J.A. Vickery, British Trust for Ornithology, The Nunnery, Thetford, Norfolk IP24 2PU, UK (email@example.com).
1The effects of agricultural intensification on biodiversity in arable systems of western Europe have received a great deal of attention. However, the recent transformation of grassland systems has been just as profound.
2In Britain, the management of grassland has changed substantially in the second half of the 20th century. A high proportion of lowland grassland is managed intensively. The major changes include a doubling in the use of inorganic nitrogen, a switch from hay to silage, and increased stocking densities, particularly of sheep. Structurally diverse and species-rich swards have been largely replaced by relatively dense, fast-growing and structurally uniform swards, dominated by competitive species.
3Most of these changes have reduced the suitability of grassland as feeding and breeding habitat for birds.
4The most important direct effects have been deterioration of the sward as nesting and wintering habitat, and loss of seed resources as food. Short uniform swards afford poor shelter and camouflage from predators, whereas increased mowing intensities and trampling by stock will destroy nests and young. Increased frequency of sward defoliation reduces flowering and seed set, and hence food availability for seed-eating birds.
5The indirect effects of intensification of management on birds relate largely to changes in the abundance and availability of invertebrate prey. The effects of management vary with its type, timing and intensity, and with invertebrate ecology and phenology, but, in general, the abundance and diversity of invertebrates declines with reductions in sward diversity and structural complexity.
6Low input livestock systems are likely to be central to any future management strategies designed to maintain and restore the ecological diversity of semi-natural lowland grasslands. Low additions of organic fertilizer benefit some invertebrate prey species, and moderate levels of grazing encourage sward heterogeneity.
7There is now a need to improve understanding of how grassland management affects bird population dynamics. Particularly important areas of research include: (i) the interaction between changes in food abundance, due to changes in fertilizer inputs, and food accessibility, due to changes in sward structure; (ii) the interaction between predation rates and management-related changes in habitat; and (iii) the impact of alternative anti-helminithic treatments for livestock on invertebrates and birds.
In the last 25–30 years, many farmland birds that were once common have exhibited widespread population declines and range contractions, both in Britain and abroad (Tucker & Heath 1994; Fuller et al. 1995; Siriwardena et al. 1998). Growing evidence links these declines to changes in farm management practices associated with agricultural intensification (Potts 1986; O’Connor & Shrubb 1986; Evans 1997; Wilson et al. 1997). However, the majority of recent, relevant, research has focused on arable farming systems (Aebischer et al. 2000), even though grassland accounts for over 65% of the agricultural land in Britain (MAFF et al. 1997) and a range of birds depend exclusively or partly on grassland habitats (Vickery et al. 1999; Perkins et al. 2000). Permanent and temporary grasslands occupy approximately 7 million ha (MAFF et al. 1997), representing a higher proportion of the agricultural land area than most European countries. Over 60% of the UK grassland resource is located in the west, with less than 12% in the eastern regions (MAFF et al. 1997; Fig. 1), and most is agriculturally improved or semi-improved.
The management and productivity of lowland grassland in Britain has been transformed during the last 50 years, largely through greater fertilizer inputs, changes in stocking practices, increases in silage production, a greater emphasis on optimizing yields of nutrients rather than dry matter per se, and development of new harvesting techniques (Chamberlain et al. 2000; Fuller 2000). These changes have undoubtedly played an important role in the recent declines of farmland birds, but the mechanisms are poorly known.
A number of correlative studies have suggested that increases in the intensity of grassland management in Britain may affect bird populations. An examination of changes in the spatial ranges of 21 farmland birds in Britain between the 1960s and 1980s revealed the greatest species loss in mainly pastoral 10-km squares in the west of the country. Changes in the management of grassland and livestock were suggested as potential causes of these higher rates of local extinction (Chamberlain & Fuller 2000). Lower numbers of six ‘grassland specialist’ bird species in Britain were associated with areas of higher stocking densities (Pain, Hill & McCracken 1997), and low frequencies of some granivorous birds were associated with large areas of young, reseeded grassland and high sheep densities (Siriwardena et al. 2000). In both studies, the patterns of change in bird populations appeared to be linked to the direct and/or indirect effects of agricultural intensification.
This review had four aims: (i) to identify and outline the range of potential mechanisms by which the intensification of grassland management may impact on bird populations in Britain; (ii) to review our current understanding of the mechanisms involved; (iii) to highlight gaps in current knowledge about the impacts of grassland management on birds; and (iv) to consider ways in which grassland management could be modified to benefit grassland birds.
We focused on lowland neutral grasslands, such as the Lolium perenne–Cynosurus cristatus, Lolio–Plantaginion and Cynosurus cristatus–Centaurea nigra communities (Rodwell 1992) found on moist mineral soils with a pH of between 5·0 and 6·5. We did this for three reasons: they are the most widespread, the least well studied and subject to intense agricultural pressure. Despite their extent, lowland neutral grasslands have received far less scientific attention than more geographically restricted grasslands, such as calcareous and floodplain types. Most previous work on grassland birds in Britain has focused on now localized types of grassland, such as grazing marshes and coastal grassland (Vickery et al. 1997; Norris et al. 1998; Milsom et al. 2000), or the autecology of rare species such as the corncrake Crex crex L. (Tyler, Green & Casey 1998). Relatively little is known about the impacts of agricultural management on the biodiversity of lowland neutral grasslands and, for this reason, we draw, inevitably, on examples from relevant research on other grassland types.
Three broad components of intensification were considered: fertilizer use, stocking practices and cutting regimes. The impacts of drainage, which have been most extensive on lowland wet grasslands (Dargie 1993; Jefferson & Grice 1998), and reseeding were not reviewed because these relate mainly to habitat loss rather than management. In addition, the effects of drainage on the availability of invertebrate prey for waders (Beintema 1983, 1991; Green 1986, 1988; Green & Cadbury 1987; Beintema & Visser 1990; Self, O’Brien & Hirons 1994), predation rates of wader nests and chicks (Beintema & Muskens 1987) and availability of roosting and feeding sites for wildfowl, particularly in winter (Thomas 1980, 1982; Mayhew & Houston 1989; Rees 1990; Vickery & Gill 1999), have been well documented elsewhere. Pesticide use in grassland systems was also not considered because, although the area of grassland treated increased in the 1980s and 1990s, pesticides remain much less frequently used than in arable systems (Thomas & Garthwaite 1994; Garthwaite et al. 1998). Within grasslands, herbicides used to control perennial weeds account for over 60% of the pesticides used on new leys and over 90% of the pesticide applied to established grassland (Garthwaite et al. 1997). In the 1980s avermectin drugs were introduced as anti-helminthic treatments for livestock. This is probably one of the more significant changes in pesticide use in terms of its implications for invertebrates and birds. Avermectin residues are excreted in the faeces of treated animals and, being insecticidal, may reduce the numbers and diversity of invertebrates associated with dung, many of which are important prey items for birds. The effects of increased use of avermectins on invertebrates and birds was considered under grazing management (McCracken & Foster 1993; McCracken 1993).
A large proportion of bird species in Britain uses grassland at some time during the year and many species show preferences for this habitat. However, few can be termed grassland specialists as most also make some use of arable habitats (Pain, Hill & McCracken 1997; Vickery et al. 1999). This review examined the link between grassland management and its use by generalist farmland birds for nesting and, in particular, for foraging, rather than the impact of management on a particular bird species. During the breeding season species considered included mainly waders, for example oystercatcher Haematopus ostralegus L., curlew Numenius arquata L. and lapwing Vanellus vanellus L., and songbirds or passerines, for example skylark Alauda arvensis L., yellowhammer Emberizacitrinella L., meadow pipit Anthus pratensis L. and yellow wagtail Motacilla flava L. In winter, grassland is used by flocks of foraging waders, for example lapwing, golden plover Pluvialis apricaria L. and gulls (especially black-headed Larus ridibundus L. and common Larus canus L.), and passerines, for example starling Sturnus vulgaris L., meadow pipit, redwing Turdus iliacus L. and fieldfare Turdus pilaris L. In addition we included a few more restricted grassland species, such as chough Pyrrhocorax pyrrhocorax L., cirl bunting Emberiza cirlus L. and corncrake, all of which may be considered grassland specialists. Waders such as snipe Gallinago gallinago L. and redshank Tringa totanus L. were also included because, although associated mainly with wet lowland grasslands, they are also found on damp neutral grasslands.
Although a range of wildfowl uses grassland habitats, we excluded them because many have relatively localized distributions and are associated largely with wet grasslands. Furthermore, the impacts of grassland management on geese have been reviewed elsewhere (Vickery & Gill 1999).
Grassland management practices – impacts on plants, invertebrates and birds
Use of fertilizer on grassland
Recent changes in grassland fertilization practice
Between 1970 and 1986, the average use of inorganic nitrogen fertilizer (N) on grassland increased by around 100% to c. 132 kg N ha−1 year−1 (ADAS/FMA 1992; MAFF et al. 1997) (Fig. 2). Since 1986, nitrogen fertilizer inputs to grassland have decreased to around 115 kg N ha−1 year−1 (MAFF et al. 1997) but, in 1997, 86% of all grassland in the UK was still receiving inorganic nitrogen fertilizer (MAFF et al. 1997). Application rates vary with grassland age and management regime. For example, in 1997, c. 165 kg N ha−1 year−1 was applied to grasslands of less than 7 years of age, compared with c. 103 kg N ha−1 year−1 to grasslands of over 5 years of age. Fields cut for silage, but not grazed, generally receive the highest inputs, c. 221 kg N ha−1 year−1, and fields mown for hay receive the lowest, c. 90 kg N ha−1 year−1 (MAFF et al. 1997). Dairy farms use approximately twice the amount of nitrogen fertilizer compared with beef/sheep farms.
Other inorganic fertilizer inputs include phosphorus (P) and potassium (K), applied on c. 67% of grassland at average rates of 15 kg P ha−1 and 45 kg K ha−1. In addition to inorganic fertilizer inputs, 48% of grassland (2·3 million ha) also receives annual applications of organic fertilizer as solid manure and/or liquid slurry, but there is little recent information on quantities (B.J. Chambers, K.A. Smith & B.F. Pain, unpublished data).
Fertilizer-related increases in net primary production, decreases in plant species diversity and sward structural complexity, and increases in the nutrient content of the vegetation may all affect grassland invertebrates, but effects differ between species. Some, such as leatherjackets (Tipulidae larvae) and nematodes (Nematoda), seem unaffected by fertilizer application (Linzell & Madge 1986), whereas others, such as mirid bugs (for example Leptoterna dolabrata L.), may show increased fecundity or reduced larval mortality (McNeill 1973). Increases in plant nitrogen content may promote the abundance of phytophagous species such as Auchenorrhyncha (Andrzejewska 1976) but, in general, additions of inorganic fertilizer, phosphorous, potassium and lime (calcium/magnesium carbonate) all tend to decrease the numbers and diversity of grassland invertebrates (Fenner & Palmer 1998). A number of invertebrate groups, including Acari, Collembola, Diptera, Coleoptera, Orthoptera and Myriapoda, exhibits moderate to severe population reductions in response to fertilizer applications (Edwards & Lofty 1975; Van Wingerden, Vankreveld & Bongers 1992). Earthworm numbers benefit from moderate fertilizer applications but suffer under high application rates (Edwards & Lofty 1975, 1982; Zajonc 1975; Standen 1984; Unwin & Lewis 1986) and, while numbers increase with increasing soluble soil phosphorous (Nuutinen et al. 1998), sulphate of ammonia appears toxic to earthworms in acid soils (Satchell 1955; Edwards 1977, 1983).
There is evidence to suggest that management intensity influences the size, as well as the abundance and diversity, of invertebrates. Intensive grassland management with high inputs of fertilizer and intensive grazing or mowing may be particularly detrimental to larger insect species (Beintema et al. 1990). Blake et al. (1994, 1996) showed that, although beetle numbers increased with increasing intensity of grassland management, size declined.
The effects of organic and inorganic fertilizers may be comparable, in terms of nutrient enrichment, but organic fertilizers provide extra food for the decomposer communities, and grassland soil invertebrate populations generally benefit from moderate applications of organic manures (Marshall 1977). Input of readily assimilated nutrients may raise the productivity of soil and turf invertebrates (Keiller, Buse & Cherrett 1995). As with inorganic fertilizer, earthworm populations seem to increase with moderate applications of farmyard manure and slurry but decrease under high applications (Edwards & Lofty 1982; Standen 1984; Unwin & Lewis 1986). High numbers of bibionids (Diptera: Bibionidae) and leatherjackets, both important prey items for birds, are associated with areas where large amounts of organic matter (dung or slurry) have been applied in the previous season (D’Arcy-Burt & Blackshaw 1991; McCracken, Foster & Kelly 1995).
Impacts on birds
Fertilized swards, which tend to be fast-growing, species-poor and dense, provide very different feeding and nesting habitats for birds compared with unfertilized swards (Bunce et al. 1998). Rapid changes in sward height, particularly in the spring, affect the foraging efficiency and the availability of nest sites for waders. Different species show distinct preferences for different sward heights. The lapwing, for example, prefers relatively short swards for nesting (Shrubb & Lack 1991). In a study on the Somerset Levels, western England, the highest breeding densities of this species were in areas where vegetation height was 10–15 cm in mid-May (Green 1986) and they may abandon sites when the grass is taller than this (Lister 1964 cited in Hudson, Tucker & Fuller 1994). Other species, such as breeding snipe and redshank, tend to prefer taller tussocky swards (Mason & Macdonald 1976) of 15–20 cm in height (Green 1986). In addition, important prey items may be less abundant and less accessible in the tall, dense, swards of heavily fertilized grassland for species such as the lapwing, which detect prey visually (Redfern 1982), and snipe, which probe for prey (Green 1986). In winter, curlew (in coastal areas only), lapwing and golden plover all show preferences for pasture with short swards (Townsend 1981; Barnard & Thompson 1985; Milsom, Holditch & Rochard 1985; Lack 1986; Milsom et al. 1998).
With respect to nesting habitat, species that show active nest defence towards potential predators require good visibility and hence favour short vegetation. The lapwing, for example, selects small patches of short swards even when breeding in hay meadows (Nairn, Herbert & Heery 1988), and heavily fertilized silage may be too high for nesting (Shrubb & Lack 1991). Although waders, such as snipe and redshank, rely on concealment and require tussocky areas for breeding (Green 1986; Nairn, Herbert & Heery 1988), they may avoid uniform dense swards, such as those maintained under silage production, possibly because these offer poor chick foraging habitat. Changes in the nesting and feeding quality of grassland associated with increased fertilizer use may have caused the wide-scale shift in the timing of breeding of some waders in the Netherlands. A range of species, including lapwing, oystercatcher, redshank and snipe, now breed 1–2 weeks earlier than in the early 1900s, although the effect of this shift, at the population level, is unclear (Beintema, Beintema-Hietbrink & Muskens 1985).
Fertilizer-related reductions in numbers and accessibility of invertebrates, such as Diptera, Coleoptera, Orthoptera and Hemiptera, are also detrimental for grassland passerines. Species such as skylark, meadow pipit and starling tend to select open areas of low vegetation cover for foraging (Feare 1984; Cramp 1988; Cramp & Perrins 1994; Wilson et al. 1997; Schön 1999). Changes in the breeding numbers of yellow wagtail, which also favour short vegetation, have been attributed to changes in management resulting in tall and dense swards (Cramp 1988). Fertilizer-related reductions in botanical diversity may also reduce seed availability in summer and winter for birds such as the linnet Carduelis cannabina L., turtle dove Steptopelia turtur L. and a number of bunting species.
Moderate use of organic fertilizer (farmyard manure) may benefit grassland birds by increasing the abundance of soil-dwelling invertebrates, or their accessibility, by bringing them closer to the surface (Scullion & Ramshaw 1987; Tucker 1992). Winter field use by lapwing, starling, fieldfare and redwing is positively associated with frequent addition of farmyard manure on permanent grassland (Tucker 1992), although associated increases in sward height may counteract these benefits for some species (Milsom, Holditch & Rochard 1985; Milsom et al. 1998).
Livestock stocking practices
Recent changes in livestock grazing practices
Between 1976 and 1997, the total number of sheep in the UK increased by over 50% (Fig. 3), with densities and rates of increase especially high in Wales and northern England (Fuller & Gough 1999a). Although several regions have shown a plateau or decline in sheep numbers since 1980, numbers in 1997 were still far higher than in the mid-1970s. However, broad regional trends conceal much variation at the local scale (Fuller & Gough 1999a). In contrast, the total cattle herd declined by about 18% (Fig. 3). The UK dairy herd has declined by c. 16% in the last 20 years, although there has been an increase in the size of the beef herd (Fig. 4). When these figures are converted to live weights for seven selected ‘grassland’ counties, they suggest an increase in live weight carried by grassland of only c. 4% between 1980 and 1997. The area of UK grassland currently grazed by different livestock is difficult to ascertain. The only comprehensive figures that exist are for south-west England in 1983, where virtually all the grassland was grazed for at least part of the year: 34% by cattle, 9% by sheep, 56% by cattle and sheep, and less than 1% by horses (Hopkins & Peel 1985).
Probably one of the more significant changes in pesticide use in grassland, in terms of impact on invertebrates and birds, is the increasing use of avermectins. These comprise a group of broad-spectrum anti-parasitic compounds derived from the naturally occurring soil actinomycete Streptomyces avermitilis Burg. The compound avermectin B1a (abamectin), isolated following fermentation of S. avermitilis, possesses high potency against a broad spectrum of endo- and ecto-parasites of livestock. The semi-synthetic 22,23-dihydro analogue of avermectin B1, ivermectin (Fisher & Mrozik 1989), is the most widely used of these drugs and its effects on dung-dwelling invertebrates and birds is the best studied. Ivermectin was introduced onto the animal health market in 1981 and is approved for use in Britain in cattle, sheep, goats, pigs and horses (McCracken & Bignal 1991).
Impacts on plants and invertebrates
Grazing acts upon individual plants and plant communities through defoliation, trampling, treading and pawing, deposition of dung and urine, and poaching, which alter the relative abundance and competitive abilities of the different plant species (Jensen 1985). Increased stocking densities create densely tillered swards (Grant et al. 1983; Orr et al. 1988; Tallowin, Williams & Kirkham 1989). For example, continuous grazing by sheep of Lolium perenne L.-dominated grassland can create tiller densities of 25–30 000 m−2, almost twice those created under cutting at 4-weekly intervals (Orr et al. 1988) and markedly higher than under continuous intensive grazing by cattle (Tallowin, Williams & Large 1986). Swards grazed intensively by sheep are extremely uniform (Kiehl et al. 1996; Berg et al. 1997). Cattle-grazed swards are more spatially heterogeneous, principally due to the more patchy distribution of dung (Richards & Wolton 1976) and the lack of grazing around freshly deposited dung. The size and extent of patches of heavily grazed and lightly grazed swards depends upon stocking density but, generally, any decline in cattle numbers coupled with an increase or maintenance of high sheep numbers will result in a more uniform sward structure (Table 1).
Table 1. Characteristics of sheep grazing, cattle grazing and mechanical cutting in agriculturally unimproved grasslands (adapted from Crofts & Jefferson 1999)
• Bite the vegetation, graze close to ground level and produce very short swards of minimum height 3 cm
Bite, pull and tear the vegetation, cannot graze as close to ground level, and maintain longer swards with minimum height 5–6 cm
• Able to manipulate vegetation and select items from very low in the grassland profile
Coarse level manipulation of vegetation, and relatively unselective grazers
• Avoid tall plants in the sward, leave grass stems and often select flowers
Take tall plants and grass stems, and occasionally select flowers (orchids)
• Dead material and litter left
Some dead material taken
• Often avoid rough, tall swards and tussocky areas
Utilize rough, tall swards and tussocky areas
• Graze preferentially in small patches, selecting the most palatable patches available
As sheep. Cattle swards are often particularly patchy
Leaves swards extremely homogenous in height
• Returns some organic matter in dung and urine
Returns organic matter in dung and urine.
Usually returns little or no organic matter. If cuttings areleft as a dense mat they cannot be utilized by decomposers
Large dung pats promote sward heterogeneity
Intensive grazing reduces botanical diversity by favouring a few species, such as Lolium perenne and Trifolium repens L., that can tolerate repeated defoliation. Competitive species, such as Poa trivialis L., and those with strongly developed defences against herbivory, such as Cirsium ravense (L.) Scop and Cirsium vulgare (Savi) Ten. (Tallowin, Brookman & Santos 1995; Tallowin & Brookman 1996), are among the few non-crop species that can survive. Intensive grazing also reduces the opportunity for flowering and seeding.
Although all methods of grazing management lower the mean height of the vegetation and reduce the standing crop, there are important differences between grazing and other sward management practices, particularly in terms of their effects on invertebrates (Morris 1990a,b). Because grazing is selective, insects associated with plants that are resistant to defoliation may survive intensive grazing but not intensive mowing (Morris 1990a,b, Table 2). Seasonal, rather than continuous, grazing is likely to promote sward heterogeneity, and hence invertebrate diversity (Morris 1971, 1973, 1979; Morris & Rispin 1987, 1988). Grazing in autumn is less deleterious than grazing in spring in terms of overall insect diversity (Table 2); Heteroptera, in particular, are enhanced by autumn grazing but reduced by spring grazing (Brown, Gibson & Sterling 1990).
Useful if dominant species are relatively unpalatable, e.g. tor-grass Brachypodium pinnatum L., mat-grass Nardus stricta L. and rushes Juncus spp.
Repeated heavy grazing of swards with early flowering plants, e.g. fritillary Fritillaria meleagris L., and annuals, e.g. hay rattle Rhinanthus minor L., can cause local extinctions and damage or eliminate a significant range of invertebrates
Can be used to check the growth of scrub seedlings, ragwort Senecio spp. and purple moor-grass Moliniacaerulea (L.) Moench (late spring)
High stocking densities may result in loss, through trampling, of eggs and chicks of some waders and passerines
Helps control tall herb species, e.g. great willowherb Epilobiumhirsutum L. and meadowsweet Filipendula ulmaria (L.) Maxim, and woody scrub (fresh leaves and shoots of species such as sycamore Acerpseudoplatanus L., oak Quercus spp. and ash Fraxinus excelsior L. are highly palatable to domestic stock)
At high stocking rates grazing removes flowers and can prevent seed set. Although perennials persist in a vegetative form, continual grazing may affect annual and biennial species. It may also damage a significant range of invertebrates, through reducing structural and species diversity of the sward, and results in a simple, specialist, short-turf fauna
Less nutrient build up from animal dung due to greater microbial processes
Changes in sward and associated invertebrate community under intensive grazing reduces food for birds directly (seed and green material) and indirectly (invertebrates)
Soil moisture is generally low so there is less chance of poaching
Least damaging time for sensitive invertebrates
Sward palatability declines markedly, many species are rejected and competitive species may dominate swards
Most plant species have finished flowering and set seed, thus grazing has little impact
Heavy grazing may remove flowers of some species, e.g. devil’s bit scabious Succisa pratensis Moench, and this may reduce invertebrate numbers
Most grassland herbs are dormant and hence not directly affected by winter grazing
Heavy trampling may lead to poaching and weed infestation
Winter grazing is less damaging for invertebrates that often over-winter in the base of tussocks
Intensive grazing may remove plant litter, which is an important over-wintering habitat for invertebrates
Modest trampling breaks up the litter layer and exposes bare ground to allow seedling recruitment
Less likely to control tall grasses
Stock may lose condition and supplementary feeding may be required
Selective moderate grazing can result in structurally heterogeneous swards, which support a range of herbivorous invertebrates such as Auchenorrhyncha (Morris 1971; Strong, Lawton & Southwood 1984). The effects of increasing grazing intensity are severe on species feeding on above-ground parts of plants, and will usually reduce numbers and biomass of phytophagous grassland invertebrates (East & Pottinger 1983). However, several soil-dwelling invertebrates, such as soldier flies Inopus rubriceps (Stratiomyidae) Macquart, grass grub Costelytra zealandica White and some other chafers (Scarabaeidae), are also influenced by changes in stocking rates, either through the indirect effects of defoliation or through trampling (Duffey 1975; East & Pottinger 1983). Soil compaction by machinery or stock will have a significant effect on soil invertebrates. Some earthworms can burrow into compacted soil (Joschko, Diestel & Larink 1989) but others have their activity restricted by compaction under conditions of high water potential (Kretzschmar 1991). Soil compaction has also been shown to decrease slug populations (Ferguson, Barratt & Jones 1988).
The introduction of avermectins (particularly ivermectin) has significant implications for dung-dwelling invertebrates because insecticidal residues from these drugs may be found in the dung pats. Halley et al. (1993) and Wratten et al. (1993) conclude that, due to the availability of residue-free dung and the mobility of the insects concerned, the effect on dung-associated insects at the population level is likely to be limited but there could be severe effects at a local level. Ivermectin in the dung at levels of 0·5 mg kg−1 dung markedly changes the invertebrate fauna in and below the dung pat (McCracken & Foster 1993). Many studies (Madsen et al. 1990; McCracken & Foster 1993; Wardhaugh et al. 1993; Floate 1998) have found that insect activity was significantly reduced in dung from treated cattle and that a diverse group of insects was affected, including coprophagous flies, predaceous and coprophagous beetles and parasitic wasps. Dung from ivermectin-treated cattle may contain fewer larval Scarabaeidae and Cyclorrhapha (Diptera) larvae and avermectin residues may also arrest development of some Scarabaeidae larvae (Strong & Wall 1994; Strong et al. 1996).
Impacts on birds
Grazing can impact on bird populations through a large number of mechanisms. In the case of sheep grazing, these have been reviewed by Fuller & Gough (1999b). Interactions between these mechanisms are complex and their relative importance is difficult to assess, but three seem central: changes in vegetation structure, food resources and predation pressure.
Alteration of the vegetation structure will affect the suitability of the sward for nesting and feeding. The foraging behaviour of many grassland birds is heavily influenced by sward height, which modifies prey availability, and grazed areas are often preferred by a range of invertebrate-feeders, possibly due to the increased availability of prey in the short sward (Milsom et al. 1998). Heavily sheep-grazed swards can provide attractive feeding sites for passerines such as starling, blackbird Turdus merula L., carrion crow Corvus corone L., jackdaw Corvus monedula L., magpie Pica pica L., rook Corvus frugilegus L. and mistle thrush Turdus viscivorus L. (Tucker 1992; Wilson, Taylor & Muirhead 1996). This may reflect preferences for feeding in short grass, the increase in numbers of subsurface invertebrates by dunging, or the localized source of invertebrates in dung. These grazing-related factors may be responsible, at least in part, for increases in numbers of magpie, jackdaw and carrion crow in recent decades, which have been especially marked on grazed farms (Gregory & Marchant 1996). Dung-dwelling invertebrates have also been shown to be particularly important as prey for lapwing chicks (Bientema et al. 1990) and chough (McCracken & Foster 1993).
Recent, as well as current, farm management also influences the availability and abundance of insect prey for a wide range of farmland birds. This has been highlighted by studies of the way in which grazing-related differences in sward height influence the availability of leatherjackets for chough (Bignal & McCracken 1996; Bignal et al. 1996; McCracken & Bignal 1998). Relatively tall swards in autumn encourage egg laying by craneflies and promote high numbers of leatherjacket (Tipulidae) larvae in the soil the following winter and spring. However, the chough requires short swards in late winter/early spring to access these larvae, so tall autumn swards need to be grazed down over winter.
Although the impact of grazing, at even moderate to heavy levels, on the suitability of grassland for foraging by birds may be beneficial, impacts related to predation pressure and trampling of nests are highly detrimental. Alteration of sward structure through grazing could be linked with predation pressure on ground-nesting birds in two ways. First, the uniform sward structure, created particularly by grazing with sheep, may increase the likelihood of a predator detecting nests, chicks or adults by reducing cover or camouflage. Baines (1990) has suggested this as a factor resulting in higher nest losses for lapwing on agriculturally improved compared with unimproved grassland. Secondly, grazing may increase predation pressure by acting to increase predator numbers. For example, increases in soil invertebrates may benefit corvids, which are common predators on ground-nesting birds at the egg stage (Grant et al. 1999).
The most direct impact of grazing animals on breeding waders, such as redshank and snipe, is the destruction of nests through trampling (Green 1986, 1988; Beintema & Muskens 1987; Shrubb 1990). The rate of nest destruction depends on the type and density of stock, the timing of grazing (Table 2) and the bird species involved. Trampling rates per individual animal are lower for sheep than for cattle but, when comparing equivalent stocking rates of cattle and sheep (one adult cow = three yearling cattle = five sheep), trampling rates are similar (Beinteima & Muskens 1987). Redshank and snipe nests seem particularly vulnerable to trampling, and a stocking rate of approximately 2·5 cows ha−1 for the whole of the incubation period leads to approximately 70% of redshank, 60% of snipe and 35% of lapwing nests being trampled (Green 1986; Beintema & Muskens 1987; Shrubb 1990). The timing of stock turn-out onto grassland will also influence the vulnerability of nests. Where early turn-out of cattle or early mowing is precluded (for example by wet soil conditions in spring), early nesters will benefit over late nesters, particularly as the opportunities for successful replacement clutches become much reduced as the season advances (Beintema 1988; Beintema & Muskens 1987).
The type of grazed livestock is also important for some species. Breeding lapwing, for example, although generally associated with grazed rather than ungrazed areas, is most closely associated with horses or sheep rather than cattle (Shrubb & Lack 1991). In winter, however, both lapwing and golden plover appear to prefer cattle- rather than sheep-grazed pasture (Tucker 1992). Breeding redshank prefers a varied, tussocky, sward structure and also tends to select cattle- rather than sheep-grazed grassland (Herbert, Heery & Meredith 1990).
There is concern about the possible indirect effects of ivermectins on birds, such as pipit, wagtail, lapwing, thrushes and corvids, which feed in, or around, animal dung (McCracken & Bignal 1991). However, although there are clear effects of ivermectins on dung-dwelling invertebrates, the impacts on birds have been relatively poorly studied. Research on chough, for which invertebrates in cow dung are a particularly important food resource (McCracken & Foster 1993), suggests breeding success and juvenile survival may be reduced in areas where ivermectins are used (McCracken 1993; McCracken, Foster & Kelly (1995).
Management of grasslands by cutting
Recent changes in management of grasslands by cutting
Approximately 40% of the enclosed grasslands in Britain are mown annually (MAFF et al. 1997). Technical innovation, improvements in silage making and an increased emphasis on optimizing yields of nutrients rather than the yield of dry matter per se are among the most significant of all developments to affect agricultural grassland in the post-1940 period (Frame, Baker & Henderson 1995; Merry, Jones & Theodorou 2000). Silage making is a relatively recent phenomenon. In the early 1970s, the ratio of hay to silage (on a dry weight basis) in the UK was around 85 : 15, but by the mid-1990s this had reversed to around 30 : 70 (MAFF et al. 1997). The ability to conserve silage with a moisture content of 25–40% compared with the need to achieve a moisture content of < 18% for hay (Merry, Jones & Theodorou 2000) allows much greater flexibility in the timing of silage making, an important attribute given the vagaries of the British climate. However, with the adoption of plastic-wrapped big-bale silage of relatively low moisture content, made from phenologically mature grassland, the distinction between this material and hay is narrow.
Contamination of silage by soil can cause serious spoilage by introducing Clostridium species; these bacteria ferment carbohydrates and lactic acid to butyric acid, making silage rancid and unpalatable or toxic to livestock (Halley & Soffe 1988). To avoid this, silage fields are rolled in the early spring, often eliminating microtopographical features (Clements & Cook 1996). This is not the case in haymaking where soil contamination is not perceived as a major problem.
Impacts on plants and invertebrates
Like grazing, mowing reduces the size and complexity of the vegetation and can increase primary production. However, unlike grazing, mowing is non-selective, so sward heterogeneity is greatly reduced, and the removal of grass also reduces the amount of organic matter returned to the soil. This applies to hay as well as silage, but the switch to the latter has had a considerable impact on both grassland structure and botanical diversity. The higher acceptable moisture content of silage and the increased emphasis on optimizing nutrient yield means that grasslands are generally cut prior to any expression of flowering. The first silage harvest is usually taken in May in much of lowland Britain, a month or more before hay, which is usually cut in late June–August. Haymaking, in contrast, allows considerable flowering (Smith & Jones 1991); because the process is so dependent on having a period of settled dry weather it tends to be delayed until mid-summer. Dung and farmyard manure from stock fed hay or late-cut big-bale silage may contain considerable quantities of seed from grassland plant species. However, seed resources are likely to be absent or very limited in manure from livestock systems based on feeding silage of high nutrient content and digestibility (Marshall & Hopkins 1990).
The structure of grassland vegetation is of the utmost importance in maintaining arthropod diversity. In general, the abundance and diversity of most arthropod groups declines with declining sward height (Morris 2000). Studies focusing on calcareous grasslands and the management of grassy arable field margins have shown grassland fauna to be depleted by cutting and removing herbage under both hay and silage management (Curry & Tuohy 1978; Morris 1990a,b; Kirby 1992; Volkl et al. 1993; Smith 1994; Feber, Smith & Macdonald 1996; Macdonald et al. 2000). Thomas & Jepson (1997) showed that cutting for silage significantly depleted linyphiid spider populations after each cut, but heavy grazing by sheep resulted in a significantly lower density of spiders than silage cutting. Coleoptera appear to be less sensitive to cutting treatment than Auchenorrhyncha. Cutting management imposed on calcareous grassland tends to favour phytophagous beetles but was found to be deleterious to numbers of those in other trophic groups (Morris & Rispin 1988). The faunas of established, but regularly managed, grassland comprise many eurytopic species with high fecundity and good colonizing abilities (Curry 1994), which facilitate rapid recovery from mowing (Andrzejewska 1979). Morris & Lakhani (1979) showed that the Auchenorrhyncha of a recently established grass sward had a larger proportion of bi- and multi-voltine species than those of mature unimproved grasslands, and the impact of mowing, on abundance and species richness, was relatively short lived on the former.
Timing, rather than frequency, of cutting may be more important for some invertebrate groups. Baines et al. (1998) showed that cutting grassy field margins once in summer had more severe effects on both the abundance and species richness of spiders than cutting twice in spring and autumn. This may be because structurally diverse swards offer better web-building sites and prey densities are higher in taller vegetation (Southwood, Brown & Reader 1979). Morris (1979) found that double-brooded mirid species were less affected by summer cutting than single-brooded species, and species in which most adults emerged in early summer were less affected by cutting in July than those emerging later. Similarly, Duffey et al. (1974) showed that cutting in May allowed a more rapid recovery of invertebrate populations than cutting in July. Most of the immediate deleterious effects of cutting, for example on the species richness of Auchenorrhyncha, have been attributed to loss of vegetation structure (Morris 1981), but the removal of flowers may also have significant effects on nectar-feeding invertebrates such as butterflies (Lepidoptera) (Feber, Smith & Macdonald 1996).
Impacts on birds
The ways in which cutting regimes will impact on birds are similar to those of grazing, namely through changes in the availability of preferred vegetation types and alterations in predation pressure and food availability. As with grazing, the interactions between these factors are complex. In addition, intensive mowing is invariably associated with silage making and hence also with high nitrogen inputs. It is therefore difficult to distinguish the indirect effects on birds of cutting per se.
The creation of short swards through regular mowing, as under moderate grazing, may increase the accessibility of prey for invertebrate feeders such as starling, wagtail and pipit. However, unlike grazing, this is likely to be offset by reductions in both the abundance and diversity of invertebrates. Cutting can result in a temporary flush of invertebrate availability, and hirundines, wagtail, pipit and starling will often congregate on freshly cut hay or silage (G. Conway, personal communication).
Intensive mowing is unlikely to be associated with any increase in predator numbers, as may be the case under intensive grazing. However, the resultant uniform sward structure will almost certainly increase the likelihood of a predator detecting nests, chicks or adults.
Regular intensive mowing can severely reduce the breeding success of ground-nesting birds through mechanical destruction of nests and young. Probably the best documented example of this is the corncrake, where earlier and more frequent mowing, associated with a switch from hay to silage, has caused high rates of adult and chick mortality and nest loss. This has been a major cause of the decline and range contraction of this species in the UK (Green & Stowe 1993; Green 1995). For breeding waders, earlier and more frequent mowing also reduces nesting success (Beintema, Beintema-Hietbrink & Muskens 1985; Beintema & Muskens 1987), and it seems likely that ground-nesting passerines will be similarly affected, although little information is available. Multi-brooded species, such as skylark, can lay replacement clutches, but earlier and more frequent mowing does reduce breeding success (P.F. Donald, unpublished data).
Conclusions and conservation management implications
Management intensification since the 1950s has created striking uniformity in floristic composition and structure within grazed and cut lowland grassland in Britain. This review identifies a close relationship between the changes in grassland management and an apparent profound and widespread loss of food resources and preferred habitats of grassland birds, culminating in a marked pattern of local extinction of some species, particularly in western Britain, since c. 1970. However, empirical data on the impacts and interactions between the grassland floristic changes and changes in faunal populations in neutral grassland are limited, particularly in relation to providing causal links with the observed large changes in bird populations.
In this review, we have identified a web of potential mechanisms through which changing grassland management may have direct and/or indirect effects on bird populations. Mechanical destruction of nests and young birds during mowing operations and trampling by livestock probably comprise the principal direct mechanism for grassland bird loss. Indirect management effects, such as the near-ubiquitous switch to silage production and elevated fertilizer inputs and the resultant creation of homogeneity in botanical composition and structure of the grassland, are more difficult to interpret but probably also highly significant. The complex interactions that exist between different components of grassland management, sward characteristics, invertebrates and the performance of breeding birds are summarized in Table 3 and Fig. 5. These mechanisms are further complicated by the impact of management history and the wider environmental characteristics of the site, such as soil and climatic conditions (McCracken & Bignal 1998).
Table 3. Changes in fertilizer, grazing and mowing practices on grassland in Britain, and the potential implications for sward composition and structure, invertebrates and birds
Changes in management practice
Impact on sward composition and structure
Impact on invertebrates
Impact on birds
Inorganic fertilizer use
Two- to threefold increase in inputs since 1940s (peaked in 1980s)
Increases density and reduces structural diversity of sward
General declines in invertebrate abundance and diversity (but responses are complex)
Tall, dense, swards are less suitable nesting habitats for some waders, e.g. lapwing, snipe and redshank, and invertebrate prey are less abundant and/or accessible, e.g. for waders, wagtail and pipit
Currently 86% of grassland receives N fertilizer (highest inputs on silage, lowest on hay fields)
Encourages growth of competitive species, favouring species-poor mesotrophic/eutrophic communities, e.g. Lolium perenne–Cynosurus cristatus or Lolium perenne leys
Reduces numbers of Acari, Collembola, Diptera, Coleoptera, Orthoptera and Myriapoda
Changes in sward characteristics may cause an advancement and extension of the breeding season of some waders, e.g. lapwing, oystercatcher and redshank
Reduces earthworm numbers (but numbers increase under moderate fertilizer applications)
Auchenorryhyncha and Cicadellids decrease in diversity but increase in numbers
Leatherjackets and nematodes are unaffected
Organic fertilizer use
Increases in use of solid manure and/or liquid slurry since 1940s
Periodic light to moderate use is compatible with the maintenance of some species-rich grasslands
Low to moderate applications benefit many invertebrates, especially the decomposer communities
Increases in soil-dwelling invertebrates (low/moderate applications) benefit some species, e.g. starling and golden plover. Reverse is true under high applications
Currently 48% of grassland receives organic manure annually (heaviest use on young, annually mown, grass)
Heavy use reduces species and structural diversity of the sward
High applications may reduce numbers and/or diversity of some invertebrates, e.g. Collembola, Acari and nematodes
Livestock stocking practices
Doubling of sheep numbers between 1940 and 1990
Intense grazing, especially by sheep, promotes densely tillered swards and reduces structural complexity
Reductions in sward structural and species diversity will result in a general decline in invertebrate numbers and diversity, especially phytophagous insects, but effects vary with timing and intensity of grazing
Intense grazing causes increased nest loss through trampling, especially for waders, e.g. lapwing and redshank, and increased sward uniformity may increase detectability of ground nests to predators
Increases in beef herd and decreases in dairy herd in the last 20 years
Intense grazing favours competitive species such as Lolium perenne and Trifolium repens or species with well-developed defences against grazing, such as Cirsium spp.
Seasonal grazing, in autumn rather than spring, may promote sward structural complexity and hence invertebrate numbers and diversity
Structural and species diversity, promoted by moderate/light grazing, may benefit some waders and passerines by providing short swards, where invertebrates are readily accessible, and taller nesting cover
Increases in livestock numbers in western Britain and decreases in eastern Britain
Moderate grazing, especially by cattle, promotes structural heterogeneity in the sward
Soil-dwelling invertebrates, e.g. slugs, may be adversely affected by trampling and soil compaction
Birds feeding on invertebrates associated with dung, e.g. wagtail and pipit, benefit from increased food
Little change in overall live weight carried by grassland in the last 20 years
Addition of dung to grassland will provide habitat for coprophagous flies and beetles
Widespread replacement of hay with silage since 1970s has led to earlier, more frequent, mowing and increased levels of reseeding and rolling
Intensive cutting reduces biomass, species and structural complexity of the sward
Reduction in floristic and structural diversity of the sward reduces the abundance and diversity of invertebrates
Reduced productivity of ground-nesting species through mechanical destruction of nests and young
Frequent cutting prolongs the vegetative phase of grass growth, usually preventing grasses flowering
Effects vary with timing of cutting. Summer is generally more detrimental than spring or autumn for a number of species, e.g. within Araneae and Hemiptera
Short swards may facilitate foraging by a range of passerines and waders but this may be offset by increased sward density and reduced invertebrate abundance
Cutting reduces the return of organic matter to the soil
Rolling may reduce populations of leatherjackets and slugs
Rolling reduces topographical diversity
Few linkages/interactions between grassland management and changes in faunal populations have been established and/or quantified, and our understanding of the relative importance of the processes involved is inadequate. A detailed understanding of the role played by farm management and wider environmental factors in influencing the invertebrate populations is essential to underpin effective measures for farmland birds within grassland systems (McCracken & Bignal 1998).
Attempts to improve the value of grassland for birds could be directed at management at the level of the whole field or field margins. The management of field margins for biodiversity is an accepted practice in arable systems. Margins provide particularly valuable foraging habitats in summer when many hedgerow nesting species are often constrained to foraging close to the nest. With the exception of skylark and ground-nesting waders, few nesting/foraging species avoid margins and many seem to prefer them (Vickery & Fuller 1998). In arable landscapes, uncultivated grass margins, in particular, are favoured by birds such as yellowhammer, cirl and corn buntings (Evans 1997; Bradbury et al. 2000; Brickle et al. 2000). It seems likely that similar benefits might accrue from grassland margins in grassland habitats, but quantitative data are lacking; for example yellowhammer avoids nesting in improved pastures and silage leys (Bradbury et al. 2000) but this effect may be reduced in the presence of uncultivated grass margins.
On intensively grazed grassland, protecting field edges from grazing during the summer has been shown to increase the abundance of chick food insects, such as true bugs (Heteroptera), sawfly larvae (Symphyta) and caterpillars (Lepidoptera), while reducing herbicides in field edges has also led to increases in sawflies and caterpillars (Haysom et al. 2000). Differences in carabid species composition on experimentally managed margins of silage have also been recorded in relation to different cutting regimes in those margins (Haysom, McCracken & Foster 1999). Other margin treatments could also benefit birds. For example, Perkins et al. (2000) have suggested that mechanical creation and maintenance of exposed soil surfaces may provide improved foraging opportunities for winter birds. Clearing strips in dense swards using herbicide or mechanical cutting may facilitate access to prey for ground foragers throughout the year. Some treatments could benefit birds by increasing the abundance or availability of prey beyond the area of the margin itself. Leaving an uncut and/or unfertilized strip at the edge of silage fields to provide a reservoir of invertebrates able to recolonize the grass sward after cutting may increase prey abundance in the cropped field as well as the margin itself. Although this review has focused on the cropped habitat, the value of any management of grass fields or margins for birds will be enhanced by sympathetic management of adjacent hedges to provide nesting and foraging habitat (Parish, Lakhani & Sparks 1994; Barr, Britt & Sparks 1995). In studies of set-aside (Gates et al. 1997) and arable field margins (Feber et al. 1995), boundary characteristics have been shown to affect invertebrate communities within the field itself, but further research in this area is needed.
Low-input extensive livestock systems have historically created and maintained the ecological diversity of unimproved grasslands in Europe, and the restoration of such systems appears to be central to any attempts to restore grassland biodiversity. To date, agri-environmental policy has been largely directed towards ameliorating damaging effects of intensive agricultural management, and relatively little support has gone towards the development of ecologically sustainable low-intensity farming systems. The latter are often highly intensive in terms of human labour requirement, and solutions are required that allow a way of life that is both socially and economically attractive to farmers, and which maintains land management practices beneficial to wildlife (Bignal & McCracken 1996). Redirection or modification of support payments within agri-environment schemes that link livestock headage payments (to reduce stocking levels) with environmental conditions are needed.
The agri-environment regulation (EEC 2078/92) introduced as part of the May 1992 Common Agricultural Policy (CAP) reforms includes provision of financial support for traditional farming systems and extensification options. Support payments based on area, rather than volume of production, in line with those operating in the arable sector (Pain, Hill & McCracken 1997), are now needed for low-intensity low-yielding pastoral farming systems. However, clear definitions of management criteria are essential because of incompatibilities between habitat requirements for some environmental ‘products’. For example, some pasture managements that maintain high botanical richness will do little to enhance bird populations (Pärt & Söderström 1999). Tailoring options to specific regions or species may increase the cost effectiveness with which such schemes meet their environmental targets, as has been the case, for example, in the management of grassland for corncrake.
There is an urgent need for farmland bird research, hitherto largely arable focused, to broaden to encompass grassland systems. Almost all the mechanisms outlined in Table 3 and Fig. 5 merit further investigation. We suggest three particularly important issues are: the interaction between changes in food abundance, due to changes in nutrient levels, and accessibility, due to changes in sward structure; the interaction between predation rates and management-related changes in habitat; and the impact of alternative anti-helminithic treatments for livestock on invertebrates and birds. There is an urgent need for mechanistic research on interactions between grassland management practices and population dynamics of individual grassland/farmland bird species but more especially of bird species assemblages in those habitats. This research is necessary to underpin the development of grassland management systems that can be incorporated into agri-environment schemes.
The report on which this review is based was funded by The UK Ministry of Agriculture Fisheries and Food. The paper was part-funded under a contract from JNCC (on behalf of English Nature, Scottish Natural Heritage, Countryside Council for Wales and Environment and Heritage Service in Northern Ireland). We thank Su Gough and Nicki Read for help with figures and formatting. Rhys Green, Andy Brown and Davy McCracken, as referees, provided invaluable comments on the manuscript.
Received 11 March 2000; revision received 1 March 2001