ISSN: (Print) (Online) Journal homepage:

Transcription

1 Bird Study ISSN: (Print) (Online) Journal homepage: Population estimates, trends and habitat associations of breeding Lapwing Vanellus vanellus, Curlew Numenius arquata and Snipe Gallinago gallinago in Northern Ireland in 1999 I.G. Henderson, A.M. Wilson, D. Steele & J.A. Vickery To cite this article: I.G. Henderson, A.M. Wilson, D. Steele & J.A. Vickery (2002) Population estimates, trends and habitat associations of breeding Lapwing Vanellus vanellus, Curlew Numenius arquata and Snipe Gallinago gallinago in Northern Ireland in 1999, Bird Study, 49:1, 17-25, DOI: / To link to this article: Published online: 29 Mar Submit your article to this journal Article views: 630 View related articles Citing articles: 14 View citing articles Full Terms & Conditions of access and use can be found at Download by: [ ] Date: 29 December 2017, At: 13:20

2 Bird Study (2002) 49, Population estimates, trends and habitat associations of breeding Lapwing Vanellus vanellus, Curlew Numenius arquata and Snipe Gallinago gallinago in Northern Ireland in 1999 IAN G. HENDERSON 1 *, ANDREW M. WILSON 1, DAVID STEELE 2 & JULIET A. VICKERY 1 1 British Trust for Ornithology, The Nunnery, Thetford, Norfolk, IP24 2PU, UK and 2 Royal Society for the Protection of Birds, Belvoir Park Forest, Belfast, BT8 4QT, UK Capsule Declines in the breeding populations of Snipe, Lapwing and Curlew were recorded between April and June 1999 and compared with previous estimates in Aims To compare populations of non-coastal breeding waders between 1987 and Methods In km 2 km square tetrads, observers recorded the number of breeding pairs of waders and habitat details on 1: scale maps. Results A significant decline of c. 60% for Lapwing and Curlew, and a non-significant decline of 30% for Snipe was recorded over 12 years. Concentrations of these species were found in County Tyrone, but Counties Antrim, Down and Armagh supported few breeding pairs of any species. Very few pairs of any species were recorded on improved grassland despite its widespread availability. Conclusion A successful conservation strategy for these species must address the wider countryside and not just key sites. Intensive pastoral farming in upland and lowland areas and activities such as drainage and peat extraction will further reduce the suitability of open habitats for these wader species. Breeding waders such as Curlew Numenius arquata, Snipe Gallinago gallinago and Lapwing Vanellus vanellus have exhibited marked population declines or range contractions across much of western Europe (e.g. Tucker & Heath 1994, Hagemeijer & Blair 1997). These declines have been attributed to land drainage, intensification of arable and grassland farming, afforestation and increased predation (Hutchinson 1989, Pain & Pienkowski 1996, Grant et al. 1999, Wilson et al. 2001). Breeding Lapwing in England and Wales, for example, have declined by over 49% in 11 years (i.e ; Wilson et al. 2001), exceeding the threshold rate of decline for classification as a Red Listed Bird of Conservation Concern (Gibbons et al. 1996). These declines were steepest in western pastoral areas (Wales and southwest England) and were attributed to increased intensification of arable and grassland agriculture and the loss of spring tillage and mixed farming enterprises (Shrubb & Lack 1991, Wilson et al. 2001). Atlas studies in and have *Correspondence author. ian.henderson@bto.org suggested declines in breeding Curlew across the midlands, southwest England and Wales and southeast Ireland, and Snipe throughout southern England, Wales and the southern half of Ireland (Gibbons et al. 1993). Northern Ireland has historically supported important populations of breeding waders (Partridge & Smith 1992). High annual rainfall and saturated gley/peat soils combine with steep slopes to reduce the potential for arable intensification. Only 6% of the land area is cultivated while 85% is occupied either by improved Rye Lolium perenne-multiflorum grassland (34%), permanent pasture (26%) or uncultivated land (25%, i.e. mainly rough grazing; Cooper et al. 1997). Damp grasslands and species-rich pastures have provided key breeding habitats for Lapwing, as well as Snipe on wetter swamp and fen habitats, whilst extensive peatlands formed the stronghold of breeding Curlew (Partridge & Smith 1992). In 1987, Curlew and Snipe populations in Northern Ireland were estimated to represent 10% and 14.3% respectively of the British and Irish population (Table 1; Partridge & Smith 1992), in addition to over 5000 pairs of 2002 British Trust for Ornithology

3 18 I.G. Henderson et al. Table 1. Population estimates for breeding waders in Northern Ireland in 1987 and Population estimate Population (±95% CL) change Species (%) Lapwing 5250 ( ) 1771 ( ) 66 Curlew 5000 ( ) 2091 ( ) 58 Snipe 5725 ( ) 3993 ( ) data from Partridge & Smith (1992); 1999 data from this survey. Lapwing. However, growing evidence suggests that these important breeding wader populations are declining from the loss of peatlands, mechanized cutting, the drainage of wetland sites, afforestation and increasingly intensive grassland management (Partridge & Smith 1992). Despite recommendations to protect key wader sites (Partridge & Smith 1992), such as the shores of Lough Neagh and Lough Erne, declines in Curlew, Snipe and Lapwing have continued (Stanbury et al. 2001). The majority of Curlew, Snipe and Lapwing are also widely dispersed away from key areas breeding at low density (Partridge & Smith 1992). These populations have not been well monitored and their current status in Northern Ireland is unknown. It is known that across Northern Ireland, over 18% of species-rich grasslands and fen meadow have been lost to agricultural conversion or afforestation between 1989 and 1993, and peatlands have become increasingly fragmented and degraded (Cooper et al. 1997). These changes are likely to have had a negative impact on habitat quality for breeding waders. Concerns for breeding wader populations in Northern Ireland highlighted the need for a large-scale survey to reassess the current status of three key species, Lapwing, Curlew and Snipe. Here we report on the results of an extensive survey of breeding waders in Northern Ireland and we compare population estimates with a previous survey carried out in 1987 (Partridge & Smith 1992) with reference to more recent work (e.g. Grant et al. 1999). METHODS Survey coverage and fieldwork procedure In total, 147 tetrads (2 km 2 km squares) were selected for the survey across Northern Ireland. This selection process was identical to that used in 1987 when each tetrad was randomly selected from within every land-based 10 km square in Northern Ireland (Partridge & Smith 1992). Observers were asked to visit each tetrad twice, the first visit between 9 April and 17 May, when the majority of breeding waders were expected to be on territories, and the second visit between 18 May and 21 June Each visit (lasting four or five hours) required observers to cover on foot as much of their designated tetrad as possible (the field by field method) recording all waders seen or heard. Inaccessible areas were viewed from lanes or vantage points. As in 1987, observers were particularly thorough in searching areas of potential wader habitat, such as wet flushes or unimproved grassland, in order to improve the detection rate of Snipe (Partridge & Smith 1992). Expansive areas of open moorland or grassland were covered by walking parallel transects approximately 200 to 250 m apart. Numbers and location of all waders observed were recorded on a 1: site map of each tetrad. The habitat type that each wader was found using was recorded, with observers using defined habitat codes for consistency (see Appendix). For the purpose of the survey (and comparable to 1987), breeding pairs were identified as incubating birds, singles or pairs of birds encountered away from sea shorelines or estuaries, or individuals displaying, mobbing or assuming other territorial behaviours. Flocks of more than four waders were recorded as individuals rather than breeding pairs and were assumed to be non-breeding birds (Cramp & Simmons 1983). Observers estimated the proportion of each 1-km square, within their tetrad, that was covered by one of a number of different habitat types. Habitat categories distinguished between wetland habitats (i.e. blanket bog/mire or marsh and fen), non-arable habitats (i.e. improved grassland, unimproved grassland, upland rough grassland and heather moorland). Observers indicated on their maps, areas of standing water, including lakes, bog and marsh, and flooded or water-logged ground and areas of woodland or scrub. Analysis Population estimates were calculated as in the previous survey to ensure comparability. Thus, the total number of pairs of each species recorded in the tetrad survey and the area of land covered (from the proportion of each tetrad covered by the observer) was used to calculate both mean species density and total breeding population for Northern Ireland, effectively by multi-

4 Breeding waders in Northern Ireland 19 plying up the birds on the area covered (369 km 2, or 1/37th, see below) to the total land area available. Correction factors were not used for Snipe although adjusted figures are discussed. Confidence limits for population estimates were calculated using a proportionately random bootstrap procedure that resamples the count data (all habitats combined for each species, 999 times) to produce population estimates on each iteration, resulting in the calculation of a mean population estimate plus 95% confidence limits (CL) for each species (Greenwood 1991). A relative measure of habitat preference (not absolute numbers for each habitat in Northern Island) was calculated for seven habitat categories (i.e. bog/mire, marsh/fen, heather moorland, upland rough grassland, unimproved grassland, improved grassland and combined arable habitats. This measure of preference was calculated using the formula: Preference index = 100 log 10 (observed pairs/ expected pairs from the area of the habitat available) Positive index values represent selection for a particular habitat and negative values indicate avoidance of a particular habitat. Where no pairs were found on a habitat type the index value is negative infinity. The expected values were calculated by assuming that pairs were distributed randomly across habitats in direct proportion to the available area of each habitat. RESULTS Tetrad distribution and coverage A total of 106 tetrads was covered in 1999 compared to 147 in In all, 62.5% of tetrads were visited twice in The mean area of coverage within each tetrad in 1999 was 87% with over 90% of tetrads receiving coverage exceeding 75% of the area [distribution as follows: area of coverage (number of tetrads): < 50% (2), 60 70% (8), 75 95% (41), 100% (49)]. Large areas of dense forests and urban areas were considered covered and given a zero count, since the wader species in question were unlikely to have bred in these habitats. Coverage excludes only areas that observers could not access, which may have held waders. The total area surveyed (i.e. 87% of 106 tetrads = 369 km 2 ; 552 km 2 in 1987) represented approximately 2.7% of the total land area of Northern Ireland. In general, surveyed tetrads were evenly distributed across Northern Ireland in 1999, as in 1987, and included extensive coverage of remote areas with potentially high wader densities, in counties Fermanagh and Tyrone, except for one potentially important area in the Antrim Hills, discussed below (Fig. 1). In Fig. 2, we show the 1999 distribution of occupied tetrads of each wader species. Absence implies absence from the tetrad but not necessarily absence from the 10-km square represented by the tetrad. Population estimates and habitat association The survey results suggest widespread population declines across Northern Ireland for all three wader species (Table 1). For Lapwing, Curlew and Snipe respectively, 90%, 96% and 100% of occupied tetrads contained some wetland habitat, comprising standing water, marshland or flooded grassland, compared with only 46% of tetrads in which no waders were recorded (e.g. for Lapwing, χ 2 1 = 11.8, P < 0.001). a b Figure 1. (a) An outline map of Northern Ireland showing key features. (b) The distribution of survey tetrads in Northern Ireland that were visited in 1987 and 1999 ( ) and those visited in 1987 only ( ).

5 20 I.G. Henderson et al. a Lapwing b Curlew c d e Snipe Figure 2. The abundance and distribution of Lapwing (a), Curlew (c) and Snipe (e) in Northern Ireland: (, no pairs recorded in the tetrad;, one to two pairs recorded;, three to four pairs recorded;, more than four pairs recorded). The change in the presence or absence of breeding birds between 1987 and 1999 is also presented for Lapwing (b), Curlew (d) and Snipe (f):, pairs recorded in 1987 but not in 1999;, pairs recorded in 1999 but not in f Lapwing The 1999 population estimate of 1771 pairs (declining from 1.4 pairs per tetrad in 1987 to 0.44 pairs per tetrad in 1999) represents a 66% decline across Northern Ireland (Table 1) for Lapwing. The majority of breeding Lapwing were recorded west of Lough Neagh particularly in Co. Tyrone (Fig. 2a). In counties Antrim and Down, pairs were recorded in only four (10%) out of 38 tetrads covered. This compares to 70% occupancy in 1987 (χ 2 1 = 15.5, P < 0.01), and thus the change map (Fig. 2b) shows that 70% of tetrads occupied in 1987 (and surveyed again in 1999) were reported without pairs in 1999, compared to 48% for the rest of Northern Ireland. Lapwing populations in the Sperrin Mountains and Lough Erne also appear to have undergone declines (Fig. 2b). Between habitat types, this species showed a positive selection for marshland, upland rough grassland, unimproved grassland and arable habitats, but fewer birds than expected (according to the area available) were found on upland bogs/mire, heather moorland (where no birds were recorded) and improved pasture (Fig. 3).

6 Breeding waders in Northern Ireland 21 Curlew The 1999 population estimate of just over 2000 breeding pairs of Curlew represents a 58% decline since 1987 (Table 1) across Northern Ireland. Mean breeding density declined from 1.37 pairs per tetrad in 1987 to 0.52 pairs per tetrad in In 1999, the highest breeding concentrations were recorded from Co. Tyrone with a notable absence within the Lough Erne basin, previously an area of high density (Fig. 2c). In 1999, breeding pairs were recorded from only two out of 20 tetrads (10%) in Co. Antrim compared to 80% of 34 tetrads in 1987 (χ 2 1 = 9.1, P < 0.01). Thus the change map shows that 94% of tetrads in Antrim occupied in 1987 (and surveyed again in 1999) were reported without pairs in 1999, compared to 62% for the rest of Northern Ireland (Fig. 2d). There was little change for Co. Down, where breeding Curlew were also scarce in 1987 (Partridge & Smith 1992). There was relatively strong selection for bog/mire or unimproved grassland (most occupied tetrads in 1999 included areas of these) but Curlew were scarcer than expected on improved grassland, upland rough grassland and arable land according to availability of these habitats (Fig. 3). Preference index 100 a Infinity or absence 100 b Infinity or absence 100 c Infinity or absence BO MA HM RG UG IG A BO MA HM RG UG IG A BO MA HM RG UG IG A Habitat Figure 3. Relative habitat preferences by (a) Lapwing (n = 46 pairs), (b) Curlew (n = 61 pairs) and (c) Snipe (n = 101 pairs), based on the index: 100 H log 10 (observed pairs/expected pairs). Habitats: BO, bog/mire; MA, marsh/fen; HM, heather moorland; RG, upland rough grassland; UG, unimproved grassland; IG, improved grassland; A, arable land. As indicated, where no birds were seen, the index equals infinity. Snipe Snipe were the most abundant and widespread breeding wader in Northern Ireland in However, a population estimate of c pairs, and a mean density of 0.97 pairs per tetrad (1.58 pairs per tetrad in 1987), represents a decline of around 30% since 1987 (Table 1), although the decline was not statistically significant (z = 2.37, ns). Snipe occurred in all counties and were most numerous west of Lough Neagh (45% of tetrads; Fig. 2e), but in Co. Antrim and Co. Down they were recorded in only 16% of tetrads surveyed. The change map (Fig. 2f) shows that in these two counties there was a decline in 60% of tetrads that were surveyed in both 1987 and 1999 (compared to 41% of tetrads across the rest of Northern Ireland; χ 2 1 = 6.2, P < 0.01). Snipe preferred fens and marshes, with very few birds being recorded in other habitats, and with a notable avoidance of improved grassland, arable land and upland rough grassland (Fig. 3). DISCUSSION Previous surveys of breeding Lapwing, Curlew and Snipe in Northern Ireland on key wader sites such as Loughs Erne and Neagh, the Tyrone peat bogs (Fairywater Valley) and the Blackwater catchment (Tyrone/Armagh) have reported declines of 64%, 55% and 29% respectively between 1986 and 200 (Stanbury et al. 2001). The results presented here provide strong evidence of population declines and range contractions for Lapwing and Curlew and a suggestion of a population decline for Snipe, throughout Northern Ireland in the last 12 years. Although fewer tetrads were covered in 1999 (106), coverage was good in remote areas of Co. Fermanagh and Co. Tyrone where high densities of breeding waders are traditionally recorded (Gibbons et al. 1993).

7 22 I.G. Henderson et al. Indeed, population changes between 1987 and 1999 may be conservative because the majority of uncovered tetrads in 1999 supported low wader densities in earlier surveys (Partridge & Smith 1992, Gibbons et al. 1993). Their inclusion would therefore have lowered the population estimates for 1999 still further. Also, a higher proportion of tetrads was surveyed twice in 1999 (62%) than in 1987 (when Most sample sites were visited once only : Partridge & Smith 1992), increasing the likelihood that waders would be detected in 1999 and, if anything, making the population changes discussed here conservative. Areas that were under-represented by the survey included parts of the Mourne Mountains, southern Co. Armagh, north-central Co. Londonderry and the Antrim Hills. Apart from the Antrim Hills, all of these localities supported low densities of Snipe, Curlew and Lapwing in 1987 as well as during the Breeding Atlas period (Gibbons et al. 1993). Their inclusion in the current survey would have contributed very few pairs to the population totals (Table 1) and would probably have reduced further the overall mean population estimate of each wader species for Breeding Lapwing, Curlew and Snipe were found in the Antrim Hills tetrads in 1987, and had they been covered in the 1999 survey they may have increased the population estimate for each species. Taking all missing tetrads into consideration, the population estimates given in Table 1 are considered likely to approach those for complete coverage of Northern Ireland. For Snipe, it is likely that the population totals were underestimated in both 1987 and 1999 for three reasons (see Partridge & Smith 1992). First, surveys based on counts of drumming males are known to underestimate breeding pairs. Green (1985) suggested counts should be multiplied by 1.74 to provide a more accurate figure (that is, 6950 pairs for 1999). Second, an accurate census requires multiple visits to overcome weather-dependent variations in the detectability of displaying birds (Fuller et al. 1983). A larger proportion of tetrads (63%) received a second visit in 1999 than in 1987 (Partridge & Smith 1992), possibly improving detectability in the second survey. Third, peak Snipe display occurs at dawn and dusk (Cramp & Simmons 1983) and only a portion of a tetrad can be covered at these optimal times. In both surveys attempts were made to minimize this error. In 1987, fieldworkers were explicitly recommended to undertake early morning visits while in 1999 at least 65% of visits included some coverage before 09:00 hours or after 17:00 hours. However, the Snipe population decline estimated between the two surveys is consistent with earlier findings from key sites (O Brien & Self 1994, Donaghy & Mellon 1998, Stanbury et al. 2001) and despite the difficulty of accurately surveying this species, we consider that the results presented here represent a real decline. The results also indicate a contraction of range for Lapwing and Curlew within Lower Lough Erne, the Sperrin Mountains, Co. Antrim, Co. Down and south Co. Armagh. In 1999, Lapwing and Curlew were concentrated in the mid- and west of Northern Ireland, in Co. Fermanagh and Co. Tyrone and, for Lapwing only, also in Co. Londonderry. Snipe, in contrast, have remained relatively widespread, albeit at lower densities than Curlew and Lapwing, but with a higher population than both of these species. The causes of range contractions in the east are difficult to ascertain from this survey but habitat preference data presented here and studies elsewhere suggest that a number of factors largely relating to land use changes may be responsible. Breeding Lapwing tended to be associated with agricultural habitats, with losses more evident from the marginal uplands, lowlands and, to a lesser extent, central lowlands in Co. Antrim and from drumlin farmlands in Co. Down (Fig. 3; Gibbons et al. 1993). In contrast, breeding Curlew were associated with bogs and heather moorland and have been lost from uplands, marginal uplands and some marginal lowlands, especially in Co. Antrim (Cooper et al. 1997). These habitat associations suggest that the most important factors affecting the abundance and distribution of these species are likely to be associated with changes in lowland grasslands for Lapwing, and upland rough grassland and peatland habitat for Curlew. Increasingly intensive grassland management has been a feature of agriculture in Northern Ireland in the last 20 to 30 years, and this has reduced arable land to only 6% of the land area (Cooper et al. 1997). Livestock densities have risen by almost 50% in 35 years with sheep numbers rising almost three-fold to 2.5 million (Donaghy & Mellon 1998). To support this number of animals, the pastoral ecosystem has moved towards increasingly intensive grassland management through a regime of drainage, regular cutting and high fertilizer (nitrogen and phosphorus) input (Donaghy & Mellon 1998). Improved rye-grass leys, a habitat that was consistently avoided by Lapwing, Curlew and Snipe, now account for over 30% of the land area of Northern Ireland (Cooper et al. 1997).

8 Breeding waders in Northern Ireland 23 Although the impacts of intensive grassland management on birds are poorly understood, increases in fertilizer inputs and intense grazing or cutting regimes lead to uniform dense swards with low plant species and structural diversity (Vickery et al. 2001, Perkins et al. 2000). A number of studies have shown reduced breeding density or productivity of waders on improved, compared with unimproved, grassland (e.g. Lapwing: Baines 1989, Berg et al. 1992; Curlew: Baines 1988, Valkama et al. 1999; Snipe: Cramp & Simmons 1983). This has been attributed variously to reduced food availability for adults and chicks, increased predation on nests and chicks, and the loss of nests and chicks through trampling by livestock or mechanical cutting. For Lapwing, productivity measured in many European studies has been insufficient to support longterm populations on improved grassland (Peach et al. 1994). The loss of mixed farming practice with nearby fields of older grassland and spring tillage has also been linked to severe population declines (down 49% in 11 years) of Lapwing in Wales and southwest England since 1987 (Wilson et al. 2001). Such mosaics provide preferred nesting (spring till) and chick foraging habitats (unimproved grassland) (Cramp & Simmons 1983) so that the loss of arable land in Northern Ireland may also have reduced the availability of optimal nesting and brood rearing habitats for this species. In 1999, as in 1987, the majority of breeding Curlew were found on bog/mires and unimproved grassland, reflecting a preference for damp and unimproved tussocky terrain and/or tall but sparse swards (Grant 1997, Valkama et al. 1999). Northern Ireland holds 44% of the UK resource of Molina/Shagnum bog (Cooper et al. 1997) but these peatlands are becoming increasingly fragmented and degraded, and in 1994 only 12 15% ( ha) of high quality, species-rich habitat was considered intact (Cooper et al. 1997). At the same time, much marginal hill land has been improved as pasture or lost to afforestation (Donaghy & Mellon 1998). Curlew in Northern Ireland have also been affected by increased predation on clutches and broods by Foxes Vulpes vulpes, Hooded Crows Corvus corone cornix and Lesser Black-backed Gulls Larus fuscus (Grant et al. 1999). This may have been due to less intensive predator control than in the past (Grant et al. 1999). However, additional factors, such as high livestock densities, drainage and habitat fragmentation (Tullinoff & Bergmann 1993, Valkama et al. 1999), may have exacerbated the impact of predators on Curlew productivity. For Lapwing, Curlew and Snipe, an important habitat component is the provision of damp ground within their territory, where invertebrate prey are both abundant and accessible (Cramp & Simmons 1983, Fuller et al. 1986, Berg, 1992). In the present survey, most of the tetrads from which each species was recorded contained areas of standing water (i.e. lakes, fens, boggy or marshy ground), in contrast to unoccupied tetrads. This crude measure of the availability of wet ground was nevertheless consistent with many previous studies that emphasize the importance of a high water table to ensure suitable foraging habitat for breeding waders (e.g. Herbert et al. 1990, Hotker 1991, Berg 1992). Major arterial drainage schemes, designed to prevent seasonal flooding and facilitate agricultural improvement, have affected large areas of wet grassland in Northern Ireland (Donaghy & Mellon 1998). Over the wider countryside, however, well established and longterm activities, such as peat cutting, forestry and grassland improvement operations (Cooper et al. 1997), may have reduced the availability of damp habitats for many breeding waders. Although further research is required to understand the mechanisms underlying these declines, particularly for Snipe, formulating measures for the conservation of populations in the wider countryside is made more difficult by their widespread distribution and low density. Declines of the order of 60% for Lapwing and Curlew over 12 years effectively exceed the threshold of one of the criteria for placing species on the Red List of Species of Conservation Concern (that is, a >50% decline in 25 years; Gibbons et al. 1996) and the results suggest that all three species should be attributed higher conservation status than at present. Effective action will need to consider farmland intensification, upland drainage activities and forest expansion. There is a need to protect remaining areas of wet grassland through agri-environemtal schemes or Area of Special Scientific Interest (ASSI) management agreements. The recently developed Countryside Management Scheme (CMS) in Northern Ireland already includes specific options to manage Lapwing breeding sites, Upland breeding wader sites and Wetland sites (DANI 1999). This new scheme is voluntary and its success will depend on the scale of uptake by land managers. The CMS could provide extensive benefits for breeding waders in Northern Ireland but continued monitoring will be vital to assess its success and provide recommendations for future option developments or modifications.

9 24 I.G. Henderson et al. ACKNOWLEDGEMENTS The authors would like to thank all volunteers who took part in the breeding wader survey, and all landowners who permitted access to their land. We would like to extend our gratitude to Ken Perry (BTO Ireland Officer), Clive Mellon (RSPB) and Anita Donaghy (RSPB) for raising volunteer support for the survey in Northern Ireland, and we thank Nicola Read for data management and secretarial assistance. We also thank Rob Fuller (BTO) and Nick Carter (BTO) for help and advice throughout the project, and Anita Donaghy (RSPB) and Murray Grant (RSPB) and referees for earlier comments on the survey results. The Breeding Wader Survey of Northern Ireland 1999 was commissioned by the Environment and Heritage Service. REFERENCES Baines, D The effects of improvement of upland, marginal grasslands on the distribution and density of breeding wading birds (Charadriiformes) in northern England. Biol. Conserv. 45: Baines, D The effects of improvement of upland, marginal grasslands on the breeding success of Lapwings Vanellus vanellus and other waders. Ibis 131: Berg, C Habitat selection by breeding Curlews Numenius arquata on mosaic farmland. Ibis 134: Berg, C., Lindberg, T. & Kallebrink, K.G Hatching success of lapwings on farmland: differences between habitats and vcolonies of different sizes. J. Anim. Ecol. 61: Cooper, A., Murray, R., & Mcann, T The Northern Ireland Countryside Survey. Environment and Heritage Service, Belfast. Cramp, S. & Simmons, K.E.L The Birds of the Western Palearctic, Vol. 3. Waders to Gulls. Oxford University Press, Oxford. DANI Countryside Management Scheme; Explanatory Booklet. Department of Agriculture for Northern Ireland, Belfast. Donaghy, A. & Mellon, C Fields for the Future: Towards the Conservation of Farmland Birds in Northern Ireland. RSPB, Sandy. Fuller, R.J., Green, G.H. & Pienkowski, M.W Field observations on methods used to count waders breeding at high density in the Outer Hebrides, Scotland. Wader Study Group Bull. 39: Gibbons, D.W., Reid, J.B. & Chapman, R.A The New Atlas of Breeding Birds in Britain and Ireland: T. & A.D. Poyser, London. Gibbons, D, Avery, M., Baillie, S., Gregory, R., Kirby, K., Porter, R., Tucker, G. & Williams, G. 1996: Bird species of conservation concern in the United Kingdom, Channel Islands and the Isle of Man: revising the Red Data list. RSPB Conserv. Rev. 10: Grant, M.C Breeding Curlews in the UK: RSPB Research and Implications for Conservation. RSPB Conserv. Rev. 11: Grant, M.C., Orsman, C., Easton, J., Lodge, C., Smith, M., Thompson, G., Rodwell, S. & Moore, N Breeding suc- cess and causes of breeding failure of curlew Numenius arquarta in Northern Ireland. J. Appl. Ecol. 36: Green, R.E Estimating the abundance of Snipe. Bird Study 32: Greenwood, J.J.D Estimating the total number and its confidence limits. Bird Study 39: Hagemeijer, W.M. & Blair, M.J The EBCC Atlas of European Breeding Birds: Their Distribution and Abundance. Poyser, London. Herbert I.J., Heery, S. & Meredith, C.R.M Distribution of breeding waders in relation to habitat features on the River Shannon callows at Shannon harbour, Ireland, Irish Birds 4: Hotker, H Waders breeding on wet grasslands in the countries of the European Community a brief summary of population sizes and population trends. Wader Study Group Bull. 61 suppl.: Hutchinson, C.D The Birds of Ireland. Poyser, Calton. O Brien, M. & Self, M Changes in the numbers of Breeding Waders on Lowland Wet Grasslands in the UK. RSPB Conserv. Rev. 8: Pain, D.J. & Pienkowski, M.W. (eds) Farming and Birds in Europe: a Common Agricultural Policy and Its Implications for Bird Conservation. Academic Press, London. Partridge, J.K. & Smith, K.W Breeding wader populations in Northern Ireland, Irish Birds, 4: Peach, W.J., Thompson, P.S. & Coulson, J.C Annual and long-term variation in the survival rates of British lapwings Vanellus vanellus. J. Anim. Ecol., 63: Perkins, A.J, Whittingham, M.J., Bradbury, R.B., Wilson, J.D., Morris, A.J. & Barnett, P.R Habitat characteristics affecting use of lowland agricultural grassland by birds in winter. Biol. Conserv. 95: Shrubb, M & Lack, P.C The numbers and distribution of Lapwings V. vanellus in England and Wales in Bird Study 37: Stanbury, A., O Brien, M. & Donaghy, A Trends in breeding wader populations in key areas within Northern Ireland between 1986 and Irish Birds 6: Tucker, G.M. & Heath, M.F Birds in Europe: Their Conservation Status. Birdlife Conservation Series No. 3. Birdlife International, Cambridge. Tullinoff, R. & Bergmann, H-H Curlew habitats (Numensius arquata) in Lower Saxoma preferred and avoidance structures of agricultural landscapes. Vogelwarte 37: Valkama, J., Currie, D. & Korpimoki, E Differences in the intensity of nest predation in the curlew Numenius arquata: A consequence of land use and predator densities? Ecoscience 6: Vickery, J.A., Tallowin, J.T, Feber, R.E., Asteraki, E.J., Atkinson Fuller, R.J. & Brown,V Effects of grassland management on birds and their food resources, with special reference to recent changes in fertiliser, mowing and grazing practices in lowland Britain. J.Appl. Ecol. 38: Wilson, A., Vickery, J.A. & Browne S.J Numbers and distribution of Northern Lapwings Vanellus vanellus breeding in England and Wales in Bird Study 48: (MS received 22 August 2000; revised MS accepted 14 May 2001)

10 Breeding waders in Northern Ireland 25 APPENDIX Habitat definitions used in the field survey are presented below. Note that for analytical purposes (see Fig. 3) some of the arable categories were combined as indicated by the code at the end of each descriptive passage. Improved grassland: Pasture for grazing or silage mainly in lowland areas that has probably been cultivated within the last five years. The sward tends to look uniform and freshly green due to high fertilizer inputs (IG in Fig. 3). Unimproved lowland grassland: Fields of long-term grazing pasture or hay meadows in lowlands and on hillsides that have not been obviously improved. The sward may be short but uneven with occasional tussocks and may be both patchy and variable in colour (UG in Fig. 3). Upland rough grazing: Unimproved grassland on open hillsides, upland areas and moorland, often interspersed with heather or bracken and rarely confined to small fields (RG in Fig. 3). Heather moorland and heath: Continuous expanses of open heather moorland or heathland, sometimes with sparse gorse scrub or other shrubs (dense areas of gorse are categorized as woodland/scrub) (HM in Fig. 3). Marsh swamp and fen: Permanently flooded ground in lowland areas by lakes, pond or rivers, including marginal and emergent vegetation (iris beds or rushes) (MA in Fig. 3). Bog/mire: Shallow but saturated ground including peat, blanket or raised bogs, possibly mixed with heather, usually with low vegetation such as cotton grass, sedges or sphagnum moss, in upland and lowland areas (BO in Fig. 3). Saltmarsh: shoreline or estuary vegetation that periodically floods at high tide. Water body: any expanse of water including a lake or large river. Cereals: Wheat, barley or oats: specified as winter-sown if cm or more in height by mid-april or spring-sown if less than cm by April (A in Fig. 3). Non-cereals: Potatoes, turnips, sugar beet, kale, vegetable crops and oilseed rape (A in Fig. 3). Bare soil: Caused by cultivation or vehicle disturbance for example (A in Fig. 3). Stubble: Stubble fields from previous cereals, linseed or oilseed rape crop, not yet ploughed in (A in Fig. 3). Woodland/scrub/forest/hedgerows: Deciduous or coniferous woodland, scrub, hedgerows and orchards. Buildings: Any part of a town, village or a group of farm buildings.