Partial recovery of the population of Corncrakes Crex crex in Britain, 1993-2004: Capsule Singing male Corncrakes increased more than two-fold over the study period

This article was downloaded by: [111.161.126.83]On: 24 March 2014, At: 15:42Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Bird StudyPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/tbis20

Partial recovery of the population of Corncrakes Crexcrex in Britain, 1993–2004: Capsule Singing maleCorncrakes increased more than two-fold over thestudy period.Mark O'Brien , Rhys E. Green & Jeremy WilsonPublished online: 29 Mar 2010.

To cite this article: Mark O'Brien , Rhys E. Green & Jeremy Wilson (2006) Partial recovery of the population of CorncrakesCrex crex in Britain, 1993–2004: Capsule Singing male Corncrakes increased more than two-fold over the study period., BirdStudy, 53:3, 213-224, DOI: 10.1080/00063650609461436

To link to this article: http://dx.doi.org/10.1080/00063650609461436

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Corncrakes Crex crex bred regularly in every county inBritain in the late 19th century (Holloway 1996), butsuffered a marked decline, which continued throughmost of the 20th century (Norris 1945, 1947, Sharrock1976, Gibbons et al. 1993). By 1990 they were largely confined to northern and western Scotland, in particu-lar the Hebridean Islands (Gibbons et al. 1993). Thefirst national survey of the number of singing males wascarried out in 1978–79 (Cadbury 1980), and further

surveys using similar methods were completed in 1988(Hudson et al. 1990), 1993 (Green 1995) and 1998(Green & Gibbons 2000). The number of singing birdsrecorded declined from at least 700–746 in 1978–79 to574 in 1988 and 480 in 1993, but recovered to 589 in1998 (Green & Gibbons 2000).

Corncrakes are trans-Saharan migrant rails which, inthe breeding season, forage, nest and rear their youngin tall grass and herbs, particularly hay and silage meadows and, less often, in arable crops (Norris 1947,Cadbury 1980). In Scotland, the vegetation in hay andsilage meadows is not tall enough to conceal birds

Bird Study (2006) 53, 213–224

© 2006 British Trust for Ornithology

Partial recovery of the population of CorncrakesCrex crex in Britain, 1993–2004

MARK O’BRIEN1, RHYS E. GREEN2,3* and JEREMY WILSON1

1Royal Society for the Protection of Birds, Dunedin House, 25 Ravelston Terrace, Edinburgh, EH4 3TP, UK,2Conservation Science Group, Department of Zoology, University of Cambridge, Downing Street,Cambridge, CB2 3EJ, UK and 3Royal Society for the Protection of Birds, The Lodge, Sandy, Bedfordshire,SG19 2DL, UK

Capsule Singing male Corncrakes increased more than two-fold over the study period.Aim To provide updated counts of the number and distribution of singing male Corncrakes in Britain anddescribe the implementation of schemes in which land management was modified to benefit Corncrakes.Methods Annual surveys of singing Corncrakes in the core of the range in Britain were undertakenbetween 1993 and 2004 and the total population of the species throughout Britain was assessed in2003. Annual information on the extent and location of the various conservation management schemeswas obtained from the authorities that administer them.Results Following a long-term decline, the count of Corncrakes in the core of the species’ British rangeincreased in ten of the 11 years since 1993. A survey of the core of the range in 2004 located 1040singing males, a 2.3-fold increase on the 446 singing males located in the same area in 1993. A full sur-vey of Britain and the Isle of Man in 2003 located 832 singing males in 81 10-km National Grid squares.This compares with 589 birds in 93 squares in the last full survey in 1998, and 480 birds in 83 squaresin 1993. During 1993–2003, more than 70% of Corncrakes in the core area occurred in 1-km NationalGrid squares in which conservation schemes were implemented. For males in squares with some conservation management, the average area of managed land to which each bird potentially had exclusive access was about 5 hectares. About half of the tall grass present in 1-km squares occupied byCorncrakes was included in schemes intended to modify mowing or grazing to reduce losses of nests andchicks.Conclusion There has been a partial recovery of the population of Corncrakes in Britain since 1993,reversing a decline that began in the late 19th and early 20th centuries. However, the geographicalrange of the species, measured as occupied 10-km grid squares, has not increased. The recoveryoccurred during a period in which a significant area of land, to which the majority of Corncrakes hadaccess, was entered into conservation management schemes intended to benefit the species. Estimatedchanges in breeding productivity likely to be associated with changes in grass-mowing management onthe observed scale were approximately sufficient to account for the observed change in population trend.

*Correspondence author.Email: [email protected]

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014

when they first arrive in April and May and they arerestricted to patches of marshland vegetation and tallherbage until grass grows sufficiently tall that they canuse the meadows as well (Cadbury 1980). A high proportion of nests and broods, particularly those fromsecond and subsequent breeding attempts, are found inhabitats that are liable to be mowed for hay or silage(Green et al. 1997). The mowing season (June–August) overlaps substantially with the Corncrakebreeding season (Green et al. 1997). Nests aredestroyed by mowing and a high proportion of chicks present in mowed fields are killed, though thismortality can be reduced substantially by altering thepattern of mowing (Tyler et al. 1998). The long-termdecline of the Corncrake population in Britain hasbeen attributed to the effects on breeding success of theintroduction of mowing machines (Norris 1947) andintensification of hay and silage production. Other factors, especially the loss of habitats with suitable tall vegetation caused by changes in land-use, haveprobably also been important in the late 20th century(Green 1995).

A study of the distribution and abundance ofCorncrakes in Scotland and Ireland in 1993 found thathigh-density populations remained only in regionswhere substantial areas of suitable tall vegetation werepresent throughout the breeding season (late April toearly September) and where mowing of hay and silageoccurred mainly in late July and August (Green 1996).This correlation was attributed to the beneficial effectson breeding success and the survival of adults andchicks of having vegetation tall enough to concealCorncrakes throughout the period when they are onthe breeding grounds, combined with the less severeeffects on nest and chick survival of late compared withearly mowing. This suggests that effective conservationmeasures should (1) increase the area of suitable tallvegetation (tall grass and herbs), including habitatsthat will provide cover at times when vegetation ingrass fields is too short for Corncrakes (early and latecover), (2) reduce the overlap between the breedingand mowing seasons by delaying the date of mowing,and (3) reduce mortality of chicks caused by mowing byadopting Corncrake-friendly mowing techniques(CFM), that allow chicks to escape from mowed areaswhilst remaining in cover (Tyler et al. 1998). It wouldbe expected that these measures would only be effec-tive if implemented in combination in the same area.

A recovery programme for Corncrakes was initiatedin Scotland in 1992 (Stowe & Green 1997, Williams etal. 1997). At first, it focused on reducing losses of nests

214 M. O’Brien, R.E. Green and J. Wilson

© 2006 British Trust for Ornithology, Bird Study, 53, 213–224

and chicks to mowing by paying farmers and crofters todelay mowing until after the end of July and to useCFM techniques in fields close to the singing places ofmale Corncrakes (the Corncrake Initiative). Morerecently, the programme has also attempted to provideearly and late cover habitats. It is currently co-ordinated by the Steering Group of the Corncrake UKBiodiversity Action Plan.a Changes to land manage-ment intended to benefit Corncrakes have beenachieved by a variety of mechanisms, including theacquisition and management of nature reserves, payments to farmers and crofters for specified manage-ment under voluntary agreements with the RoyalSociety for the Protection of Birds (RSPB), ScottishNatural Heritage (SNH) or National Trust for Scotland(NTS). Participation of farmers and crofters in volun-tary agri-environment schemes administered by theScottish Executive Environment and Rural AffairsDepartment (SEERAD) has also grown substantially.

In this paper, we report the results of annual monitor-ing of Corncrakes in the core of their range in the period1993–2004 and a survey of Corncrakes throughoutBritain in 2003. We also document the implementationof conservation measures and the proportion ofCorncrakes that are likely to be affected by them.

METHODS

Systematic surveys in the main Corncrake range

Counts of singing male Corncrakes were made bymethods described previously (Green & Gibbons2000). In each of the years 1993–2004, systematicsearches were made at night on the Hebridean islandsand Orkney, which together held most (>90%) of theBritish population in the 1988, 1993 and 1998 surveys.These core areas are listed in Table 1. It is believed thatall suitable habitat was surveyed within these areas. In2003, Scottish islands other than those comprising thecore areas (Table 1), and all 32 10-km National Gridsquares within mainland Scotland north and west ofthe Great Glen with breeding season records ofCorncrakes since 1993, were also surveyed (see maps in Green 1995, Green & Gibbons 2000). Systematiccoverage in 2003 was similar to that achieved in 1998.In 1998, a random sample of seven of the 15 10-kmsquares in which Corncrake had been recorded since1988 were also searched but no birds were found. Thispart of the survey was not attempted in 2003. Recordsof birds in these areas were sought from other sources(see below).

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014

Additional records from other areas

To obtain records of Corncrakes from outside the areassurveyed systematically, requests for information weremade to the general public and birdwatchers through-out Great Britain. National and local press releases inMay encouraged people who believed that they hadheard or seen a Corncrake to telephone the RSPBnational headquarters or regional and country offices.Callers were asked for details of their record and birdidentification experience and were usually played arecording of a singing Corncrake. Promising recordswere passed to RSPB regional and country conserva-tion staff and, wherever possible, field visits were madeby experienced field observers who sought to confirmwhether or not a Corncrake was present. Records confirmed by experts or considered to be reliable on thebasis of the observers’ previous experience were

© 2006 British Trust for Ornithology, Bird Study, 53, 213–224

Corncrake recovery 215

included if the bird was present within the period 20May to 10 July. Information on Corncrakes in the Isleof Man was contributed by the Manx Bird Atlas.

County Bird Recorders outside the core areas werecontacted during the winter following the 2003 survey.They were given the records for their county that weconsidered to be acceptable and asked for additionalrecords that they had received.

Calculation of geographical range size

We used the numbers of 1-km and 10-km NationalGrid squares in which Corncrakes were present as ameasure of geographical range size. In cases where several singing places were attributed to the same male,we calculated the ‘centre of gravity’ (centroid) of thelocations and attributed that bird to the square inwhich the centroid lay.

Table 1. Numbers of singing male Corncrakes recorded in Britain and the Isle of Man from the full surveys of 1978–79, 1988, 1993, 1998and 2003 and annual counts in core areas in 1994–97, 1999–2002 and 2004. Core areas are those with counts in all years. A range isgiven where the total is uncertain.

1978–79 1988 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Shetland 3–5 1 3 – – – – 4 – – – – 4 2Orkney 102–105 33 6 20 39 43 21 13 15 11 9 16 31 17Lewis 31–38 71 106 89 96 104 114 77 50 55 60 57 105 118Harris 7 5 4 7 8 5 5 3 4 3 4 5 11 24Berneray 5 2 2 0 0 1 4 2 2 1 1 1 1 3North Uist 75 58 66 51 53 66 78 73 66 66 74 71 77 112Benbecula 33 23 17 19 23 27 45 41 36 31 34 30 30 35South Uist 83 96 50 49 58 81 62 65 101 98 90 98 101 119Barra & Vatersay 23–25 59 26 40 44 58 74 47 46 56 64 61 68 72Pabbay & Mingulay – – 0– – – – – 0 – – – – 2 5Skye 31–34 27 8 9 9 21 23 12 16 23 21 23 23 32Canna 9 2 3 2 0 0 0 2 0 0 0 2 4 4Rum 1–2 0 1 0 0 0 0 0 0 0 0 0 0 0Eigg 4–6 0 1 2 0 0 0 0 0 0 1 0 0 1Muck 4 0 1 1 0 0 0 0 0 3 2 4 5 3Coll 28 20 20 25 37 40 42 40 48 53 62 74 90 134Tiree 85 101 111 126 140 117 136 136 144 153 143 172 184 260Mull 1 3 1 1 0 0 0 2 2 0 0 1 0 5Iona 25 3 4 3 4 5 8 12 13 9 7 12 24 24Treshnish Isles 0 0 2 – – – – 3 – – – – 2 5Colonsay/Oronsay 22 18 10 6 12 8 8 14 21 21 18 25 32 46Islay 22–24 19 9 13 14 7 17 4 7 8 7 8 10 31Jura 1 1 0 – – – – 0 – – – – 0 0Gigha 1 4 0 – – – – 0 – – – – 1 0McCormaig Isles 0 0 1 – – – – 0 – – – – 0 0Scottish Mainland 94–103 24–29 18 – – – – 24 – – – – 17 12England 11–17 4 8 – – – – 13 – – – – 2 –Wales 1 0 1 – – – – 2 – – – – 0 –Isle of Man 0–3 1 1 – – – – 0 – – – – 6 –Total in core area 594–692 540 446 463 537 583 637 543 571 591 597 660 796 1040Total 662–773 574 480 589 832 1067

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014

Information on Corncrake conservation management

For parcels of land whose management was modifiedwith the intention of benefiting Corncrakes, we collected information from the organizations thatadministered the schemes. We documented the areaand location of each parcel, the actions which the landmanager had agreed to implement and the years inwhich the management was applied. Details of theschemes are given in Table 2. The EnvironmentallySensitive Area (ESA) scheme in the machair of theUists, Benbecula, Barra and Vatersay was introduced in 1988, but at first it did little to influence the mowing date of hay and silage because mowing wasallowed relatively early (after 15 July). Hence, we onlyincluded land entered into the scheme after it wasrevised in 1995 to require that mowing occurred after31 July.

Comparison of the distribution of managementschemes and singing male Corncrakes

We wished to consider the area of land under conser-vation management in areas where not just singingmales, but also nests, females and broods would beexpected to be present. The nest of a female Corncrakeand her brood-rearing area is usually within 250 m ofthe singing place of the male with which she paired(Tyler & Green 1996, Tyler 1996). Several locations ofsinging male Corncrakes are frequently attributed tothe same individual because it is known that malesoften move between singing places during the course ofa breeding season and may obtain more than one mate(Tyler & Green 1996). Rules for allocating singingrecords at different locations to individual males, forthe interpretation of survey results, are given by Green& Gibbons (2000). For each singing place attributed toa given male, we measured the proportion of a 250-m-radius circle centred on a singing place that lay withinall 1-km National Grid squares overlapped by the cir-cle. For each of the squares overlapped by any of the250-m-radius circles around the male’s singing places,we then averaged these proportions across the singingplaces attributed to that male. This procedure wasrepeated for all males and the results for a given 1-kmsquare were then summed to give a measure of thenumber of Corncrakes that might be influenced bymanagement within that square.

The proportion of Corncrakes potentially affected bya particular type of scheme was calculated by dividing

216 M. O’Brien, R.E. Green and J. Wilson

© 2006 British Trust for Ornithology, Bird Study, 53, 213–224

the number of Corncrakes within the core area thatwere in 1-km squares which included that type ofrecovery scheme, calculated as described above, by thetotal number of Corncrakes in the core area. The average area of land managed in a particular type ofscheme per male Corncrake was calculated as the areaof land in schemes within 1-km National Grid squareswith Corncrakes divided by the number of Corncrakesin those squares. Hence, this measure represents theamount of managed land per Corncrake for those birdswith potential access to some managed land.

Average area of tall grass in 1-km National Gridsquares occupied by Corncrakes

We wished to estimate the proportion of all tall grass(defined as grass >20 cm tall in July; Green 1996) insquares with Corncrakes that was subject to conserva-tion management of mowing or grazing intended toreduce losses of nests and chicks. We estimated the areasubject to management as described above, but did nothave annual estimates of the total area of tall grass present. We therefore used information from a habitatsurvey carried out in 1993 (Green 1996). The averagearea of tall grass was obtained from surveys of 50 1-kmsquares in the core area that held at least one centroidof the singing places of a male Corncrake in 1993. Forsurvey methods see Green (1996).

RESULTS

Total population of Corncrakes in Britain in 2003

A total of 832 singing birds was detected over all areasduring the survey period (Table 1), of which 12 wouldnot have been detected without information from thepublic. Eleven Corncrakes were heard in the 32 10-kmsquares searched in the northern mainland of Scotland.However, all of these birds had been reported toregional RSPB staff prior to being found by theCorncrake fieldworkers. A request for further informa-tion on Corncrake numbers in 2003 made to CountyBird Recorders in spring 2004 produced no additionalbirds.

The numbers of Corncrakes counted in 2003 in thegeographical subdivisions used in previous surveys aregiven in Table 1, together with results from the surveysof 1978–79, 1988, 1993, 1998 and annual countswithin the core areas. The distribution and abundanceof Corncrakes by 10-km squares in northern Britain in2003 is shown in Fig. 1.

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014

© 2006 British Trust for Ornithology, Bird Study, 53, 213–224

Corncrake recovery 217

Scheme OrganizationYear introduced Areas covered

Eligible if birds not present Early/late cover Mowing/grazing delay

Corncrake-friendly mowing

Corncrake Initiative

RSPB with SNH and SCF

1992 All No No Livestock and mowing excluded from 15 May until after 31 July on compartments within 250 m of singing male

On compartments within 250 m of singing male

RSPB Nature Reserves and management agreements, including Nadair-funded projects

RSPB Varies between sites, from 1991

Islay, Coll, Tiree, Oronsay, Colonsay, South Uist, North Uist, Lewis, Egilsay, Copinsay

Yes Fenced-off field corners and edges. Suitable plant species established artificially where necessary

Livestock and mowing excluded from early March until after 10 August on compartments within 250 m of singing male

All fields within reserve or agreement areas

Skye Grassland Scheme

RSPB and SNH

1993 Skye Yes No grazing between 1 March and 15 October

Livestock and mowing excluded between 15 April (1 May on reseeded grasslands) and 1 August

All fields in scheme

NTS Early Cover Scheme (included in RSPB nature reserves in analysis)

NTS 1997 Iona, Canna Yes Fenced off areas. Livestock excluded between March and mid-October. Adjacent to hay meadows

N/A N/A

SNH Corncrake SPA Management Schemes and Section 15 sites

SNH SPAs from 1999, Section 15 sites from 1993

Islay, Coll, Tiree, Barra, South Uist, Benbecula, North Uist, Lewis

Yes Fenced-off field corners and edges. Livestock excluded between end February and mid-October

Livestock and mowing excluded from 15 May until after 31 July

All fields in scheme

ESA Uist Machairs SEERAD 1995 Vatersay, Barra, South Uist, Benbecula, North Uist

Yes No Mowing excluded until after 31 July

All fields in scheme

ESA Argyll Islands SEERAD 1995 Islay, Jura, Colonsay, Oronsay, Coll, Tiree, Iona, Mull

Yes Provision and safeguarding of tall vegetation adjacent to hay crop encouraged, (funding available from 2001)

Livestock excluded by 1 June. Grass may not be cut before 1 August

Countryside Premium Scheme

SEERAD 1998 All (Corncrake-specific in Argyll Islands, Orkney, Skye or Western Isles)

Yes Must be adjacent to hay or silage meadow

Livestock and mowing excluded between 15 May and 31 July unless no Corncrakes present on or near the holding

All fields in scheme

Rural Stewardship Scheme

SEERAD 2001 All Yes Must be adjacent to hay or silage meadow. 0.15–1 ha in size. No grazing between 28 February and 30 September

Livestock and mowing excluded between 15 May and 31 July

All fields in scheme

Table 2. Details of main Corncrake conservation schemes in Britain (all schemes listed were in operation up to at least 2003).

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014

Population of Corncrakes in the core area

The annual survey of the core area for Corncrakes wasundertaken in 2004 by RSPB staff. This showed a substantial recent increase in the numbers recorded: thepopulation increased from 796 singing males in 2003 to1040 in 2004 – a 31% increase between the two years.This follows a 21% increase between 2002 and 2003 inthe core area (from 660 to 796 singing males). Table 1includes a summary of all information on Corncrakenumbers from Scotland in 2004, including minimumcounts outside the core area, where available. It is clearthat the numbers have increased substantially in boththe Inner and Outer Hebrides, but have declined inOrkney. There may also have been a decline on themainland of Scotland, but this is not within the corearea and it is possible that survey coverage was lowerthan in the previous full survey year.

Trends in Corncrake numbers and range

Average annual rates of population changeb were

218 M. O’Brien, R.E. Green and J. Wilson

© 2006 British Trust for Ornithology, Bird Study, 53, 213–224

calculated for the intervals between each of the full surveys of Corncrakes in Britain and the Isle of Man.Corrections to allow for incomplete surveys of parts ofthe Outer Hebrides in 1978–79 were first applied, asdescribed previously.b Between 1978–79 and 1988, theCorncrake population declined at an average annualrate of 3.4% per year and the decline continued at asimilar rate (3.5% per year) between 1988 and 1993.The estimate for the whole period 1978–79 to 1993 isa decline of 3.4% per year. From 1993 to 1998 the population increased by an average of 4.2% per year,which accelerated to 7.2% per year between 1998 and2003. If only the data from full surveys (1993, 1998,2003) are used, the average rate of increase between1993 and 2003 is 5.7% per year. Using the data for thecore area allows the estimate to be made over theslightly longer period 1993–2004, during which thepopulation in the core area increased at an average rateof 5.6% per year. The pattern was not even acrossregions. The fringes of the British distribution, in theNorthern Isles, on the mainland of Scotland and inEngland and Wales, showed fluctuating numbers(Table 1).

Annual monitoring within the core of the Corncrakerange indicates that Corncrake numbers increased inten of the 11 years from 1994 to 2004 (Fig. 2). Annualchanges ranged between a 15% decrease in 1997–98and a 31% increase in 2003–04. The count in the corearea increased by a factor of 2.33 between 1993 and2004.

The geographical range, measured in terms of 10-km

Figure 1. Map of Scotland showing 10-km National Grid squareswith singing male Corncrakes during the breeding season of 2003.The size of the circles indicates the number of singing males (seekey). The number in brackets indicates the number of 10-km squaresin each category. Two 10-km squares further south in England thateach held one Corncrake, two 10-km squares on the Isle of Manthat held a total of six Corncrakes, and three 10-km squares onShetland that held a total of four Corncrakes are not shown.

Figure 2. Number of singing male Corncrakes in Britain1978–2004. The bars indicate the numbers of Corncrakesrecorded in each of the full surveys, while the crosses show resultsfor the core areas that have been surveyed in each year since1993. A correction has been made for incomplete surveys of Lewisand Barra in 1978.b

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014

National Grid squares, decreased from 93 to 81 squaresbetween 1998 and 2003 (Table 3, Fig. 3). This is thelowest number of 10-km squares recorded as beingoccupied in any full survey year. However, the numbersof occupied 1-km squares and 10-km squares in the coreareas have increased (Fig. 3).

The area of land affected by Corncrake management schemes

The area of land within the core of the Corncrakerange that is affected by management schemesintended to benefit Corncrakes increased almost five-fold between 1992 and 2003 (Fig. 4). In 1992 virtuallyall of the area managed was provided through theCorncrake Initiative and on nature reserves and management agreements administered by the RSPB.However, the area affected by agri-environmentschemes, initially the Environmentally Sensitive Areasin the Argyll Islands and Uists Machair, and later alsothe Countryside Premium Scheme and RuralStewardship Scheme, increased markedly, both inabsolute and relative terms. The area affected by RSPBreserves and management agreements and by the SNHCorncrake Special Protection Area (SPA) Scheme andSection 15 agreements also increased substantially, butthe area covered by the Corncrake Initiative decreased(Fig. 4), because many farmers and crofters who wereeligible for Corncrake Initiative payments switched toagri-environment schemes with similar prescriptionsinstead. Simultaneous payments under both schemeswere not permitted.

Although there is a large and confusing number ofschemes intended to benefit Corncrakes, they fall intotwo groups: (a) management of tall grass by delayed

© 2006 British Trust for Ornithology, Bird Study, 53, 213–224

Corncrake recovery 219

mowing or grazing and/or CFM; and (b) creation ormanagement of early/late cover. Within these twogroups, the different schemes have broadly similar prescriptions and intended outcomes. After groupingthe schemes in this way, we found that modification ofgrass-mowing or grazing was the most common form ofhabitat management for Corncrakes, with managementof early/late cover affecting a much smaller area (Fig.5). The area in both types of management increasedsubstantially, but the area of land with modification ofmowing or grazing in 1-km grid squares containingCorncrakes has increased more slowly than the totalarea in this type of management since 1997 (Fig. 5).This is because of increased switching from theCorncrake Initiative, which requires that a singingmale is present, to agri-environment schemes (ESA,CPS, RSS) that do not require Corncrakes to be present for the land to be eligible for the scheme (Fig.4). As a result, 46% of payments for modified mowingin the core area in 2003 were in 1-km squares without

Table 3. Number of 10-km squares occupied by Corncrakes inBritain and the Isle of Man in 1968–72, 1978–79, 1988, 1993,1998 and 2003.

Region 1968–72a 1978–79 1988 1993 1998 2003

Shetland 9 2 1 3 4 3Orkney 30 27 12 4 10 15Outer Hebrides 31 31 29 29 30 29Inner Hebrides 54 36 31 20 21 26Rest of Scotland 216 54 14 18 15 5England 133 9 3 7 11 2Wales 42 1 0 1 2 0Isle of Man 12 0 1 1 0 1Total 527 160 91 83 93 81

a For 1968–72 the combined total of squares with confirmed andprobable breeding from Gibbons et al. (1993) is used.

Figure 3. Geographical range of the Corncrake 1993–2004.Numbers of occupied 10-km (a) and 1-km (b) National Gridsquares within the core area (�) of the range. ��, The total numberof occupied squares in Britain and the Isle of Man from full surveysin 1993, 1998 and 2003. Occupied squares are those that con-tain at least one centroid calculated from all locations attributed toan individual calling bird.

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014

Corncrakes. Most land in early/late cover schemes hascontinued to be in squares with Corncrakes present(Fig. 5).

Proportion of Corncrakes in the core area affectedby management schemes

The proportion of male Corncrakes in the core areasthat occurred in 1-km squares that include land inschemes that require the modification of mowing or graz-ing and the provision of early/late cover have bothincreased (Fig. 6). The average proportion of birds insquares with modified mowing or grazing during theperiod 1993–2003 was 72% (range of annual values60–80%) and has shown a small but significant increaseover time (r = 0.895, P < 0.001). The average proportionof birds in squares with land in schemes supportingearly/late cover during the period 1993–2003 was muchlower (24%), but showed a much larger increase (3% in1993 to 36% in 2003; r = 0.830, P < 0.001).

Area of managed land available per Corncrake insquares in which Corncrakes and managed landwere both present

Although a high proportion of Corncrakes occurred in

220 M. O’Brien, R.E. Green and J. Wilson

© 2006 British Trust for Ornithology, Bird Study, 53, 213–224

1-km National Grid squares where land managed tobenefit them was present, it might be that the area ofmanaged land present in these squares was too small tohave a significant effect on their survival or breedingsuccess. We therefore calculated a measure of the quantity of habitat subject to conservation schemesavailable to the average male Corncrake in those 1-kmsquares of the core area that contained bothCorncrakes and land in schemes (see Methods). Duringthe period 1993–2003, the average male Corncrake inthese squares potentially had exclusive access to 4.6 haof land on which efforts were made to modify the timing and/or method of mowing. There were annualfluctuations in this measure and a slight tendency for itto increase (Fig. 7; r = 0.437, P = 0.12). The area ofearly/late cover in schemes per male Corncrake wasconsiderably lower, averaging 0.5 ha, but showed aclearer tendency to increase (r = 0.718, P < 0.02) fromabout 0.4 ha per bird in 1993 to 0.8 ha per bird in 2003(Fig. 7).

Proportion of tall grass under management insquares in which Corncrakes were present

Based on habitat surveys in 1993 in 50 1-km NationalGrid squares in the core area with at least oneCorncrake present, there was an average of 15.0 ha oftall grass present per square. During the period1993–2003, the average area of land in schemes thatrequire the modification of mowing or grazing in 1-km

Figure 4. The area of land in each of the management schemeswithin the core area of the Corncrake range between 1992 and2003. Further details about the schemes can be found in Table 2.Parts of the diagram with diagonal hatching represent schemes runby statutory and non-governmental conservation bodies, whilstthose with vertical or horizontal hatching represent agri-environmentschemes run by the statutory agriculture department. ,Environmentally Sensitive Areas; , Rural Stewardship Scheme;

, Countryside Premium Scheme; , Scottish Natural Heritage;, Skye Grassland Scheme; , RSPB reserves and agreements

(including The National Trust for Scotland Early Cover Scheme); , Corncrake Initiative.

Figure 5. Area of land within the core area of the Corncrakerange managed within schemes as early/late cover (open symbols)or by modifying the timing and/or method of mowing or grazing ofgrass (filled symbols). The total area (squares) and the area within1-km National Grid squares with Corncrakes (diamonds) are shownseparately.

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014

squares in the core area that were occupied byCorncrakes was 7.1 ha per square. Hence, if we assumethat the 1993 estimate for the area of tall grass persquare applies to the whole period, an average of 47%(7.1/15.0) of tall grass in core-area 1-km squares occu-pied by Corncrakes was subject to modification ofmowing or grazing.

DISCUSSION

Change in population trend

The trend in the Corncrake population in Britain haschanged markedly since about 1993. Between 1978 and1993 the population declined at an average rate ofabout 3.4% per year, with the rate in the two compo-nent periods (1978–79 to 1988, and 1988 to 1993)being virtually the same (Green & Gibbons 2000).Strong evidence that this was the continuation of along-term decline that began in the late 19th and early20th centuries comes from studies of distribution andhistorical reports of local population collapses thatwere carried out before the first census (Norris 1945,1947, Sharrock 1976). In the period 1993 to 2003 thetotal population increased by an average of 5.7% peryear, whilst the population of the core area increased byan average of 5.6% per year between 1993 and 2004.

Changes in demographic rates required to accountfor the change in population trend

The change in population trend that occurred in about

© 2006 British Trust for Ornithology, Bird Study, 53, 213–224

Corncrake recovery 221

1993 must have been caused by a change in adult survival, recruitment (the product of the number ofyoung reared to independence per adult and first-yearsurvival), net immigration, or some combination ofthese three demographic rates. Immigration and emigration can probably be ignored because the popu-lation of Corncrakes in Scotland is remote from otherpopulations of comparable size (Hagemeijer & Blair1997). Furthermore, individual Corncrakes fromScotland are significantly larger in mean dimensionsand weight than those from the nearest neighbouringpopulations (R.E. Green unpubl. obs.), a differencethat would probably not be maintained with anythingother than low levels of exchange with other popula-tions.

If we refer to the two periods of population moni-toring before and after 1993 as periods 1 and 2, anddesignate the population multiplication rate, adult survival and recruitment in the two periods as λ1, λ2,S1, S2, R1, R2, respectively, then if birds first breed whenone year old (Green 2004), it can be seen that λ1 = S1+ R1 and λ2 = S2 + R2. If the change in population trendwas caused solely by a change in recruitment then theratio of recruitment in the second and first periodswould be R2/R1 = (λ2 − S.)/( λ1 − S.), where S. is adultsurvival, S. = S1 = S2. However, if the change in popu-lation trend was caused solely by a change in adultsurvival then the ratio of adult survival in the secondand first periods would be S2/S1 = S2/( λ1 − λ2 + S2). Athird scenario to consider might be that both recruit-ment and adult survival changed by the same

Figure 6. Proportion of male Corncrakes within the core area ofthe Corncrake range that were in 1-km National Grid squares withland managed within a scheme as early/late cover (��) or by mod-ifying the timing and/or method of mowing or grazing of grass (�).

Figure 7. Area of land managed within schemes as early/latecover (��) or by modifying the timing and/or method of mowing orgrazing of grass (�) per male Corncrake present. For each type ofscheme, only Corncrakes in 1-km National Grid squares with someland managed within the scheme are included in the calculation.

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014

proportion between periods 1 and 2, then R2/R1 = S2/S1= λ2/λ1.

We have estimates of λ1 and λ2 from the populationcounts described above (λ1 = 0.966, λ2 = 1.056). Anestimate of recent adult survival, S2 = 0.286, based onbirds captured almost entirely during the period ofincrease since 1993 is also available from Green(2004). Substituting these values into the expressionsgiven above, it can be seen that the observed change inpopulation trend could be accounted for by an increasein recruitment only of 13%, by a 46% increase in adultsurvival only, or by an increase in both recruitment andadult survival of 9%.

Estimated effects of conservation management ondemographic rates

The reversal of the population decline of Corncrakeshas coincided with the introduction of the recoveryplan. However, is there evidence of a causal link? Hasconservation management been implemented on ascale sufficient to have altered demographic ratesenough to be the cause of the change? Over the period1993 to 2003 an average of 72% of male Corncrakes inthe core area occurred in 1-km National Grid squareswith land in conservation schemes intended to modifymowing or grazing to reduce losses of nests and chicks.In most cases, these measures consisted of delayedmowing, CFM and usually both (Table 2, Fig. 4). Onaverage, 47% of tall grass in squares occupied byCorncrakes was subject to this type of scheme. Schemesthat delay mowing or grazing and provide for CFM areexpected to increase Corncrake productivity (the number of independent young reared per adult). It hasbeen estimated, using a simulation model of the timingof breeding and mowing, that delayed mowing andCFM together increase productivity by 23% whereapplied (Green et al. 1997, Green 1999). Combiningthese estimates gives an assessment of the averageeffect of mowing and grazing schemes on the produc-tivity of the whole core area population; 0.47 × 0.23 =0.11, an 11% increase in overall productivity. This estimate is likely to be too low. Tall grass subject toconservation management of mowing is more likely tobe close to Corncrake nests and brood-rearing areasthan other tall grass in the same square. This is becausethe Corncrake Initiative is only applied close to singingplaces of males and because several schemes combineprovision of early/late cover, which attracts breedingCorncrakes (Green 1996), with management of mow-ing and grazing. Hence, the estimate of a potential

222 M. O’Brien, R.E. Green and J. Wilson

© 2006 British Trust for Ornithology, Bird Study, 53, 213–224

improvement in productivity of at least 11% frommowing and grazing management schemes is broadlyconsistent with the improvement in recruitment by13% required to produce the observed change in popu-lation trend if this was caused by a change inrecruitment alone. Delayed mowing and CFM mightalso increase adult survival by preventing adults frombeing killed by mowers, but we do not have an estimateof the magnitude of this possible effect from field observations. However, watches of meadows duringmowing indicate that few adults are killed (Tyler et al.1998), so this effect is likely to be small.

Provision of additional early/late cover might also beexpected to change demographic rates. However, themechanisms by which this might occur are much moredifficult to quantify than the assessment of the effectsof modified mowing and grazing on productivity andrecruitment. By providing concealment when othersuitable vegetation is scarce at the beginning and end of the breeding season, extra cover might reducepredation on adults and juveniles, and competitionamong adults for singing and nesting sites. Losses ofnests, chicks and adults to mowing are also likely to bereduced by providing more early/late cover. This isbecause Corncrakes prefer typical early/late cover vegetation, such as umbellifers (mostly Anthriscussylvestris and Heracleum sphondylium), Yellow Flag IrisIris pseudacorus and nettles Urtica dioica, to tall grass inmeadows during the mowing season. Radiotrackingdata for adult Corncrakes, including females withchicks, showed that the three early/late cover vegeta-tion types mentioned above had numbers ofradio-locations per unit area 4.7, 3.8 and 2.4 timeshigher, respectively, than for grass meadows in July(table 2.11 of Tyler 1996). The density of nests in theseearly/late cover types in July and August was alsohigher than for grass meadows, though the number ofnests built in meadows is higher because the area ofmeadow is usually greater (table 3.8 of Tyler 1996).Hence, providing more early/late cover vegetation ofthe preferred types should result in a smaller portion ofadults, chicks and nests being in tall grass, where theyare vulnerable to mowing. However, an average of only24% of male Corncrakes in the core area occurred in 1-km National Grid squares with land in conservationschemes intended to provide or protect early/late cover.Hence, it seems unlikely that this has yet had mucheffect on demographic rates or population trends.

Overall, our analysis indicates that conservationmanagement has occurred on a scale sufficient toaccount for the change in trend of the Corncrake

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014

population. Information of this kind forms an impor-tant part of the case for establishing a causal linkbetween a change in an external factor and a change ina bird population trend (Green 2002). However, otherkinds of evidence, such as replicated randomized fieldexperiments and correlation of spatial variation of pop-ulation trends with spatial variation of the externalfactor, are also desirable. In this case, the assessment bythese approaches of effects on population trend of management that influences breeding success is prob-lematic, because natal dispersal of young Corncrakes islikely to result in their recruitment as adults into areaswith different management (Green 1999, 2002).Spatially explicit population modelling that includesthe geographical distribution of conservation manage-ment and natal dispersal is a potential solution to thisproblem.

Recommendations for improving conservationmanagement for the Corncrake

Currently about 20% of male Corncrakes occur in 1-km National Grid squares where no scheme that seeksto modify mowing methods is in operation. The grass insome of these areas may not be mowed or may bemowed sufficiently late without the application of theschemes that there is no lost opportunity to benefitCorncrakes. However, it would be worth attempting toidentify areas where Corncrakes occur but where mowing is early and CFM not used, so that schemes tomodify mowing can be applied there, if appropriate.

There appears to be a much larger shortfall in theapplication of schemes designed to create or maintainearly/late cover vegetation. Currently about 64% ofmale Corncrakes occur in 1-km National Grid squareswhere no scheme of this kind is in operation. Theshortfall may not be as important as it appears becauseearly/late cover habitats used by Corncrakes sometimesoccur in areas that are not in agricultural use (e.g. atthe edges of wetlands) and that may be protected inother ways. These habitats may remain suitable with-out further conservation effort. Even so, there is a needto identify areas where the availability of suitableearly/late cover vegetation is limiting Corncrake density, and to devise ways to provide it. Of particularlyhigh priority for this are areas where farmers andcrofters receive payments for modifying mowing tobenefit Corncrakes, but where there are no Corncrakespresent.

Finally, it must be emphasized that the suitability forCorncrakes of many of their current habitats in

© 2006 British Trust for Ornithology, Bird Study, 53, 213–224

Corncrake recovery 223

Scotland depends upon the continuation of certaintypes of farming, especially those that include mowingof grass for winter food for livestock and low intensitygrazing. Fine-tuning farming methods to provide con-ditions for Corncrakes is desirable, but will not bepossible if the farming systems themselves cease toexist. Income to farmers and crofters from conservationschemes funded by government departments and othershas increased recently, but is unlikely on its own toensure that these systems persist.

ACKNOWLEDGEMENTS

We are grateful to the following people who took part in thesurveys on Corncrakes in 2003 and 2004: P. Ellis, E. Meek, A.Knight, D. Fox, M. Wilkinson, K. Fairclough, A. Mitchell, K.Graham, D. Mitchell, A. Rothwell, A. Sillence, A.MacLennan, M. Scott, C. Murray, A. Murray, J. Grant, Y.Benting, L. Coutts, B. Lowe, J. Boyle, G. Evans, A. Patterson,J. Bowler, S. Money, S. Wellock, D. Sexton, M. Peacock, C.McKay, A. Keys, J. McGhie, J. How, K. Chisholm, R. Broad,J. Tilley, G. Wilkinson, S. Foster, M. Edwards, R. Setchfield,K. Marshall, N. Aspey, I. Court, M. Cade and N. Melville. C.Sharpe kindly provided Manx Bird Atlas data. Many thanksto Judith Smith for advising on liaison with County BirdRecorders. The following SEERAD staff facilitated the avail-ability of data on agri-environment schemes: J. Hood, D.Bailey, D. Cowie and M. Christal. I. Hislop and R. Lilley pro-vided data relating to distribution of RSPB and SNHmanagement schemes, respectively. I. Peters, K. Peck, T.Partridge and A. Middleton forwarded information reportedby the general public to RSPB wildlife enquiries. RussellSlack at Birdguides very kindly incorporated a request forsightings on the national website. Richard Gregory andHelen Riley commented on previous drafts. EHS, NorthernIreland and Scottish Natural Heritage contributed to thecosts of the survey.

ENDNOTES

a. The UK Biodiversity Action Plan steering group forCorncrakes is chaired by the Royal Society for the Protectionof Birds (RSPB) and includes Scottish ExecutiveEnvironment and Rural Affairs Department (SEERAD),Scottish Natural Heritage (SNH), Scottish CroftingFoundation (SCF), National Farmers Union for Scotland(NFUS), Scottish Agricultural College (SAC), NationalTrust for Scotland (NTS), Scottish Landowners Federation(SLF) and the Farming and Wildlife Advisory Group(FWAG). The plan is located at the website:http://www.ukbap.org.uk/UKPlans.aspx?ID = 244

b. To calculate the population multiplication rate λ from justtwo counts, n1 and n2, in years t1 and t2, with no intervening

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014

counts, we took λ = (n1 /n2)1/(t2– t1). To calculate λ for a set ofthree or more counts, we first fitted a least-squares linearregression of loge(n) on year and then back-transformed theestimated slope parameter b, i.e. λ = exp(b). For both types ofestimates we took 100(λ – 1) to be the annual percentagerate of change of the population. Where upper and lower lim-its were given for a count (see Table 1), their average wasused. Counts in 1978–79 on Barra and Lewis were incom-plete, so estimates for those islands were imputed bymultiplying the count in 1988 by the ratio of the 1978–79 tothe 1988 counts for the other islands of the Outer Hebrides,excluding Lewis and Barra (Green & Gibbons 2000).

REFERENCES

Cadbury, C.J. 1980. The status and habitats of the Corncrake inBritain 1978/79. Bird Study 27: 203–218.

Gibbons, D.W., Reid, J.B. & Chapman, R.A. 1993. The NewAtlas of Breeding Birds of Britain and Ireland: 1988–1991. T. &A.D. Poyser, London.

Green, R.E. 1995. The decline of the Corncrake Crex crex in Britaincontinues. Bird Study 42: 66–75.

Green, R.E. 1996. Factors affecting the population density of the Corncrake Crex crex in Britain and Ireland. J. Appl. Ecol. 33:237–248.

Green, R.E. 1999. Survival and dispersal of male Corncrakes Crexcrex in a threatened population. Bird Study Suppl. 46: 218–229.

Green, R.E. 2002. Diagnosing causes of population declines andselecting remedial actions. In Norris, K. & Pain, D.J. (eds) Conserv-ing Bird Biodiversity: 139–156. Cambridge University Press,Cambridge.

Green, R.E. 2004. A new method for estimating the adult survival rate

of the Corncrake Crex crex and comparison with estimates fromring–recovery and ring–recapture data. Ibis 146: 501–508.

Green, R.E. & Gibbons, D.W. 2000. The status of the CorncrakeCrex crex in Britain in 1998. Bird Study 47: 129–137.

Green, R.E., Tyler, G.A., Stowe, T.J. & Newton, A.V. 1997. Asimulation model of the effect of mowing of agricultural grassland onthe breeding success of the Corncrake (Crex crex). J. Zool. Lond.243: 81–115.

Hagemeijer, W.J.M. & Blair, M.J. 1997. The EBCC Atlas of Euro-pean Breeding Birds: Their Distribution and Abundance. T. & A.D.Poyser, London.

Holloway, S. 1996. The Historical Atlas of Breeding Birds in Britainand Ireland 1875–1900. T. & A.D. Poyser, London.

Hudson, A.V., Stowe, T.J. & Aspinall, S.J. 1990. Status and dis-tribution of Corncrakes in Britain. Br. Birds 83: 13–186.

Norris, C.A. 1945. Summary of a report on the distribution and statusof the Corncrake (Crex crex). Br. Birds 38: 142–148, 162–168.

Norris, C.A. 1947. Report on the distribution and status of the Corn-crake. Br. Birds, 40: 226–244.

Sharrock, J.T.R. 1976. The Atlas of Breeding Birds in Britain and Ire-land. British Trust for Ornithology, Tring.

Stowe, T.J. & Green, R.E. 1997. Response of Corncrake Crex crexpopulations in Britain to conservation action. Vogelwelt 118:161–168.

Tyler, G.A. 1996. The ecology of the Corncrake, with special refer-ence to the effect of mowing on breeding production. PhD Thesis,University College Cork.

Tyler, G.A. & Green, R.E. 1996. The incidence of nocturnal song bymale Corncrakes Crex crex is reduced during pairing. Bird Study43: 214–219.

Tyler, G.A., Green, R.E. & Casey, C. 1998. Survival and behav-iour of Corncrake Crex crex chicks during the mowing of agriculturalgrassland. Bird Study 45: 35–50.

Williams, G., Green, R.E., Casey, C., Deceuninck, B. &Stowe, T.J. 1997. Halting declines in globally threatened species:the case of the Corncrake. RSPB Conserv. Rev. 11: 22–31.

(MS received 15 March 2005 revised MS accepted 11 August 2005)

© 2006 British Trust for Ornithology, Bird Study, 53, 213–224

224 M. O’Brien, R.E. Green and J. Wilson

Dow

nloa

ded

by [

111.

161.

126.

83]

at 1

5:42

24

Mar

ch 2

014