be a very large extent of suitable breeding habitat available to the species. If it is true that numbers can be reduced by heavy nest predation, then it is possible that the land-use changes which promote increase in the predators concerned may be a more potent factor for future population decline. It might then come about that high to very high breeding densities would become increasingly difficult to maintain, and while the species would remain widespread, its numbers might become sparser overall. As long as grouse moors are maintained, the Golden Plover will probably continue to breed at high density locally. Increasing recreational use of the hills is unlikely to have a significant effect on numbers, since the species resorts mainly to featureless moorland and bogs which attract relatively few people, who mostly keep to well-defined tracks. ACKNOWLEDGEMENTS I owe much to Ernest Blezard for encouraging my interest in Golden Plovers and for valuable insight into nest-finding; and I am indebted to Dorothy Blezard for giving me her late husband's careful analyses of food of this species. Desmond Nethersole- Thompson has given freely of his great knowledge of the Golden Plover and I am grateful for his encouragement, stimulating discussion of problems, and helpful additions to this paper. My thanks are due to J. C. Coulson for information and comments on the food potential of different habitats; to I. Mitchell for his special help on breeding distribution and numbers; to W. Gibbs for an account of his unique Peak District experience; and to I. Newton for helpful discussion. I acknowledge the kindness of J. T. R. Sharrock, the British Trust for Ornithology and the Irish Wildbird Conservancy for permission to use the distribution map to be published in the forthcoming Atlas of Breeding Birds in Britain and Ireland. I am grateful also to R. Morgan and the British Trust for Ornithology for use of the Nest Record data for the Golden Plover. The map of breeding habitat was prepared by G. Ledgard and I thank her and R. A. Fenton, both of the Nature Conservancy Council, for allowing me to publish it. Ian Prestt gave me valuable criticism of the final draft. The following have given me information on distribution and numbers, and I much appreciate their help : A. F. Airey, E. K. Allin, A. Archer-Lock, P. Aspin, C. Badenoch, M. E. Ball, A. Barnard, J. Bates, H. M. S. Blair, E. Blake, W. Brotherston, G. Bundy, A. W. Colling, W. M. Condry, M. Cowlard, A. Currie, East Lancashire Ornithologists' Club, T. K. Edwards, P. Evans, I. Findlay, N. Foster, H. Galbraith, M. Garnett, A. Gladwin, C. J. Gent, R. G. Gibbs, D. A. Goode, A. G. Gordon, N. J. Gordon, R. W. Grainger, F. Hamilton, J. M. Harrop, C. J. Henty, D. Holmes, P. Holness, E. G. Holt, D. Humphrey, J. F. Hunt, D. Jenkins, A. J. Kerr, R. R. Lovegrove, D. MacDonald, R. Mackechnie, A. McArthur, J, Martin, C. J. Mead, N. C. Morgan, H. Morrey Salmon, J. Parkin, R. Parr, I. D. Pennie, A. D. K. Ramsay, M. Rawes, J. Richardson, H. Robb, If. Sharrock, R. Stokoe, D. Summers-Smith, W. Thompson, B. Thompson, P. Thompson, R. Thompson, R. Tweddle, A. Walker, K. Walton, E. Ward, A. Watson, A. D. Watson, D. Welch, N. Westwood, G. Williams and P. Wormell. Finally, I thank C. Durell, E. Gordon, J. Mitchell, W. Thompson and A. D. Watson for their help and companionship in the field; M. Fullwood for help with references, and M. J. Chan-Mo for assistance with the preparation of this paper.

57 114 BIRD STUDY SUM MARY 1. With the Red Grouse, the Golden Plover is par excellence the bird of the northern and western moorlands of Britain, showing an increasingly continuous breeding distribution and descent of lower breeding limits with distance north. It has shown substantial decline in many districts during the last 100 years, but not synchronously or in any clear geographical pattern. The total breeding population of the British Isles is estimated now at c. 30,500 pairs. 2. The favourite nesting habitat is on flat or gently sloping ground with submontane ( m) blanket bog, wet heath, heather moor and acidic grassland; and montane (610-1,100 m) blanket bog, dwarf shrub heath, acidic grassland and moss/ lichen heath. Locally in the Pennines species-rich limestone grassland is a noteworthy but atypical nesting habitat, and in most districts nearly all nests are on vegetation overlying acidic, base-poor substrata. 3. Difficulties of nest finding vary according to the nature of the ground, breeding density and behaviour of the incubating birds. Cold searching is profitable only at very high density, and watching birds back to the nest is not usually a rewarding technique either. The best method is to watch carefully for the incubating bird flying from the nest as one walks across the moor. In areas with above moderate density, the regular spacing of nests helps to narrow the search. 4. The onset of laying is delayed by severe weather, but usually begins early in April and is complete by the end of April on lower moors. High altitude pairs are up to two or three weeks later. The time-spread of nesting is wide in many areas, through loss of earlier clutches and the laying of repeats, but most nesting grounds are deserted by the end of July. Mean clutch size was 3.84 eggs for 394 full clutches, and mean brood size at hatching was estimated at c. 3.5 chicks. 5. Food of the Golden Plover is largely invertebrate, especially earthworms, insects, arachnids and molluscs, but some vegetable matter is usually taken, either deliberately or incidentally. Although the species typically breeds on barren moorland, during the incubation period it utilises the most fertile areas within reach of the nest. The sexes share incubation and the off-duty bird flies to enclosed grassland of the upper farms or to especially good feeding habitat on the moor. After hatching, both chicks and parents appear to depend on the nesting area for food, though they may move some distance from the nest site. 6. Variations in breeding density are considerable and are influenced partly by topographic suitability for nesting. On suitable terrain there is frequently a correlation between breeding density and soil fertility, evidently reflecting a response to feeding value of the habitat. The present highest known density is on 110 ha of Pennine limestone grassland, with 16 pairs /km 2. Blanket bog and heather moor within reach of rich grassland for feeding, in parts of the Pennines, Southern Uplands and eastern Highlands have 5-7 pairs /km2. Density is lowest on the barren moorlands of western and northern Scotland (less than 1 pair /km 2). 7. Except at the lower densities, Golden Plovers disperse themselves territorially on the nesting grounds, and nests are often evenly spaced out. The strength of this dispersion appears to be related to the food carrying capacity of the habitat. Yet in many areas the birds resort to communal feeding places away from the nest during incubation, and only depend on the nesting area for food after hatching, by which time territorialism has broken down. It is suggested that behaviour which spaces pairs out in relation to carrying capacity of habitat is advantageous to survival of the young and so is subject to individual selection, but the effect is delayed between one breeding season and the next. 8. Post-glacial deforestation greatly extended the area available to nesting Golden Plovers, but recent afforestation of moorland has reversed this trend and has caused decline, notably in the Cheviots and Southern Uplands. Long-term decline in the western Highlands and Islands may reflect reduced productivity of moorland through extractive land use. The declines reported from most districts probably result from factors not yet fully understood, but evidently operating during the

58 BREEDING OF THE GOLDEN PLOVER 115 breeding season. Predation on nests may be an important influence, locally, and the possible effects of climatic changes are discussed. Yet, during the last years, the breeding populations on two main study areas in the Pennines, and a third in the Southern Uplands, have remained remarkably constant. 9. The Golden Plover shares nesting haunts in one place or another with other Lunicolae: Lapwing, Curlew, Redshank, Greenshank, Snipe, Dunlin, Common Sandpiper and Dotterel. These species are differentiated partly by food requirements; their respective niches, and their overlap with Golden Plovers are discussed briefly. Only the Lapwing appears to be a significant competitor, but is usually separated from the Golden Plover in nesting habitat. 10. The Golden Plover appears to present few conservation problems. It is adaptable in its lowland wintering farmland habitats, and a substantial part of the population breeds on uplands which are unlikely to be afforested or developed agriculturally. The species could decline further if grouse moors became extensively converted to sheep walks, but it seems unlikely to be at serious risk to increasing recreational pressure. REFERENCES BALFOUR, E Orkney Birds, Status and Guide. Stromness. BANNERMAN, D. A The Birds of the British Isles. Vol. 10. Edinburgh and London. BAXTER, E. V. and L. J. RINTOUL The Birds of Scotland. Edinburgh. BLEZARD, E Lakeland Natural History. Trans. Carlisle Nat. Hist. Soc. Vol. VII. BROWN. L. H. and A. WATSON The Golden Eagle in relation to its food supply. Ibis, 106, BUCKLEY, T. E. and J. A. HARVIE-BROWN A Vertebrate Fauna of the Orkney Islands. Edinburgh. BURTON, P. 3. K Feeding and the feeding apparatus in waders. British Museum (Natural History). London. COULSON, J. C Observations on the Tipulidae (Diptera) of the Moor House Nature Reserve, Westmorland. Trans. R. Ent. Soc. London. 111: COULSON, J. C The biology of Tipula subnodicornis Zeiterstedt, with comparative observations on Tipula paludosa Meigen. J. Anim. Ecol., 31 : COULSON, 3. c. and J. B. WHITTAKER (in press). The ecology of moorland animals. In The Ecology of some British Moors and Mountain Grasslands. (Ed. by 0. W. Heal and D. Perkins). Berlin. DE KORTE, j Golden Plover Pluvialis apricaria breeding in Jameson Land, East Greenland. Dansk Orn. Foren. Tidss., 69: FORREST, H. E The Vertebrate Fauna of North Wales. London. GILBERT, H. A. and A. BROOK Secrets of Bird Life. London. GORANSSON, G. J. }CARLSSON, S. G. NILSSON, and D. ULFSTRAND Predation on birds' nests in relation to antipredator aggression and nest density: an experimental study. Oikos, 26: GORDON, 5 R. HEWSON, D. NETHERSOLE-THOMPSON, R. TEWNION and A. WATSON "Northern" Golden Plovers in northern parts of Scotland. Scott. Nat., 69: GRAHAM, R Bird notes from the Solway, Pennines and Lakeland, North-west Nat. 9: GRAY, R The Birds of the West of Scotland. HARVIE-BROWN, J. A A Fauna of the Tay Basin and Strathmore. Edinburgh. HARVIE-BROWN, J. A. and H. A. MACPHERSON A Fauna of the North-west Highlands and Skye. Edinburgh. HAVILAND, M. D Notes of the breeding habits of the Asiatic Golden Plover. Brit. Birds, 9: HOLMES, R. T Differences in population density, territoriality, and food supply of Dunlin on Arctic and Subarctic tundra. In: Animal Populations in Relation to their Food Resources (Ed. A. Watson). Brit. Ecol. Soc. Symposium No. 10. Aberdeen. HOUSTON, W. W. K. Carabidae (Coleoptera) from two areas of the Northern Pennines. Ent. Month. Mag., 107:1-4. KEARTON, R Our Rarer British Breeding Birds. London. LEA, D. and W. R. P. BOURNE The Birds of Orkney. Brit. Birds, 68: , MACKENZIE, A Hundred Years in the Highlands. London. MCVEAN, D. N. and J. D. LOCICIE Ecology and Land Use in Upland Scotland. Edinburgh. morton, R. K. and N. J. WESTWOOD Some effects of agricultural change on the English avifauna. Brit. Birds, 67: NETHERSOLE-THOMPSON, n. and c. NETHERSOLE-THOMPSON. 1961(a). The breeding behaviour of the British Golden Plover. In Bannerman, D.A. The Birds of the British Isles. Edinburgh and London. Vol. 10, NETHERSOLE-THOMPSON, D. 1961(b). The breeding behaviour and breeding biology of the Lapwing. In Bannerman, D. A. The Birds of the British isles..edinburgh and London. Vol. 10, NETHERSOLE-1110MPSON, D The Snow Bunting. Edinburgh and London. NETHERSOLE-THOMPSON, D The Dotterel. London.