Development of an agri-environment scheme option: seed-bearing crops for farmland birds


*Corresponding author. Email:


Winter mortality, resulting from reduced food supply during a period of agricultural intensification, is thought to have driven population declines for some farmland bird species. Planting of game crops has increased over this period in order to provide food and cover for gamebirds. We investigate the potential of this managed habitat for farmland songbird conservation, using intensive single-site studies, and an extensive national survey. Game crops were used more than other farmland habitats by a wide range of bird species. Kale and quinoa were used by many species, whereas maize was used by very few. Cereals such as triticale and millet were used by many species, including several not associated with brassicas such as kale. Crop species differed in the rate of seed shedding, and therefore in the amount of seed food that they provided through the winter. Crop location influenced use by some bird species, with crops close to hedges or other cover generally being favoured. Use of nitrogen fertilizer influenced seed yield, and therefore crop value as a source of food for birds. Our results suggest that, if managed and sited correctly, a combination of two or three crop species can provide a valuable winter food resource for many nationally declining farmland bird species, but further attention needs to be given to their agronomy. This form of management is now incorporated as an option within agri-environment schemes in England, Scotland and Wales. It enables farmers to apply existing skills to conservation and is compatible with their cultural values.

Declines in breeding populations of some farmland passerines are thought to have been driven by increased winter mortality resulting from reduced availability of seed food in modern farming systems. Adult mortality in some form has been suggested as the cause of decline for Song Thrush Turdus philomelos, Yellowhammer Emberiza citrinella, Tree Sparrow Passer montanus, Reed Bunting Emberiza schoeniclus, Blackbird Turdus merula, Chaffinch Fringilla coelebs and Lesser Redpoll Carduelis flammea (Thomson et al. 1997; Siriwardena et al. 1999; Siriwardena et al. 2000; Peach et al. 1999).

Within agricultural ecosystems, abundance of winter food for granivorous birds has declined over the past half-century as livestock and their feed sites have been lost from many farms. Increased autumn cultivation has been associated with loss of winter stubbles on some soils, and improved herbicides have reduced abundance of seeding weeds in any stubbles remaining (Campbell et al. 1997). Donald and Evans (1994), and other studies, have documented the considerable use of stubbles by seed-eating birds on farmland in winter. Hole et al. (2002) have shown that winter seed food resources can influence survival of House Sparrows Passer domesticus, something that may be true for a range of species (Siriwardena & Stevens 2004). Provision of weed-rich cereal stubbles has been a central element of the currently successful restoration of Cirl Buntings Emberiza cirlus in Devon, although management practices designed to benefit Cirl Buntings in the breeding season have also been adopted (Peach et al. 2001).

Loss of livestock from many farming systems has also resulted in a decline in the area of kale, a formerly widely grown fodder crop for cattle and sheep. As well as providing winter feed for livestock, this crop also provides cover over winter for birds, especially game species such as Pheasant Phasianus colchicus, Grey Partridge Perdix perdix and Red-legged Partridge Alectoris rufa. Where these species were valued for shooting, kale was increasingly grown specifically for this purpose from the 1960s, and by the 1970s, additional crops such as sunflower, maize and canary grass were being advocated as cover for gamebirds (Coles 1970). Cereals such as maize and sorghum had long been regarded as important game cover crops in America (Kozicky & Madson 1966). The list of crops recommended for planting in the UK continued to grow through the 1980s.

It was soon realized that game crops provided a habitat for birds other than game species. Instead of re-sowing kale every year, some shoot owners took advantage of its biennial character and left kale plots for a second year. These appeared to support particularly high numbers of birds. Research into the use of game crops by non-game bird species started in the 1990s. Stoate and Szczur (1997) found that Yellowhammers made considerable use of grain in standing cereal crops during the first half of the winter at Loddington (Leicestershire), the Allerton Research & Educational Trust's research and demonstration farm. A range of seed-bearing crop species was subsequently grown in trial plots and their use by birds was monitored as part of a wider conservation initiative (Boatman et al. 2000).

Although planting of seed-bearing crops has been permissible on set-aside under the Wild Bird Cover option since 1993, payments to farmers for growing such crops have only become available since 2002, following the research described in this paper. Options are currently available under England's Countryside Stewardship Scheme, Scotland's Rural Stewardship Scheme, and the Welsh Tir Gofal. This paper reviews research into the use of seed-bearing crops by farmland birds, with emphasis on those species that are declining nationally, and makes recommendations for the management of seed-bearing crops for the conservation of these species.


The research that this paper reviews falls into four main categories. Studies of the use of seed-bearing crops have continued at Loddington, including an experimental study carried out in 1998–2000 at Loddington, a site in Norfolk and another in Hertfordshire (Boatman et al. 2001). Running simultaneously with this project was a national survey carried out by the British Trust for Ornithology, which investigated relative use of various game crops as well as commercial crops on up to 130 farms in each of the three years throughout England (Henderson et al. 2004). In Northern England, a third study investigated the use of seed-bearing crops and the effect on winter bird numbers of introducing such crops to an area that previously lacked them (Stoate et al. 2003). More recently, similar research has been carried out in Scotland, although no attempt was made to investigate the use of individual crop species (Parish & Sotherton 2004, in press). This study assessed general game crop use during the summer as well as in winter, and also investigated the use of these crops by butterflies and bumblebees, two insect taxa known to be declining on farmland in recent decades.

All studies reported greater use of seed-bearing crops by birds than of adjacent commercial crops during winter (October–March). Exceptions to this general rule were Grey Partridge, Skylark Alauda arvensis and Corn Bunting Miliaria calandra, which were also strongly associated with cereal stubbles (Henderson et al. 2004). Parish and Sotherton (2004) found up to 100 times as many birds per hectare in game crops than in commercial crops, which included unmanaged set-aside and cereal stubbles, the farmland habitat traditionally associated with winter food for farmland birds. Henderson et al. (2004) found that bird numbers were 12 times more abundant on game crops than on commercial crops, but that this figure increased to 50 times if only ‘preferred’ game crop types were considered.

On the Raby Estate in County Durham, Stoate et al. (2003) found a bird density of 420/km2 across the farmland landscape when game crops were introduced, compared with the background level of 30–40/km2 for lowland England revealed by data from the national survey. Stoate et al. (2003) also reported an increase in bird abundance when game crops were planted in an area that previously lacked them. Such differences in bird abundance between areas with and areas without seed-bearing crops are likely to vary, depending on the farmland landscape into which they are introduced, on the presence of food provided for gamebirds and on the species composition of the crops grown.


The main focus of attention for researchers has been on identifying appropriate crop species to provide food for the range of nationally declining farmland bird species. Studies have differed slightly in the range of crop species considered, and in the bird communities present. However, together, these research projects have provided a comprehensive understanding of the seed-bearing crops used by a wide range of farmland bird species, including several Biodiversity Action Plan (BAP) species. The main results are summarized in Figure 1. This highlights the use of kale made by several bird species, including the BAP species Song Thrush, Linnet Carduelis cannabina and Reed Bunting, and similarly high use made of quinoa by BAP species Skylark, Tree Sparrow Passer montanus, Lesser Redpoll, Reed Bunting and Corn Bunting.

Figure 1.

Summary of use of seed-bearing crops by farmland birds in winter, based on the studies described in this paper. Black cells represent significant use reported by at least one of the studies reviewed in this paper. Grey cells represent use based on anecdotal information from other sites. Biodiversity Action Plan (BAP) species are in bold type.

Figure 1 reveals that some crop species are used by very few bird species. Other crops that have been recorded as providing food for only one or two species include teasel and evening primrose, which are used only by Goldfinch Carduelis carduelis, and hemp, which is used only by Greenfinch Carduelis chloris (Boatman et al. 2000). Buckwheat, a species sometimes incorporated into game crop mixtures, was rarely used by birds in any of the studies. Chicory was found to be used by Blackbird and Song Thrush in the experimental study (Boatman et al. 2001).

Millet was used by more bird species than other cereal species. Maize was used almost exclusively by gamebirds and Woodpigeons Columba palumus. Some indication of greater use of maize in the national study (Henderson et al. 2004) may be due to the fact that millet is often planted with maize when grown as a game crop and may have been overlooked by volunteer bird watchers carrying out the fieldwork. Barley was also used by Yellowhammers (Stoate et al. 2003) and is likely to be used by other species, but wheat and the wheat/rye hybrid ‘triticale’ are likely to have greater potential as a source of winter food for birds for the reasons given below. Overall, Henderson et al. (2004) showed a weak correlation (r2 = 0.0354) between winter use of game crops by gamebirds and BAP passerines, but with crops such as maize being used more by gamebirds, and others such as kale being used more by passerines (Fig. 2). These results confirm that there is considerable potential to design crop mixtures to meet the requirements of bird communities or individual species within a defined area. Maize and kale are two crops commonly grown as cover for gamebirds, with a maize-to-kale ratio currently of 1.8 in England (J. Ewald pers. comm.), but greater planting of kale than maize would provide greater benefits in terms of food resource for passerines in winter. This is the case in Scotland where a colder climate greatly restricts the growing of maize.

Figure 2.

Crop preferences (mean rank) of Pheasants and Red-legged Partridges with Biodiversity Action Plans (BAP) species (Grey Partridge, Song Thrush, Tree Sparrow, Bullfinch, Linnet, Corn Bunting and Reed Bunting). Crop types are: bare ground (BA), buckwheat (BK), cereals in seed (CE), canary grass (CG), cereal stubble (CS), grassland (GR), 1st-year kale (K1), 2nd-year kale (K2), linseed (LI), millet (ML), mustard (MU), maize (MZ), non-cereals (NC), phacelia (PH), quinoa (QU), rape (RA), sugar beet (SB), sunflowers (SU), teasel (TE), turnips (TU) and winter cereals (WC) (Henderson et al. 2004).

In summer (June–September), Parish and Sotherton (in press) found that game crops supported up to 80 times as many birds as nearby commercial crops. Murray and Stoate (2002) also reported use of Wild Bird Cover crops by breeding Skylark and Yellowhammer, although habitat use varied between species. Yellowhammers made more use of cereals, whereas Skylarks made more use of kale-based crop mixtures. However, breeding birds do not always make greater use of game crops than commercial crops, as recorded for Tree Sparrow at Loddington (C. Stoate unpubl. data).

Whereas, in winter, birds are highly mobile, in summer, foraging range from the nest constrains the availability of Wild Bird Cover crops to breeding birds foraging for invertebrate food for nestlings. At Loddington, maximum foraging ranges (based on observations of provisioning birds) of Yellowhammer, Tree Sparrow and Skylark are approximately 300, 220 and 200 m, respectively. Wild Bird Cover crops that are intended to provide foraging habitat during the breeding season should therefore be distributed across the farm to accommodate these constraints on breeding birds.


Seed-bearing crops are versatile and applicable to many situations. For example, Henderson et al. (2004) found that the difference in densities between the highest ranking winter bird crops and adjacent fields was greater for six bird species (Pheasant, Dunnock Prunella modularis, Tree Sparrow, Linnet, Chaffinch and Reed Bunting) in grassland areas than in arable areas. Seed-bearing crops of this sort could have a special role in grassland areas, effectively introducing arable pockets to grassland landscapes and increasing both the food resource and habitat heterogeneity (Robinson et al. 2001). Henderson et al. (2004) also found that use of seed-bearing crops by Pheasant, Dunnock, Song Thrush, Greenfinch and Yellowhammer was influenced by hedgerow length, and that Tree Sparrow, Greenfinch and Linnet were present in higher numbers where adjacent hedges were relatively tall. Corn Bunting densities were correlated with open field boundaries and lower hedges (Henderson et al. 2004). For optimal use by birds, the location of game crops is therefore an important consideration.

For Grey Partridge, Tree Sparrow and Reed Bunting, Henderson et al. (2004) found some evidence of higher bird densities in game crops that were relatively weedy. However, this has not been reported for other species, and crop seed abundance is likely to be more generally a major determinant of bird abundance. Crops differ in the rate of seed depletion through the winter (Boatman et al. 2001). Sunflower and borage, for example, became depleted early in the autumn, whereas species such as kale and quinoa retained their seed well into the winter. Within the cereals, wheat and triticale retained grain in the ear for longer than barley, and especially oats, which shed grain on ripening.

Probably for this reason, Henderson et al. (2004) found that, although bird densities in seed-bearing crops generally peaked in mid-winter, there was a significant interaction between crop type and count-month. Kale, quinoa and to a lesser extent cereals supported relatively high densities of birds in late winter compared with sunflowers that were largely depleted by December. Between October and February, bird densities declined by 31% in second-year kale compared with 60% in sunflowers. In the experimental study, Greenfinch made considerable use of sunflower in the autumn, but was present at low (but increased) densities on linseed in late winter (Boatman et al. 2001). Stoate et al. (2003) also found that Greenfinch numbers were significantly higher in the first half of the winter than in the second half, whereas Song Thrush numbers were significantly higher in March than during the rest of the winter. Stoate and Szczur (1997) found that Yellowhammers fed in cereals until the supply of grain became exhausted in mid-winter, then fed much more from feed sites where wheat was provided for pheasants. Providing a mixture of food sources may therefore be beneficial for wintering birds, according to which crop species are grown.

Crop structure may also influence use by birds. Stoate and Murray (2002) found that both Yellowhammers and Skylarks used cereal Wild Bird Cover crops during the winter, but that whereas Yellowhammer numbers declined through the winter as grain became depleted, Skylark numbers increased as the crop broke down and the structure became more open.


Parish and Sotherton (in press) found that game crops in Scotland contained about 180% more broad-leaved weed species than commercial crops. These weeds may have contributed to the larger numbers of butterflies and bumblebees found in game crops than in commercial crops, and many of the weed species present support invertebrate groups that are important food for birds during the breeding season. As already noted, weeds can influence positively the abundance of some bird species in winter. However, after growing game crops on the same land for several years, weed cover can reach levels that threaten seed-bearing crops and their ability to produce seed, and therefore meet the objectives of agri-environment schemes. An objective of the ‘winter stubble’ options of agri-environment schemes is to encourage seeding weeds on arable land as a source of seed food for birds in winter. This requires a very much larger area (and therefore funding) to be devoted towards this objective than the production of crops that are higher yielding if managed appropriately. Costs of establishment vary considerably between crops, being lowest for widely grown commercial species such as wheat and linseed.

Soil nutrient depletion can also limit crop seed production after two or three years of cultivation. Stoate et al. (2003) compared winter bird abundance on farmland in a year without and in a year with game crops. Numbers of birds expected to occur were predicted from counts of birds at a nearby site where game crops had been grown for many years. Bird abundance on the study site following introduction of game crops greatly exceeded that predicted from the established site. This is most likely because crop growth and seed production were greater where crops were grown on ‘fresh’ ground with high residual nitrogen, than on the established site where crops had failed completely in a previous year. Previously unpublished data gathered from experimental plots sown with kale at Loddington reveal that kale cover is significantly higher at 90 and 120 kg of nitrogen per hectare than at 30 or 60 kgN/ha (F3,24 = 8.42, P < 0.001; Fig. 3). The effect this may have on seed production is currently being investigated. However, as nitrogen fertilizer use on set-aside Wild Bird Cover and on Countryside Stewardship Scheme Wildlife Seed Mixtures is currently restricted to 30 kgN/ha, there are considerable implications for meeting the objectives of these options in terms of providing winter food for birds. Leaching rates at 90 kgN/ha are likely to be minimal if applications are timed appropriately, as rates of 240 kgN/ha are commonly applied to optimize seed yield of the closely related commercial crop, oilseed rape.

Figure 3.

Kale vegetative cover in July (mean ± se) in relation to nitrogen fertilizer application in four strips of experimental plots at Loddington, Leicestershire.

Given that different crops can often require different management, whether in terms of weed control or fertilizer application, the crop mixtures prescribed under set-aside and agri-environment regulations are also a constraint on successful crop management. For example, current Countryside Stewardship prescriptions require the planting of a mixture of kale, quinoa and triticale, even though drilling dates differ between these crops by more than a month. However, the results presented in this paper suggest that at least two crops are required in order to provide seed food for a range of bird species through the winter. We suggest that it should be permissible to plant seed-bearing crop combinations, in separate single-species strips. This permits seedbed preparation, timing of drilling, pest and weed control, and fertilizer application to be managed according to each crop's needs. As farmers are often frustrated by not being able to grow seed-producing crops effectively, because of legislative constraints, the option to plant combinations of single-species strips could increase farmer commitment to achieving the conservation objectives of seed-bearing crops. Single-species strips may require use of narrower drills than are routinely available on arable farms, but the increasing adoption of seed-bearing crops for gamebirds and within agri-environment schemes is already encouraging greater adoption of narrow drills by farmers and contractors.

Flexibility is also required to enable farmers to select crop species that are appropriate to their local conditions. For example, suitability of millet varies between regions, with this crop being best suited to light soils and southern regions, but soil type and aspect can also influence decisions on crop choice within regions, and even within farms.

Recommendations for best practice therefore might include:

  • • decision on spring- or autumn-sown species according to farming system (e.g. seasonal labour availability) and soil type;
  • • concentration on species that have been shown to provide food for a range of bird species (e.g. kale, quinoa, millet and tritcale);
  • • rotating crops around the farm in order to reduce soil nutrient depletion and weed seedbank, or
  • • application of at least 90 kgN/ha after 2 years and
  • • application of herbicides when necessary to control competing weeds;
  • • planting of combinations of single-species crops in order to facilitate appropriate agronomy.

The ‘Wildlife Seed Mixture’ option is one of many options within the Countryside Stewardship Scheme that provide an opportunity for farmers to create wildlife habitats outside their commercial crop area. A similar option is available in the pilot Entry Level Scheme, which is intended to be applied nationally in England in 2005. Stoate (2004) asked six farmer focus groups to score a range of options from the pilot Entry Level Scheme, according to how applicable they would be to their own farms (given contemporary ELS pilot scheme rewards). The Wildlife Seed Mixture ranked highly (Table 1). Reasons given included compatibility with farmers’ shooting interests, the independence of this habitat from the cropped area and encouragement for songbirds. The agronomy of seed-bearing crops, although an additional management cost, also provides an opportunity for farmers to apply their existing skills and experience to bird conservation on farmland.

Table 1.  Ranking of agri-environment scheme options, according to scores allocated by six farmer focus groups using pairwise comparisons at Loddington (Stoate 2004).
Rank Agri-environment scheme option
  • *

    Options considered by one group only.

 1Field corner
 2Wildlife seed mixture
 36-m margin
 42-m margin
 5Pollen and nectar mixture
 6Skylark plots*
 7Winter stubble*
 9Beetle bank*
10Conservation headland*
11Conservation headland (no fertilizer)*

The results discussed in this paper suggest that maize, one of the most commonly grown game crops in the southern part of England, has minimal conservation value for farmland birds compared with some other crops. Although game and passerine conservation objectives are clearly compatible, some selection of crop species is necessary if optimum benefits are to be achieved. However, farmers’ motivation for successful crop establishment is often influenced by their shooting interests, and by an interest in crops which may have low conservation value (but high visual impact, e.g. sunflower), but which increase farmer commitment to successful crop establishment. Flexibility in crop combinations grown is therefore important to their overall success as a conservation measure. The growing of seed-bearing crops for birds within agri-environment schemes provides an opportunity for farmers to apply and develop their knowledge of crop management and interest in game and other birds to the benefit of farmland bird species.

Although the research reported in this paper has shown that seed-bearing crops can provide an important source of winter food for farmland birds, resulting in local increases in abundance in winter, the impact on overwinter survival and breeding abundance or productivity in subsequent years requires investigation.


Numerous people have contributed to the research reported in this paper. John Szczur carried out the fieldwork at Raby, and Nicholas Aebischer contributed to statistical analysis. Nigel Boatman and Juliet Vickery administered some of the research presented in this paper. Sarah Bence collected the kale data presented in Fig. 2, and Peter Thompson organized sites for the national study. We also acknowledge the help of the Raby Estate and numerous farmers who allowed access to their land, and in some cases grew crops specifically for research purposes. We also thank all the volunteer observers who helped with the national survey.