the impact of growing miscanthus for biomass on farmland bird populations

9
The impact of growing miscanthus for biomass on farmland bird populations P.E. Bellamy*, P.J. Croxton, M.S. Heard, S.A. Hinsley, L. Hulmes, S. Hulmes, P. Nuttall, R.F. Pywell, P. Rothery NERC Centre for Ecology and Hydrology, Monks Wood, Abbots Ripton, Huntingdon, Cambridgeshire PE28 2LS, UK article info Article history: Received 19 March 2008 Received in revised form 18 June 2008 Accepted 22 July 2008 Published online 10 September 2008 Keywords: Breeding season Energy crops Farmland birds Invertebrates Weeds Winter Food supply Shelter abstract Miscanthus is a newly introduced crop grown primarily to produce biomass for energy production and the area grown in the UK is anticipated to increase. Major differences in crop management from conventional arable crops have led to speculation that miscanthus may also have effects on farmland biodiversity. Six miscanthus fields were paired with six of a conventional crop, winter wheat, and bird diversity and abundance were compared in winter and during the breeding season along with potential food sources. Miscanthus fields had a greater abundance and diversity of birds than did wheat in winter and summer. In winter, the greater numbers of birds in miscanthus fields were probably attracted by the shelter provided by the crop and by the abundance of non-crop plants. During the breeding season, the abundance of non-crop plants in miscanthus fields, and greater numbers of insects associated with these plants, provided food resources. However, the miscanthus crop plants provided less insect food than wheat crop plants. Changes in crop structure during the breeding season influenced the breeding birds. The results from this study suggested that an increase in the area of miscanthus grown in the UK may have temporary benefits for farm- land bird populations during establishment. These benefits are likely to diminish with age of crop and as crop management improves with experience. Management for wildlife will be required to maintain the diversity of features attractive to birds because many of these will be lost if miscanthus is managed solely to maximise crop yields. ª 2008 Elsevier Ltd. All rights reserved. 1. Introduction The use of renewable energy is crucial to meet energy and environmental objectives in the European Union. Biomass crops and biofuels have potential to reduce carbon dioxide (CO 2 ) emissions from fossil fuels and help meet targets for mitigating climate change. The renewables’ obligation gives market incentives to UK energy producers to provide increasing amounts of electricity from renewable sources including biomass combustion [1]. Current government incentives aim to stimulate the supply and demand for biomass crops and dramatically increase the areas grown in the UK. This includes planting incentives for farmers to establish energy crops [2]. The main biomass crops grown specifically for energy production are perennial short-rotation coppice willow Salix spp. and miscanthus Miscanthus x giganteus. They differ substantially from conventional crops, in the habitat they provide, crop life-cycle and management. Short-rotation coppice is the longer established crop, and the bird pop- ulations using it have been studied in several countries, e.g. [3–5]. Miscanthus is a relatively recent crop and we have found * Corresponding author. Tel.: þ44 (0) 1487 772553/772400; fax: þ44 (0) 1487 773467. E-mail address: [email protected] (P.E. Bellamy). Available at www.sciencedirect.com http://www.elsevier.com/locate/biombioe 0961-9534/$ – see front matter ª 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.biombioe.2008.07.001 biomass and bioenergy 33 (2009) 191–199

Upload: pe-bellamy

Post on 26-Jun-2016

213 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: The impact of growing miscanthus for biomass on farmland bird populations

b i o m a s s a n d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 9 1 – 1 9 9

Avai lab le a t www.sc iencedi rec t .com

ht tp : / /www.e lsev i er . com/ loca te /b iombioe

The impact of growing miscanthus for biomass on farmlandbird populations

P.E. Bellamy*, P.J. Croxton, M.S. Heard, S.A. Hinsley, L. Hulmes, S. Hulmes, P. Nuttall,R.F. Pywell, P. Rothery

NERC Centre for Ecology and Hydrology, Monks Wood, Abbots Ripton, Huntingdon, Cambridgeshire PE28 2LS, UK

a r t i c l e i n f o

Article history:

Received 19 March 2008

Received in revised form

18 June 2008

Accepted 22 July 2008

Published online 10 September 2008

Keywords:

Breeding season

Energy crops

Farmland birds

Invertebrates

Weeds

Winter

Food supply

Shelter

* Corresponding author. Tel.: þ44 (0) 1487 77E-mail address: [email protected]

0961-9534/$ – see front matter ª 2008 Elsevidoi:10.1016/j.biombioe.2008.07.001

a b s t r a c t

Miscanthus is a newly introduced crop grown primarily to produce biomass for energy

production and the area grown in the UK is anticipated to increase. Major differences in crop

management from conventional arable crops have led to speculation that miscanthus may

also have effects on farmland biodiversity. Six miscanthus fields were paired with six of

a conventional crop, winter wheat, and bird diversity and abundance were compared in

winter and during the breeding season along with potential food sources. Miscanthus fields

had a greater abundance and diversity of birds than did wheat in winter and summer. In

winter, the greater numbers of birds in miscanthus fields were probably attracted by the

shelter provided by the crop and by the abundance of non-crop plants. During the breeding

season, the abundance of non-crop plants in miscanthus fields, and greater numbers of

insects associated with these plants, provided food resources. However, the miscanthus crop

plants provided less insect food than wheat crop plants. Changes in crop structure during the

breeding season influenced the breeding birds. The results from this study suggested that an

increase in the area of miscanthus grown in the UK may have temporary benefits for farm-

land bird populations during establishment. These benefits are likely to diminish with age of

crop and as crop management improves with experience. Management for wildlife will be

required to maintain the diversity of features attractive to birds because many of these will be

lost if miscanthus is managed solely to maximise crop yields.

ª 2008 Elsevier Ltd. All rights reserved.

1. Introduction biomass crops and dramatically increase the areas grown in

The use of renewable energy is crucial to meet energy and

environmental objectives in the European Union. Biomass

crops and biofuels have potential to reduce carbon dioxide

(CO2) emissions from fossil fuels and help meet targets for

mitigating climate change. The renewables’ obligation gives

market incentives to UK energy producers to provide

increasing amounts of electricity from renewable sources

including biomass combustion [1]. Current government

incentives aim to stimulate the supply and demand for

2553/772400; fax: þ44 (0)m (P.E. Bellamy).

er Ltd. All rights reserved

the UK. This includes planting incentives for farmers to

establish energy crops [2].

The main biomass crops grown specifically for energy

production are perennial short-rotation coppice willow Salix

spp. and miscanthus Miscanthus x giganteus. They differ

substantially from conventional crops, in the habitat they

provide, crop life-cycle and management. Short-rotation

coppice is the longer established crop, and the bird pop-

ulations using it have been studied in several countries, e.g.

[3–5]. Miscanthus is a relatively recent crop and we have found

1487 773467.

.

Page 2: The impact of growing miscanthus for biomass on farmland bird populations

b i o m a s s a n d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 9 1 – 1 9 9192

only one previous study of biodiversity associated with its

cultivation at a field-scale [6,7]. Although the first crop trials in

Northern Europe were started in 1983, widespread assessment

of its suitability as an energy crop was not started until the

early 1990s [8], and more recent trials have assessed its suit-

ability for North America [9]. The UK planting rate of mis-

canthus under Defra’s Energy Crops Scheme has increased

from 52 ha in 2002, when it was first planted commercially for

energy production, to 2345 ha in 2006 and the total area of

miscanthus exceeded that of short-rotation coppice by nearly

40% by 2007 [10]. Future policies are likely to encourage the

planting of biomass crops and it is estimated that up to

350,000 ha could be in production by 2020 [11].

Large differences in the agronomy of miscanthus compared

to most annual arable crops have potential to affect bird pop-

ulations. Miscanthus is a tall, perennial grass which although

cropped annually, remains in the field for 15–20 years [12]. It is

harvested in spring rather than late summer/autumn, so there

is a standing crop over winter providing shelter from weather

and predators. After establishment, there are low agricultural

inputs with no routine applications of herbicides, insecticides

or fertilisers [12]. The lack of soil disturbance and low chemical

inputs may also allow non-crop plant and/or invertebrate

populations to persist in greater abundance than in annual

arable crops, potentially benefiting both seed and invertebrate

eating bird species (e.g. [13,14]). It thus seems possible that

growing miscanthus could both improve farmland biodiversity

and help to reduce CO2 emissions. On the other hand, there

have been concerns that the large difference in crop height and

structure may have adverse effects on specialist farmland

birds of open fields [15]. Birds are often a focus of conservation

concern [16], it is thus important to assess the impacts of

potentially large changes in patterns of agriculture, such as the

introduction of biomass crops, on farmland biodiversity from

both a policy and ecological view point [17].

The aim of this study was to quantify the effects on

farmland birds at the field-scale of switching land-use from

intensive annual cereal cropping to commercial production of

miscanthus biomass crops. Bird usage of miscanthus and of

autumn sown wheat fields was compared in both winter and

summer and differences were related to differences in food

resources and cover. A priori, we hypothesised that in winter

miscanthus would provide significantly better shelter and

weed seed resources for birds compared with the adjacent

cereal crop. Similarly, in summer, the miscanthus would

provide nesting habitat and provide a greater abundance and

diversity of invertebrate prey than the cereal crop.

2. Methods

2.1. Study area

The study was conducted in an area within 50 km of CEH

Monks Wood, near Huntingdon, Cambridgeshire in eastern

England (52� 240N, 0� 140W). The land-use of the study area is

mainly intensive arable farming with large field sizes, little

woodland or few hedges. The main crop was winter wheat;

other crops included oilseed rape, sugar beet, potatoes and

other vegetables. All miscanthus crops were managed by Bio-

Renewables Ltd, Cambridgeshire and used as feedstock for

Energy Power Resources’ 38 MW straw burning power station

at Ely, Cambridgeshire. All farms in the study area that had

planted miscanthus in 2004 or earlier were visited from a list of

planting grants made under Defra’s Energy Crop Scheme and

six farms were selected that had apparently well established

crops. Study farms were geographically separated from each

other by at least 1 km. On the selected farms, miscanthus was

grown in blocks of 2–7 adjacent fields. Observations were made

on one field from each block, chosen to provide the densest and

most uniform cover available in order to maximise the crop

differences and give the best possible approximation to

a mature crop. The minimum size of study fields was 3 ha to

maximise numbers of birds per field and reduce edge effects

(mean area¼ 5.9 ha, range 3.0–9.6 ha). Two of the miscanthus

study fields were planted in 2002 and four were planted in 2004.

Each miscanthus field was paired with a nearby winter wheat

field on the same farm (mean area¼ 9.5 ha, range 3.8–24.0 ha,

<0.3 km from miscanthus field) to provide a comparable

conventional arable crop with a similar local/regional bird

species pool. The wheat and miscanthus fields were as similar

as possible in terms of size, shape and boundary type. The

fields were mostly on fenland or fenland edge, with peat (two

sites), silt (two sites) or clay soils (two sites). The same sampling

and census protocols were used on wheat and miscanthus

fields. However, because of the obvious abundance of weeds in

the miscanthus fields additional invertebrate sweep samples

were taken from the non-crop plants below the miscanthus

canopy to separate invertebrate differences due to the crop and

those due to differences in cultivation. Non-crop plants were

largely absent from wheat fields.

2.2. Bird census

Both wintering and breeding birds were censused using the

same method, i.e. a whole area search of each field [18]. On

initial arrival at the field, the whole field was scanned and the

locations of large birds noted on a map of the field before they

were disturbed. The observer then walked slowly around the

perimeterof thefield noting birds in the crop, thefield boundary

and hedges if present. Next, linear transects were walked

through the crop at a spacing of 40–60 m. In miscanthus, the

transects were between the crop rows, but in wheat fields these

followed the tramlines. The directions and distances flown by

flushed birds were marked to reduce the possibility of double

counting. Wintering birds were recorded from three visits:

November/December 2005, January and February 2006.

Breeding birds were recorded from four visits starting after

harvest of the miscanthus in spring: late April/early May, late

May/early June, late June and midJuly 2006. All visits started

shortly after sunrise. The greater vegetation density and poorer

visibility in miscanthus fields are likely to lead to a bias in bird

detectability between crops. To reduce the effects of this bias

more time per hectare was spent censusing miscanthus thus

increasing the probability of detecting birds [19].

2.3. Sampling potential food supply

Earthworms, ground invertebrates, canopy invertebrates and

weeds were sampled as potential food resources for birds. Ten

Page 3: The impact of growing miscanthus for biomass on farmland bird populations

b i o m a s s a n d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 9 1 – 1 9 9 193

samples were taken from each field, five from equally spaced

locations along a short edge of the field and five from a long

edge, each 30 m in from the edge in all but the ground inverte-

brates. Ground invertebrates were sampled with pitfalls in two

linesoffive placedapproximately in the centre ofone shortedge

and one long edge of the field and running perpendicular to the

field edge. The first pitfall was placed 5 m into the crop from the

edge and the distance between subsequent traps was 2 m.

2.3.1. EarthwormsEarthworm (Annelida) abundance was assessed using soil

samples taken in November 2005. Approximately 3.4 L of soil

were taken to a depth of 0.15 m using a small spade. Any

deeper burrowing worms, extracted by pouring a solution of

10 ml of detergent dissolved in 2 L of water into the hole and

waiting for 20 min, were put in labelled bags with the soil

samples. Samples were then stored at 5 �C and processed

within 3 days. Soil samples were searched by hand in the

laboratory and worms extracted were counted as adults or

juveniles based on the presence or absence of a clitellum.

Total live weight of worms was measured to 0.01 g after

rinsing and surface drying.

2.3.2. Ground invertebratesPitfalls were one third filled with a 50% mixture of propylene

glycol and water and covered for protection from rainfall and

detritus. Pitfalls were collected after 2 weeks in the field and

stored in a freezer. The samples were rinsed and the numbers

of invertebrates counted in broad taxonomic groups. The

winter pitfalls were put out in November at the same time as

worm samples were taken. The summer pitfalls were put out

in mid July.

2.3.3. Canopy invertebratesInvertebrates in the canopy of the crops were sampled using

sweep nets during midJuly. At each sample location 10 sweeps

of the crop canopy were made whilst walking through the

crop, along tramlines in wheat and between crop rows in

miscanthus. Samples were bagged and stored in a freezer.

Samples were sorted in the laboratory and the numbers of

invertebrates were recorded in broad taxonomic groups. Every

effort was made to keep sampling efficiency equal between

the two crops with the net brushing through the densest

foliage. Sweep netting was chosen as more practical than the

alternative vacuum sampling and although not tested, the

potential difference in capture efficiency due to differences in

crop structure was thought to be less.

2.3.4. WeedsBoth the seeds and leaves of weeds, defined as non-crop plant

species, can be important food sources for some species of

birds. Weed abundance was assessed using 0.5� 0.5 m quad-

rats in November and July. For each quadrat, the percentage

cover was estimated for each species. The scientific nomen-

clature of plants follows Stace [20].

2.4. Bird shelter

In winter, shelter can be important for feeding birds, reducing

wind chill and obscuring them from predators. In the breeding

season, nest site concealment is also important for reducing

nest predation so shelter as measured here can also represent

the potential for nest concealment as well as protection for

foraging birds. Both density and height of crop can contribute

to the amount of shelter provided. During each bird census

visit, a whole field estimate of mean crop height and crop

cover was made to obtain a broad comparison of the avail-

ability of shelter in the two crops. An additional sample

measure of percentage canopy cover of the crop was also

estimated for ten 5� 5 m quadrats per field, at the same

locations and time as weeds were sampled.

2.5. Analysis

Bird densities were analysed in functional groups based on

taxonomic and ecological similarities, cf. [21,22]. To help

interpret effects of food availability on the birds recorded they

were also analysed in larger groups based on diet, each species

was categorised as mainly herbivorous (including green

vegetation and seeds), insectivorous (including all inverte-

brates), omnivorous or predatory based on information in the

literature [23]. Birds were also categorised by habitat using

national woodland and farmland bird indices [24]. These

habitat groupings are used by UK government and conserva-

tion organisations as indices of habitat specific populations.

Some habitat groups (e.g. wetland birds) are not analysed here

as too few birds occurred in the study and other species are

not habitat specific enough to be included in habitat groups,

although all species were included in totals and dietary

groupings. A full list of species recorded and their diet and

habitat classification is given in Table S1, scientific nomen-

clature of bird species follows Dudley et al. [25].

A comparison of bird numbers between visits within each

season showed there was no pattern of change for any group

during the winter. In summer, warblers increased and pigeons

and game birds decreased in abundance during the breeding

season in miscanthus fields. In all cases there was no shift in

preference during the season and abundance was always

greater in one crop than the other throughout a season.

Therefore, mean abundance of birds across visits within each

season was used for between crop comparisons of bird use.

The statistical significance of comparisons between crops, of

densities of birds and potential food sources, was assessed

using the Friedman test [26], using the crops as treatments

and blocked by site, with a sample size of six paired fields for

all tests. This compared crop differences within sites allowing

for variability in numbers between sites. This non-parametric

equivalent of two way ANOVA was used to provide a consis-

tent test over a range of data types, i.e. % cover, counts and

densities, without the need for different data transformations.

Although some of the data could have been analysed using

ANOVA, non-parametric methods were used throughout to

keep the analysis simple and avoid over interpretation of

a small sample of sites. The results provided a conservative

assessment of the statistical significance of differences

between the two crops. All analyses were undertaken using

Minitab14 [27].

To test whether there was a consistent preference across

all bird species, we calculated the proportion of all species

occurring at a site which was more abundant in the

Page 4: The impact of growing miscanthus for biomass on farmland bird populations

b i o m a s s a n d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 9 1 – 1 9 9194

miscanthus field than in wheat. A Wilcoxon signed rank test

was used to test if significantly more than half of the species

were more abundant in miscanthus than in wheat.

Table 2 – Mean winter abundance of potential food groupsin miscanthus and wheat fields for groups found at fouror more sites.

Potential food groups Mean abundance� SE

Miscanthus Wheat

Soil samples

Adult earthworms

(numbers sample�1)

1.7� 0.35 1.1� 0.28

Juvenile earthworms

(numbers sample�1)

7.1� 1.8 3.9� 0.6

3. Results

3.1. Winter birds

A total of 24 bird species was recorded from miscanthus fields

during winter and 11 species from wheat fields (see Table S1 in

Supplementary Material). The mean number of species per

field was significantly greater in miscanthus (10.0 species)

than in wheat (3.0 species) (n¼ 6, S¼ 6, P¼ 0.031). The total

density of all bird species was greater in miscanthus fields

than wheat fields, but the difference was not statistically

significant (Table 1). The species groups occurring in at least

four of the six sites were, granivorous passerines (chaffinch

Fringilla coelebs, greenfinch Carduelis chloris, linnet Carduelis

cannabina, lesser redpoll Carduelis cabaret, common redpoll

Carduelis flammea, reed bunting Emberiza schoeniclus, yellow-

hammer Emberiza citrinella, corn bunting Emberiza calandra),

game birds (pheasant Phasianus colchicus, red-legged partridge

Alectoris rufa, grey partridge Perdix perdix), thrushes (blackbird

Turdus merula, song thrush Turdus philomelos, redwing Turdus

iliacus, fieldfare Turdus pilaris), skylark Alauda arvensis, snipes

(snipe Gallinago gallinago and woodcock Scolopax rusticola), and

raptors (sparrowhawk Accipiter nisus, kestrel Falco tinnunculus,

merlin Falco columbarius). Game birds, snipes and granivorous

passerines were found at significantly higher densities in

miscanthus (Table 1). Only skylark was more abundant in

wheat, although this difference was not statistically signifi-

cant. All groups except game birds and granivorous passerines

were found in low numbers. Most of the frequently occurring

species in miscanthus were woodland species, whereas no

Table 1 – Mean winter abundance of bird species recordedin miscanthus and wheat fields for species groupsoccurring at four or more of the six sites (for species ineach group see Section 2 and Table S1).

Species Mean density� SE (birds ha�1)

Miscanthus Wheat

Granivorous passerines 2.62� 0.08 0

Game birds 1.47� 0.34 0.04� 0.02

Thrushes 0.46� 0.16 0.14� 0.14

Skylark 0.04� 0.02 0.41� 0.21

Snipes 0.23� 0.08 0.07� 0.07

Raptors 0.10� 0.03 0.01� 0.01

All herbivores 3.0� 1.6 2.1� 1.7

All insectivores 0.69� 0.17 0.22� 0.15

All omnivores 1.2� 0.28 0.38� 0.32

All predators 0.10� 0.03 0.02� 0.01

Woodland spp. 2.4� 1.2 0

Farmland spp. 0.86� 0.51 2.1� 1.7

All birds 5.0� 1.8 2.7� 1.8

Statistically significant differences (P< 0.05) between crops are in

bold.

woodland species were found in wheat. More than half of the

species occurring at a site were more abundant in miscanthus

(median¼ 79%, n¼ 6, T¼ 21, P¼ 0.018, Wilcoxon signed rank

test). Although there is potentially lower detectability of birds

in the taller miscanthus, the results show higher densities in

miscanthus. Thus the patterns of difference would not be

affected, even if the densities of some species were

underestimated.

3.2. Winter food supply

There was no consistent difference between crops in the

numbers of invertebrates from pitfall samples or earthworms

from soil cores (Table 2). Although numbers and biomass of

earthworms were greater in miscanthus this difference was

not statistically significant probably due to variability between

sites in soil type and age of miscanthus crop.

There was no significant difference in the total number of

invertebrates found in pitfall traps or in numbers for most of

the individual invertebrate taxa. However, slugs and snails

(Mollusca) and millipedes (Diplopoda) were more than twice

as abundant in miscanthus fields, and flies (Diptera) and

Wet weight

earthworms (g)

1.8� 0.5 1.1� 0.19

Pitfall samples (numbers sample�1)

Collembola 19� 8.3 12� 8.1

Diptera 1.3� 0.51 6.5� 3.8

Coleoptera 2.9� 0.36 4.4� 1.4

Mollusca 4.6� 1.5 2.0� 1.2

Insect larvae 3.1� 1.6 2.3� 0.53

Aranaea 2.6� 0.43 1.8� 0.58

Acarina 0.72� 0.47 1.1� 0.72

Diplopoda 1.1� 0.88 0.17� 0.09

Annelida 0.18� 0.09 0.49� 0.18

All invertebrates 37� 10 31� 14

Plants (% cover)

Crop 36� 6.1 31� 7.9

Poaceae 16� 9.3 0.03� 0.02

Asteraceae 5.8� 3.0 0.24� 0.23

Polygonaceae 7.9� 6.8 0

Onagraceae 3.0� 1.5 0

Scrophulariaceae 1.1� 0.87 0.02� 0.01

Violaceae 0.20� 0.16 0.03� 0.02

All weeds 38� 8.6 0.40� 0.27

Numbers are the mean abundance from 10 samples in each field.

Statistically significant differences (P< 0.05) between crops are in

bold. Details of all invertebrates and non-crop plants recorded are

in Table S2.

Page 5: The impact of growing miscanthus for biomass on farmland bird populations

Table 3 – Mean breeding season abundance of birdspecies groups and density of individual bird speciesrecorded breeding (territories 50–100% within the crop) inmiscanthus and wheat fields, for species occurring at fouror more of the six sites.

Species Mean density� SE

Miscanthus Wheat

Birds using fields (birds ha�1)

Game birds 0.81� 0.28 0.16� 0.05

Granivorous passerines 0.53� 0.18 0.10� 0.04

Skylark 0.46� 0.17 0.46� 0.20

Warblers 0.38� 0.11 0

Pigeons 0.22� 0.11 0.03� 0.02

Crows 0.21� 0.12 0.03� 0.03

All herbivores 0.89� 0.17 0.10� 0.05

All insectivores 1.5� 0.29 0.63� 0.26

All omnivores 0.72� 0.21 0.15� 0.05

b i o m a s s a n d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 9 1 – 1 9 9 195

earthworms (Annelida) were more than twice as abundant in

wheat fields. Of these, only the difference in earthworms,

which occurred in low abundance, was significant.

There was a large and significant difference in the total

abundance of non-crop plants between crops, with 38.2%

cover in miscanthus fields compared with 0.4% in wheat

(Table 2). Species representing 19 plant families were found in

miscanthus, and 10 families in wheat (see Table S2). Of these,

11 families were only found in miscanthus fields, but only two

were unique to wheat fields. Of the six most frequently found

families, all were more abundant in miscanthus fields than in

wheat (Table 2), though not statistically significant for two

families, Violaceae (Viola arvensis) and Scrophulariaceae

(mostly Veronica persica), which occurred at low abundance.

Most of the non-crop plants in wheat fields were young

seedlings, compared with a majority of senescent plants with

seed heads in miscanthus.

Woodland spp. 0.16� 0.07 0.01� 0.01

Farmland spp. 1.35� 0.20 0.65� 0.25

All birds 3.1� 0.29 0.87� 0.22

Breeding pairs (pairs ha�1)

Pheasant (5,4) Phasianus colchicus 0.38� 0.19 0.15� 0.06

Skylark (4,4) Alauda arvensis 0.33� 0.12 0.24� 0.10

Red-legged partridge (6,1) Alectoris rufa 0.30� 0.09 0.04� 0.04

Reed bunting (4,3) Emberiza schoeniclus 0.18� 0.07 0.07� 0.04

Reed warbler (5,0) Acrocephalus scirpaceus 0.41� 0.14 0

3.3. Winter shelter

Although the percentage cover of crops did not differ between

crops in winter (Table 2), the miscanthus crop was much taller

(miscanthus, mean height¼ 2.1� 0.15 m; wheat, mean

height¼ 0.08� 0.01 m; n¼ 6, S¼ 6, P< 0.031). This greater

height, combined with the greater cover of non-crop plants,

would provide more shelter in miscanthus fields.

All species 1.8� 0.12 0.59� 0.12

Other species breeding less frequently: meadow pipit Anthus pra-

tensis (3,0), sedge warbler Acrocephalus schoenobaenus (3,0), yellow

wagtail Motacilla flava (0,1), yellowhammer Emberiza citrinella (1,0).

Bird abundance was also grouped on diet and habitat (for species in

each group see Table S1). The number of miscanthus and wheat

fields, respectively, that each breeding species was recorded is

given in brackets after the species name. Statistically significant

differences (P< 0.05) between crops are in bold.

3.4. Breeding birds

A total of 24 bird species was recorded from miscanthus fields

during the breeding season and 12 species from wheat fields

(see Table S1). The mean number of species per field was

greater in miscanthus (2.0 species) than in wheat (0.68 species)

(n¼ 6, S¼ 6, P¼ 0.031). The species groups occurring in at least

four of the six sites were game birds (pheasant, red-legged

partridge), granivorous passerines (linnet, reed bunting,

yellowhammer, corn bunting), skylark, warblers (reed warbler

Acrocephalus scirpaceus, sedge warbler Acrocephalus schoeno-

baenus, whitethroat Sylvia communis), pigeons (woodpigeon

Columba palumbus, stock dove Columba oenas), and crows

(carrion crow Corvus corone, rook Corvus frugilegus, jackdaw

Corvus monedula, magpie Pica pica, jay Garrulus glandarius). The

overall number of birds was higher in miscanthus fields than

in wheat fields and for all species groups except skylark (Table

3). However, the only group in which the difference was

statistically significant was warblers, which occurred in all

miscanthus fields, but no wheat fields. Numbers of both

farmland and woodland birds were greater in miscanthus

(Table 3). Most species occurring at each site during the

breeding season were more abundant in miscanthus

(median¼ 78%, n¼ 6, T¼ 21, P¼ 0.018, Wilcoxon signed rank).

The species recorded were a mixture of those breeding in the

crop and the adjacent boundary (e.g. skylark and reed bunting)

and other species which were feeding on the fields, but bred in

the wider area (e.g. woodpigeon and linnet). As in winter, the

potentially lower detectability of some birds in the taller

miscanthus crop may mean that recorded densities are lower

than actually present. However, this does not affect the find-

ings as the results indicate higher densities in miscanthus.

Considering only birds whose breeding territories were

completely within the crop or half in the crop and half in the

adjacent field boundary, a total of seven species were found

breeding in miscanthus and five species in wheat (Table 3).

There was a greater density of breeding pairs (mis-

canthus¼ 1.8� 0.12 pairs ha�1; wheat¼ 0.59� 0.12 pairs ha�1;

n¼ 6, S¼ 6, P¼ 0.031) and breeding species (miscanthus¼ 0.92

� 0.14 species ha�1; wheat¼ 0.28� 0.05 species ha�1; n¼ 6,

S¼ 6, P¼ 0.031) in miscanthus than in wheat fields. Two

species were found breeding at significantly higher densities

in miscanthus and none were found at higher densities in

wheat (Table 3).

3.5. Breeding food supply

There was no clear difference in ground invertebrate abun-

dance between the crops. Although there were greater total

numbers of invertebrates from pitfall traps in miscanthus

than in wheat fields, this was not statistically significant

(Table 4). Of the individual taxa, springtails (Collembola), slugs

and snails (Mollusca), aphids (Hemiptera) and centipedes

(Chilopoda) were more than twice as abundant in miscanthus

fields, although not statistically significant for mollusca and

Page 6: The impact of growing miscanthus for biomass on farmland bird populations

Table 4 – Mean summer abundance of potential foodgroups in miscanthus and wheat fields for groups foundat four or more sites.

Potential food groups Mean abundance� SE

Miscanthus Miscanthus Wheat

Non-cropplants

Crop Crop

Sweep samples (numbers sample�1)

Hemiptera 28.5� 5.1 5.6� 1.5 19� 6.0

Diptera 6.8� 1.7 1.7� 0.37 2.0� 0.22

Hymenoptera 4.8� 0.91 1.9� 0.75 1.2� 0.34

Thysanoptera 1.2� 0.45 3.6� 1.2 3.9� 2.4

Coleoptera 2.8� 0.78 0.50� 0.16 4.4� 2.1

Mollusca 2.7� 2.6 0.53� 0.51 0.65� 0.65

Collembola 1.6� 0.88 1.5� 1.2 0.20� 0.13

Aranaea 1.8� 0.05 0.35� 0.08 0.63� 0.24

Insect larvae 1.4� 0.53 0.30� 0.18 1.00� 0.25

Neuroptera 0.07� 0.03 0 0.03� 0.03

All invertebrates 52� 7.4 16� 2.8 33� 5.8

Pitfall samples (numbers sample�1)

Collembola – 26.7� 7.5 5.4� 2.6

Coleoptera – 14� 5.5 13� 5.3

Aranaea – 9.5� 1.4 15� 4.1

Diptera – 5.3� 1.6 7.5� 1.2

Mollusca – 7.2� 3.9 0.50� 0.36

Hymenoptera – 5.9� 1.2 4.8� 1.3

Hemiptera – 5.6� 0.86 2.6� 1.2

Opilliones – 2.4� 1.6 2.4� 1.3

Isopoda – 2.3� 1.0 1.2� 0.7

Insect larvae – 1.2� 0.84 1.4� 0.80

Chilopoda – 0.23� 0.09 0.02� 0.02

Diplopoda – 0.13� 0.06 0.12� 0.10

Thysanoptera – 0.02� 0.02 0.13� 0.08

All invertebrates – 80� 16 54� 10

Plants (% cover)

Crop – 67� 4.0 91� 1.2

Asteraceae – 35� 6.1 0.08� 0.08

Poaceae – 14� 3.2 0

Onagraceae – 10� 4.7 0

All non-crop plants – 59� 6.6 0.12� 0.12

Numbers are the mean abundance from 10 samples in each field.

Statistically significant differences (P< 0.05) between crops are in

bold.

Date

Cro

p h

eig

ht (m

)

1 July1 June1 May

2.5

2.0

1.5

1.0

0.5

0.0

miscanthus

wheat

a

Date

Cro

p co

ver (%

)

1 July1 June1 May

100

80

60

40

20

0

b

miscanthus

wheat

Fig. 1 – Changes in (a) crop height and (b) crop cover during

the breeding season. Lowess lines are shown to indicate

trends.

b i o m a s s a n d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 9 1 – 1 9 9196

hemiptera (Table 4). Only thrips (Thysanoptera) were more

than twice as abundant in wheat fields, although this was not

significant and numbers were small (Table 4).

The numbers of invertebrates found in the crop canopy by

sweep net sampling showed a different pattern of abundance

to that of ground invertebrates, with significantly more

invertebrates in the wheat crop than in the miscanthus crop

(Table 4). However, there were even more invertebrates in the

weeds within miscanthus fields than in either crop (Table 4).

Most invertebrate taxa occurred in greatest abundance in the

weed under-storey of the miscanthus fields, except Thysa-

noptera (thrips) and Coleoptera (mostly weevils) which were

at their most abundant in the wheat crop (Table 4). When only

comparing the two crops, Collembola were more than twice as

abundant in the miscanthus, although the difference was not

statistically significant, and Hemiptera, Coleoptera and insect

larvae were more than twice as abundant in wheat (Table 4).

There were very few non-crop plants present in the wheat

fields in contrast to miscanthus fields which had a much

greater cover (Table 4). Species from 13 plant families were

found in miscanthus fields, although only three families were

widespread, and only four plant families were found in wheat

(see Table S2). The most abundant non-crop plants were

members of the Asteraceae (mostly thistles Cirsium spp. and

sow thistles Sonchus spp.), the other families were Poaceae

(mostly Poa annua and Alopecurus myosuroides) and Onagraceae

(willowherbs Epilobium spp.). The lower diversity found in

summer compared with winter wasprobablydue to thehot,dry

summer in 2006 reducing the presence of short lived annuals.

3.6. Breeding shelter

The wheat fields provided a dense and uniform cover of crop,

the only breaks in the canopy being along the tramlines. There

were relatively small changes over the breeding season in crop

height and cover for wheat (Fig. 1). In contrast, miscanthus

had a low (mean height¼ 0.32� 0.07 m) and open (mean

cover¼ 17� 11 %) structure in May which rapidly increased in

both height and cover over the breeding season (mean

height¼ 2.2� 0.08 m and mean cover¼ 73� 10 % in July)

(Fig. 1). There was a decline in the numbers of birds recorded

in miscanthus as the crop grew (correlation between total bird

density and crop height, n¼ 24, r¼�0.66, P¼ 0.001). Two

species in particular showed a strong relationship with crop

height, red-legged partridge numbers declined as the crop

grew (n¼ 24, r¼�0.56, P¼ 0.005) and reed warbler numbers

increased (n¼ 24, r¼ 0.58, P¼ 0.003). Reed warblers were not

found in the crop at all until it was greater than 1 m tall, even

Page 7: The impact of growing miscanthus for biomass on farmland bird populations

b i o m a s s a n d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 9 1 – 1 9 9 197

though they occurred in neighbouring rape fields and adjacent

vegetated ditches. Red-legged partridges are birds of open

habitats so the fields are likely to be less attractive as the crop

grows. However, there is also likely to be a sampling effect of

reduced detectability for this ground dwelling species.

4. Discussion

4.1. Bird assemblages of miscanthus

For all species combined, the densities of birds in miscanthus

fields during the breeding season were similar to those found

in short-rotation coppice willow [5] and set-aside fields [21]

which in turn had greater bird densities than the conventional

arable crops in these other studies. For individual species, the

densities recorded in miscanthus were similar to the highest

values reported in studies from other crops, e.g. for skylark,

highest densities in set-aside and short-rotation coppice

[5,28,29] and for reed warbler and reed bunting in oilseed rape

[30,31]. Thus miscanthus within the first 5 years of establish-

ment may constitute a crop of potentially high value for birds

within intensively farmed lowland landscapes.

In winter, the bird species using the two crops were very

different. Woodland birds predominated in miscanthus

reflecting the extra shelter provided by the standing crop,

whereas wheat fields had mainly farmland birds. There was

relatively little overlap between the two crops, only seven of

the 28 species recorded occurred in both. During the summer,

there was less of a difference in species using the two crops, 11

of the 25 species occurring in both. During the breeding

season, there were many fewer woodland birds in mis-

canthus, but widespread farmland species were found in both

crops. The presence of some birds of open habitats in mis-

canthus fields may be explained by the presence of areas of

short and/or thin crop within most fields. All of the breeding

species were species that regularly breed in fields and field

boundaries in the study area. Thus, miscanthus did not

appear to attract new species, but provided additional

resources for those already present in the landscape.

4.2. Miscanthus as winter bird habitat

In winter, most birds were ground feeding species with

a range of diets. Of the 24 bird species found in miscanthus

fields 12 were mainly herbivorous in winter, of which eight

were predominantly seed-eating finches and buntings recor-

ded only in miscanthus fields. This predominance of herbiv-

orous species reflected the greater abundance of non-crop

plants found in the miscanthus fields. However, not all non-

crop plants species are equally important in supporting seed-

eating birds. Five plant families (Poaceae, Polygonaceae,

Chenopodiaceae, Brassicaceae and Caryophyllaceae) have

been identified as important in the diets of farmland birds in

winter [32,33]. Of these, only Poaceae and Polygonaceae were

abundant or widespread in the miscanthus fields studied

here, the other important plant families were only recorded

from a few fields. The local distribution of many of the weeds

was probably related to site specific factors such as soil type

and previous cropping regimes [34]. The distribution of

important non-crop plant families between fields was reflec-

ted in the distribution of seed-eating birds; flocks of seed

eaters were only observed in the two fields with the greatest

abundance of non-crop plants important in bird diets. Two

other plant families which occurred in all miscanthus fields

were Compositeae and Onagraceae which are mostly wind

dispersed, small-seeded species and are used as a food source

by fewer bird species. Weed seed resources in miscanthus are

likely to decline over time with a lack of tillage and increasing

competition from the crop.

The link between invertebrate food supply and bird abun-

dance was less clear. There was no significant difference

between crops in the abundance of invertebrate taxa. The

apparent differences in earthworm results between soil cores

and pitfalls are due to the fact that pitfalls measure surface

activity, rather than abundance, which can be related to worm

species and food availability, i.e. worms travelling further to

feed in areas of low organic detritus. Of the most important

taxa (Coleoptera, Hemiptera, Lepidoptera and Annelida) in

winter farmland bird diets outside the breeding season [33]

only Coleoptera and Annelida were recorded regularly in the

study fields. Six of the bird species recorded in miscanthus

fields were mainly invertebrate feeders in winter, compared to

three species in wheat fields. The presence of woodcock and

three thrush species only in miscanthus fields was more likely

to be related to the shelter provided and more favourable

foraging conditions (e.g. damp soil) than the slightly higher

worm abundance. However, the lack of tillage and a possible

increase in soil organic matter in miscanthus fields may

favour an increase in soil invertebrate numbers and diversity

with time [35,36]. Therefore the results suggested that in

winter the key role of miscanthus plantations in these open,

intensively managed landscapes was the provision of shel-

tered foraging refuges.

4.3. Miscanthus as breeding habitat for birds

During the breeding season many seed eaters also feed them-

selves and their young on invertebrates. Five of the species

recorded were mainly herbivorous during the breeding season

and all were more abundant in miscanthus. Six invertebrate

taxa (Coleoptera, Orthoptera, Diptera, Lepidoptera, Hyme-

noptera, Hemiptera and Aranaea) have been found to be

important in diets of farmland birds in the breeding season

[32,33]. There was little difference between crops in the

numbers of any of these taxa found in pitfalls, and the only two

taxa (Collembola and Chilopoda) which where significantly

more abundant in miscanthus were eaten by few species and

were not important for bird diets. However, of the invertebrate

taxa important for bird diets, all except Coleoptera were more

abundant in sweep net samples from the non-crop plants than

either crop. Coleoptera (mainly Curculionidae) were most

abundant in wheat, while Lepidoptera and Orthoptera were

infrequent in all sweep net samples. Thus the non-crop plants

in miscanthus fields supplied both direct (seeds and leaves)

and indirect (invertebrates) food for farmland birds.

During the breeding season, the initial crop structure and

rapid growth of miscanthus increased the range of bird

species present. For example, red-legged partridges preferred

the early open structure while reed warbler colonised the

Page 8: The impact of growing miscanthus for biomass on farmland bird populations

b i o m a s s a n d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 9 1 – 1 9 9198

taller crop. Heterogeneity in crop height and cover within

fields throughout the season can also increase the range of

species using them. For example, gaps in the canopy and areas

of low, sparse crop allowed species such as skylark and

meadow pipit Anthus pratensis, which usually avoid tall, dense

cover, to continue to use the fields as the crop matured. The

results suggested that during the breeding season the main

benefits of miscanthus plantations were the provision of

invertebrate and plant food resources and suitable nest sites.

4.4. The potential impact of large-scale miscanthusplanting on farmland bird populations

A number of species recorded from miscanthus fields were

Red List Species of High Conservation Concern [37]. These

included two of the most frequent breeding species, skylark

and reed bunting, which occurred at slightly higher densities

than in wheat. Linnets were also regularly found feeding in

miscanthus fields during the first half of the breeding season.

Another four Red List Species (grey partridge, song thrush,

yellowhammer and corn bunting) occasionally used mis-

canthus in winter and/or summer. Increased acreage of mis-

canthus is unlikely to have a major impact on halting the

declines of these species due to the small percentage of land

under miscanthus, but their greater incidence in miscanthus

compared to wheat suggests that effects are more likely to be

beneficial than detrimental during the first 5 years of crop

establishment. However, more information will be required

before the consequences of large-scale and widespread agri-

cultural production of miscanthus can be properly assessed

for bird populations and other wildlife. The only other pub-

lished study of wildlife in miscanthus also found few insects

on the crop plants but high diversity and numbers on weeds

within the fields [7]. There have been concerns that the dense

growth of miscanthus will shade out all weeds, create a deep

litter layer and the fields will become impenetrable for most

birds. The weed flora of miscanthus fields will inevitably

decline over time, but as this is a new crop, studies of the

mature crops have yet to be undertaken to establish the long

term consequences of growing miscanthus; the oldest fields in

this study were in their fifth growing season. Also, there will

always be a proportion of the miscanthus fields being grown

in the establishment phase and these seem to be beneficial for

weeds, insects and birds.

As with all novel crops both aspects of the crop, e.g.

palatability of leaves or seeds to invertebrates or birds, and

crop management, e.g. types of herbicide/insecticide applied

or cutting times, can affect farmland wildlife. This study

suggests that the effects on birds found here were due mainly

to crop management. The crop itself did not provide much

food, but provided nesting structure for some species and

foraging cover in winter for others. However, the management

of the crop allowed the growth of non-crop plants within the

miscanthus crop which were important as both a direct and

indirect source of food for birds and the presence of these non-

crop plants depended on gaps and low density areas within

the crop. The crop structure and heterogeneity within fields

were also important for providing nest sites for a variety of

species. Even relatively small changes in management for

different varieties of the same crop can lead to significant

differences in the abundance of weed seeds available for birds

[38]. If the benefits for biodiversity found during establish-

ment in the first 5 years are to be maintained in the mature

crop then effective agri-environment options need to be

developed for miscanthus. Options could include manage-

ment of the wide headlands between the crop and field

boundaries to provide seed food in winter or leaving unplan-

ted patches within the crop, similar to unsown patches in

cereal crops for nesting skylarks [39]. Active management

using seed mixes on headlands could also target pollinators

and nectar feeding insects. If miscanthus is managed solely to

maximise crop height, density and uniformity, much of the

value of this crop for wildlife reported here will be lost, and

especially so for breeding birds.

The need for good soil to establish miscanthus means that

it is normally grown on fields previously under arable crop-

ping. The effect on farmland birds at the landscape scale will

depend on which arable crops it replaces and the extent of the

blocks planted. If large areas of miscanthus are planted in

regions which hold vulnerable populations of birds typical of

open fields, e.g. yellow wagtail, corn bunting or grey partridge,

this could have detrimental effects on these populations. As

miscanthus becomes established over wider areas, studies of

the effects on bird populations at the larger scale will be

needed to assess possible detrimental or beneficial effects on

farmland bird populations.

Acknowledgements

We would like to thank Mark Stevenson from Defra for the

information detailing sites in the energy crops scheme and

the landowners and managers who gave us access to their

fields. This work was funded by the UK Energy Research

Centre and the Natural Environment Research Council.

Appendix.Supplementary data

Supplementary data associated with this article can be found,

in the online version, at doi:10.1016/j.biombioe.2008.07.001.

r e f e r e n c e s

[1] Anonymous. The renewables obligation order 2006. StatutoryInstrument 2006 No. 1004. London: HMSO; 2006.

[2] Anonymous. Energy Crops Scheme establishment grantshandbook. London: Defra, http://www.naturalengland.org.uk/planning/grants-funding/energy-crops/docs/ecs-handbook.pdf; 2007 [accessed 13.06.08].

[3] Berg A. Breeding birds in short-rotation coppices onfarmland in central Sweden – the importance of Salix heightand adjacent habitats. Agriculture Ecosystems &Environment 2002;90:265–76.

[4] Dhondt AA, Wrege PH, Sydenstricker KV, Cerretani J. Clonepreference by nesting birds in short-rotation coppice

Page 9: The impact of growing miscanthus for biomass on farmland bird populations

b i o m a s s a n d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 9 1 – 1 9 9 199

plantations in central and western New York. Biomass &Bioenergy 2004;27:429–35.

[5] Sage R, Cunningham M, Boatman N. Birds in willow short-rotation coppice compared to other arable crops in centralEngland and a review of bird census data from energy cropsin the UK. Ibis 2006;148(Suppl. 1):184–97.

[6] Semere T, Slater FM. Ground flora, small mammal and birdspecies diversity in miscanthus (Miscanthus x giganteus) andreed canary-grass (Phalaris arundinacea) fields. Biomass &Bioenergy 2007;31:20–9.

[7] Semere T, Slater FM. Invertebrate populations in miscanthus(Miscanthus x giganteus) and reed canary-grass (Phalarisarundinacea) fields. Biomass & Bioenergy 2007;31:30–9.

[8] Jorgensen U, Schwarz K-U. Why do basic research? A lessonfrom commercial exploitation of miscanthus. NewPhytologist 2000;148:190–3.

[9] Heaton EA, Clifton-Brown J, Voigt TB, Jones MB, Long SP.Miscanthus for renewable energy generation: Europeanunion experience and projections for Illinois. Mitigation andAdaptation Strategies for Global Change 2004;9:433–51.

[10] Anonymous. Area statistics for non-food crops. New York:National Non-Food Crops Centre, http://www.nnfcc.co.uk/metadot/index.pl?id¼2179;isa¼Category;op¼show; 2007[accessed 18.06.08].

[11] Anonymous. UK biomass strategy. London: Defra, http://www.defra.gov.uk/environment/climatechange/uk/energy/renewablefuel/pdf/ukbiomassstrategy-0507.pdf; 2007[accessed 18.06.08].

[12] Anonymous. Growing miscanthus best practice guidelines.London: Defra, http://www.defra.gov.uk/erdp/pdfs/ecs/miscanthus-guide.pdf; 2007 [accessed 18.06.08].

[13] Taylor RL, Maxwell BD, Boik RJ. Indirect effects of herbicideson bird food resources and beneficial arthropods. AgricultureEcosystems & Environment 2006;116:157–64.

[14] Smith J, Potts SG, Woodcock BA, Eggleton P. Can arable fieldsmargins be managed to enhance their biodiversity,conservation and functional value for soil macrofauna?Journal of Applied Ecology 2008;45:269–78.

[15] Anderson GQA, Haskins LR, Nelson SH. The effects ofbioenergy crops on farmland birds in the United Kingdom –a review of current knowledge and future predictions. In:Parris K, editor. Biomass and agriculture: sustainability,markets and policies. Paris: OECD; 2004. p. 199–218.

[16] Gregory RD, Noble DG, Custance J. The state of farmlandbirds: population trends and conservation status oflowland farmland birds in the United Kingdom. Ibis 2004;146(Suppl. 2):1–13.

[17] Sutherland WJ, Armstrong-Brown S, Armsworth PR,Brereton T, Brickland J, Campbell CD, et al. The identificationof 100 ecological questions of high policy relevance in theUK. Journal of Applied Ecology 2006;43:617–27.

[18] Atkinson PW, Fuller RA, Gillings S, Vickery JA. Counting birdson farmland habitats in winter. Bird Study 2006;53:303–9.

[19] Slater PJ. Factors affecting the efficiency of the area searchmethod of censusing birds in open forests and woodlands.Emu 1994;94:4158–97.

[20] Stace C. New flora of the British Isles. 2nd ed. Cambridge:Cambridge University Press; 1997.

[21] Henderson IG, Vickery JA, Fuller RJ. Summer bird abundanceand distribution on set-aside fields on intensive arable farmsin England. Ecography 2000;23:50–9.

[22] Bradbury RB, Browne SJ, Stevens DK, Aebischer NJ. Five-yearevaluation of the impact of the arable stewardship pilotscheme on birds. Ibis 2004;146(Suppl. 2):171–80.

[23] Cramp S, Simmons KEL, Perrins CM, editors. The birds of thewestern palearctic, vols. I–IX. Oxford: Oxford UniversityPress; 1977–1994.

[24] Anonymous. e-Digest statistics about: wildlife – birds.London: Defra, http://www.defra.gov.uk/environment/statistics/wildlife/wdbirdspop.htm#witb10; 2006 [accessed18.06.08].

[25] Dudley SP, Gee M, Kehoe C, Melling TM, BritishOrnithologists’ Union Records Committee. The British list:a checklist of birds of Britain. 7th ed. Ibis 2006;148:526–63

[26] Siegal S. Nonparametric statistics for the behaviouralsciences. 2nd ed. London: McGraw-Hill; 1988.

[27] Ryan BF, Joiner BL, Ryan T. Minitab handbook. 4th ed.Florence, KY: Brooks Cole; 2000.

[28] Poulsen JG, Sotherton NW, Aebischer NJ. Comparativenesting and feeding ecology of skylarks Alauda arvensis onarable farmland in southern England with special referenceto set-aside. Journal of Applied Ecology 1998;35:131–47.

[29] Mason CF, MacDonald SM. Influence of landscape and land-use on the distribution of breeding birds in farmland ineastern England. Journal of Zoology 2000;251:339–48.

[30] Burton NHK, Watts PN, Crick HQP, Edwards PJ. The effects ofpreharvesting operations on reed buntings Emberizaschoeniclus nesting in oilseed rape Brassica napus. Bird Study1999;46:369–72.

[31] Gruar D, Barritt D, Peach WJ. Summer utilization of oilseedrape by reed buntings Emberiza schoeniclus and otherfarmland birds. Bird Study 2006;53:47–54.

[32] Wilson JD, Morris AJ, Arroyo BE, Clark SC, Bradbury RB. Areview of the abundance of invertebrate and plant foods ofgranivorous birds in northern Europe in relation toagricultural change. Agriculture, Ecosystems andEnvironment 1999;75:13–30.

[33] Holland JM, Hutchison MAS, Smith B, Aebischer NJ. A reviewof invertebrates and seed-bearing plants as food forfarmland birds in Europe. Annals of Applied Biology 2006;148:49–71.

[34] Moorcroft D, Whittingham MJ, Bradbury RB, Wilson JD. Theselection of stubble fields by wintering granivorous birdsreflects vegetation cover and food abundance. Journal ofApplied Ecology 2002;39:535–47.

[35] Beuch S, Boelke B, Belau L. Effect of the organic residues ofMiscanthus x giganteus on the soil organic matter level ofarable soils. Journal of Agronomy and Crop Science 2000;18:111–20.

[36] Kahle P, Belau L, Boelke B. Effects of 10 years of miscanthuscultivation on different properties of mineral soil in north-east Germany. Journal of Agronomy and Crop Science 2002;188:43–50.

[37] Gregory RD, Wilkinson NI, Noble DG, Robinson JA,Brown AF, Hughes J, et al. The population status of birds inthe United Kingdom, channel islands and the Isle of man:an analysis of conservation concern 2002–2007. British Birds2002;95:410–50.

[38] Gibbons DW, Bohan DA, Rothery P, Stuart RC, Haughton AJ,Scott RJ, et al. Weed seed resources for birds in fields withcontrasting conventional and genetically modified herbicide-tolerant crops. Proceedings of the Royal Society of LondonSeries B – Biological Sciences 2006;273:1921–8.

[39] Evans AD, Green RE. An example of a two-tiered agri-environment scheme designed to deliver effectively theecological requirements of both localised and widespreadbird species in England. Journal of Ornithology 2007;148:S279–86.