Factors affecting hedgehog presence on farmland as assessed by a questionnaire survey

ORIGINAL PAPER

Factors affecting hedgehog presence on farmland as assessedby a questionnaire survey

Anouschka R. Hof & Paul W. Bright

Received: 7 November 2010 /Accepted: 9 May 2011 /Published online: 20 May 2011# Mammal Research Institute, Polish Academy of Sciences, Białowieża, Poland 2011

Abstract The West European hedgehog, (Erinaceus euro-paeus, Linnaeus 1758) is widely distributed in WesternEurope. However, there is evidence of decline in parts of itsrange. Changes in agricultural management have partly beenthe driving force behind the loss of species diversity andabundance, and it has been argued that these changes play arole in the decline of hedgehogs as well. We used aquestionnaire to investigate the current distribution ofhedgehogs on farmland throughout Great Britain with afocus on different environmental zones. Additionally,we identified environmental correlates that related to thedistribution of hedgehogs with the aim to get a betterunderstanding of what is needed to design appropriatestrategies targeted at the conservation of hedgehogs. Ourstudy illustrates that, although the impact of several variableswas rather ambiguous, displaying positive effects in someenvironmental zones and negative effects in other, major roadsand Eurasian badgers (Meles meles, Linnaeus 1758) can havelarge scale negative effects on hedgehogs. Farm managementrelated factors did not show a consistent impact on hedgehogpresence. Conservation strategies should therefore be aimed atlessening the impacts of major roads and badger presence.Wildlife passages, for instance, may provide hedgehogs safe

passages across roads. Additionally, increasing the habitatcomplexity in order to reduce the impact of predators can bebeneficial for prey species, such as hedgehogs, and should beconsidered as a conservation strategy for them.

Keywords Erinaceus europaeus . Agri-environmentschemes . Distribution . Hedgerows . Predation

Introduction

Numerous species of flora and fauna are currently threat-ened. Recently, extinctions are occurring at an unprece-dented rate due to the increased levels of humandisturbance and other factors directly and indirectly causedby mankind (McNeely et al. 1995; Pimm et al. 1995; Salaet al. 2000). So far, the main focus of conservation biologyhas been on rare species at risk of (local) extinction.However, recently the significance of studying morecommon species has been recognised, since proportionally,relatively small declines in the numbers of common speciescan mean large absolute losses, which might disruptecosystems (Gaston and Fuller 2008).

Without the availability of long-term intensive surveying, itis difficult to detect changes in wildlife populations.Nevertheless, it has been found that the West Europeanhedgehog (Erinaceus europaeus), a common widespreadspecies, is currently in significant decline in part of itsrange (Davey and Aebischer 2006; Hof 2009). Changes infarm management are suggested to partly cause theirdecline (Davey and Aebischer 2006; Hof 2009). In largeparts of Europe, changes in agricultural management havenegatively affected biodiversity on farmland (Krebs et al.1999; Robinson and Sutherland 2002). The hedgehog is amobile species and a generalist predator of macro-

Communicated by Jan M. Wójcik

A. R. Hof : P. W. BrightSchool of Biological Sciences, Royal Holloway,University of London,Egham, Surrey TW20 0EX, UK

Present Address:A. R. Hof (*)Department of Ecology and Environmental Science,Umeå University,SE-90187 Umeå, Swedene-mail: [email protected]

Acta Theriol (2012) 57:79–88DOI 10.1007/s13364-011-0044-y

invertebrates, which are an important source of prey formany other taxa as well. The hedgehog’s decline mightimply general deterioration of the agricultural landscape inBritain. It is therefore important to understand the currentvalue of rural areas for hedgehogs and to identifymitigation strategies where necessary.

Questionnaire surveys are a useful tool to study thedistribution of elusive species and have become increas-ingly popular in ecology (White et al. 2005). Usingquestionnaires enables a specific group of people to betargeted and gathers extensive datasets in a relatively shortperiod. Nevertheless, data obtained from the questionnaireshave to be regarded with some care since the targeted groupis often not a specialist in the field. Hedgehogs, however,are easily recognized, and it is thus felt that reliable datacould come from a questionnaire survey. A questionnairestudy has therefore been used to identify factors that mayinfluence the presence of hedgehogs on farms. The aim ofthis research was to study the current distribution ofhedgehogs in the countryside of Great Britain and to gaininsight into which aspects are related to hedgehog presenceon farms. A model was developed for Great Britain, and forthe environmental zones separately, in order to investigatethe factors that had an impact on the presence of hedgehogsat a somewhat smaller scale, since specific local landscapefeatures might be important in determining the presence ofhedgehogs on a more local scale but be irrelevant on thecountrywide scale.

Study area

The study area comprised of farms throughout GreatBritain. The distribution of hedgehogs on farms wasdetermined from the data collected by means of a postal-based questionnaire survey aimed at a random selection oflandowners. Landowners were chosen at random bygeographic county in order to get a representative sample.Addresses were obtained from the ‘yellow pages’ (availableat URL: http://www.yell.com/).

Material and methods

The questionnaire

The aim was to cover at least 1% of the land underagricultural management by the questionnaires, distributedthroughout Great Britain, and representing all governmentoffice regions of England, Scotland, Wales and allenvironmental zones. The environmental zones (arable-dominated lowlands of England, pasture-dominated low-lands of England, uplands of England and Wales, lowlandsin Scotland, marginal uplands and islands of Scotland, and

the true uplands of Scotland) were defined by the BritishCountryside Survey 2000 (available at URL: http://www.countrysidesurvey.org.uk/archiveCS2000/) and wereformed from aggregations of the 40 base classes of theITE Land Classification of Great Britain (Bunce et al. 1996)(Fig. 1). Based on the total area under agricultural holdings,the total number of holdings (URL: https://statistics.defra.gov.uk/esg/publications/fiuk/2002/c5_t3-4.xls) and theaverage response rate from questionnaires used in ecology(63%) [White et al. 2005] just over 2,500 landownersneeded to be sent a questionnaire. In order to obtain a gooddistribution and to account for a lower response rate, thequestionnaire was sent to 4,000 landowners in 2006. Postalcodes of the respondents were retained in order to locate the

Fig. 1 The environmental zones of Great Britain as defined by theCountryside Information System (available at http://www.countrysidesurvey.org.uk/archiveCS2000/)

80 Acta Theriol (2012) 57:79–88

farms on the 1-km2 level. The questionnaire was as simpleand short as possible in order to maximize the return. Asmall pilot study aimed at ten landowners was set up inorder to identify possible problems in the questionnaire.Recommendations as set out by White et al. (2005) werefollowed where possible. Unfortunately, logistics did notallow a ground-truthing study to assess the accuracy of thedata obtained from the questionnaire.

Data collection

The main data obtained from the questionnaire concernedthe location of the farm provided by a postal code and thepresence of hedgehogs. The following data regardingagricultural management were also obtained from therespondents:

& Size of the farm (in hectares)& Percentage of grassland, woodland and arable land (0%,

1–25%, 26–50%, 51–75%, 76–100%)& Length of hedgerows (in meters)& Type of farm (arable, dairy, livestock, market garden/

horticulture, other)& Organic status (yes/no)

& Type of agri-environment scheme [none, entry levelstewardship, higher level stewardship, organic entry levelstewardship, other (for more information on the differentlevels please refer to URL: http://www.naturalengland.org.uk/ourwork/farming/funding/es/default.aspx accessed06-11-2010)]

& Duration of agri-environment scheme (just started,<5 years, >5<10 years, >10 years)

The classes of the percentage of grassland, woodland andarable land were converted to mean number of hectares basedon the stated size of the farm when available. Unfortunately, afew respondents failed to state the size of their farm; we had toomit those records. We chose to include both the ‘farm type’and the ‘dominant habitat type on the farm’ into the models asseparate variables, since the type of farm did not necessarilyrelate to its dominant habitat. Many farms are for instancebased upon a mixture of activities.

In addition to the above mentioned variables obtainedfrom the questionnaire, other variables (predator presence,landscape and habitat features not covered by the question-naire and soil data), which we thought might have animpact on hedgehog presence on farms, were obtained fromvarious sources. Table 1 shows all variables used and from

Variable Explanation Source

Arable and horticulture Area of arable and horticultural land in 2000 (ha) BCS

Badger Relative abundance of badgers 2003–2006 (no. of animals) LWM

Broadleaf Area of broadleaved woodland in 2000 (ha) BCS

Built-up Area of built-up in 2000 (ha) BCS

Conifer Area of coniferous woodland in 2000 (ha) BCS

Farm habitat Amount of arable/woodland/grassland on the farm (ha) Q

Farm type Arable/dairy/horticulture/livestock/poultry Q

Hectares Land area of the farm (ha) Q

Hedgehog Relative hedgehog abundance (estimated no.) Q

Hedgerow Length of hedgerows (km) Q

Improved grass Area of improved grassland in 2000 (ha) BCS

Major roads Major roads coverage in 2000 (km) BCS

Minor roads Minor roads coverage in 2000 (km) BCS

Neutral grass Area of land characterised by vegetation dominated bygrasses and herbs in 1998 (ha)

BCS

Residential urban Coverage residential buildings, gardens and groundsin 2000 (ha)

BCS

Scheme Presence of an agri-environment scheme on the farmland (yes/no)

Q

Semi-natural Area of semi-natural grassland (an amalgamation ofneutral, calcareous, and acid Grassland, bracken andfen, marsh and swamp) in 2000 (ha)

BCS

Soil type The soil types of England (a) soils with a clayey texture,(b) soils with a peaty texture, (c) soils with a sandy texture,(d) soils with a loamy texture and rich in lime, (e) soilswith a loamy texture and a low fertility, (f) soils with aloamy texture and a moderate to high fertility

NSRI

Upland Area of dwarf shrub heath, bog, montane/inland rock in 2000 (ha) BCS

Table 1 Variables used for theanalyses of the presence ofhedgehogs on farms in GreatBritain

BCS British Countryside Survey2000, LWM Living withMammals Survey, NSRINational Soil Research Institute,Q questionnaire

Acta Theriol (2012) 57:79–88 81

which source we obtained them. Data on badger presence,the main predator of hedgehogs in Great Britain (Reeve1994; Young et al. 2006), were derived from the survey‘Living with Mammals’ from the ‘Mammals Trust UK’,that recorded, amongst others, badger presence in greenspaces throughout Great Britain between 2003 and 2006.These data were spatially linked to the questionnaire bymeans of geostatistical interpolation. Environmental data,i.e. landscape features and habitat types, were derivedfrom the British Countryside Survey 2000. Landscapefeatures and habitat types were selected based upon theirbiological relevance to hedgehogs. In addition to thevariable ‘farm habitat’ (obtained from the questionnaire),which provided detailed information about the farm, wechose to include landscape variables such as ‘the area ofimproved grassland’ (obtained from the British CountrysideSurvey 2000), which provided information on the landscapelevel, as well. Soil data for England were obtained from theNational Soil Research Institute. All environmental data andthe soil data were available at the 1-km2 level, which was alsoused for the representation of the respondents to thequestionnaire. Not any of the variables showed strongcollinearity (r≤0.60; Graham 2003) with one another. Otherfactors such as invertebrate abundance, rainfall, temperatureand pollutants might impact geographical differences inhedgehog presence as well. Unfortunately, detailed data werenot available; we could therefore not include them in theanalyses.

Of all variables used for the analyses (see Table 1), thecoverage of roads was thought to be negatively related to thepresence of hedgehogs, since they add to habitat fragmen-tation and increase the risk of mortality due to traffic (Clarket al. 2001; Forman and Alexander 1998; Huijser 2000).Unfortunately, detailed data on traffic flow were notavailable. The areas of arable and horticultural land, built-up, coniferous woodlands and of upland were expected torelate negatively to hedgehog presence. Hedgehogs arefrequently found to avoid landscapes with a high percentageof these habitat types in radio-tracking studies (e.g. Huijser2000; Reeve 1994). The area of broadleaved woodland andthe amount of hedgerows, on the other hand, were expectedto relate positively to the presence of hedgehogs. Thesehabitats are frequently thought to be favoured by hedgehogs(e.g. Doncaster 1992; Huijser 2000; Reeve 1994; Riber2006). Results from previous studies are divided with regardto the suitability of grassland for hedgehogs (Doncaster1992; Huijser 2000). It was expected that badgers would benegatively related to the presence of hedgehogs. Residentialurban was expected to relate positively to hedgehogpresence, since hedgehogs are often associated with urbansettlements (Reeve 1994). The impact of agri-environmentschemes was also expected to be positive due to theassociation with vegetation structures, such as hedgerows,

that are often thought to be favourable for hedgehogs. Weincluded soil data in the analyses since the texture of the soilcan have a large influence on the abundance of invertebratessuch as earthworms, which in turn may have implicationsregarding the presence of species such as the hedgehog andthe badger that feed on them (Neal and Cheeseman 1996;Reeve 1994). Unfortunately, the organic status of the farmcould not be used in the analyses due to lack of sufficientrespondents with an organic farm (4%, n=42).

Statistical analyses

Generalized linear modelling (GLM) of binomial proportionswith logit transformation function was used in GenStat (forWindows eighth edition, VSN International Ltd, LawesAgricultural Trust, Oxford, UK) for the analyses of thepresence of hedgehogs. GLM with the normal distributionand identity link function was used to analyse the presence ofhedgehogs. Names and explanations of the variables used inthe analyses are given in Table 1. The models of best fitwere selected using the backward stepwise method. Mapswere produced in a Geographic Information System(Mapinfo Professional Version 8, MapInfo Corporation,New York, USA). The variable built-up was used as aweighting factor to account for differences in the amountof area available for farmland at the 1-km2 level. Built-up,available at the 1-km2 scale, was used rather than size ofthe farm since we worked on the 1-km2 scale and not onthe farm scale. We made the assumption that when morearea was used for built-up, less area was available forfarmland. Statistical analyses other than GLM wereconducted using SPSS (for Windows 14th edition, SPSSInc., Chicago, USA).

Results

Presence of hedgehogs

Of the 4,000 questionnaires sent out, 26% (n=1054) werereturned. Figure 2 shows the distribution of the respond-ents. Questionnaires were returned from 724 (26%) of the10-km2 grid cells present in Great Britain. Over 310,000 haof farmland was covered by the farms of the respondentscomprising about 1.8% of the total agricultural area inGreat Britain in 2006. Although the return rates were biasedgeographically on a regional basis [median, 24%; range20–33%, (chi-square test), χ2=23.559, df=10, p=0.009],they were not biased geographically on the environmentalzone scale [median, 26%; range 23–31%, (chi-square test),χ2=8.116 df=5, p=0.115)]. Since analyses were carried outon the latter scale, we did not omit returned questionnairesfrom the analyses.

82 Acta Theriol (2012) 57:79–88

Hedgehogs were seen by 55% (n=580) of the respond-ents. Figure 3 shows the total number of respondents perregion and the percentage of those who had seen hedge-hogs. The majority of the respondents had seen hedgehogson their farm in all northern regions. In the southernregions, except for East England and the South East ofEngland, the majority of people did not see hedgehogs ontheir farm (chi-square test, χ2=49.803, df=10, p<0.001).The respondents most often encountered hedgehogs in theirgarden followed by grassland and deciduous woodland.Hedgehogs were, on the other hand, least often seen inconiferous woodland and in heathland.

A large range of variables had a negative, a positive or anambiguous impact on the presence of hedgehogs on farms.Table 2 displays in how many models a particular variableappeared as a positive explaining factor, as a negativeexplaining factor and the respective average partial r2. The

Fig. 3 Percentage of respondents to the questionnaire that had seenhedgehogs on their farm in Great Britain, shown per region. Datalabels show the number of respondents. EE East of England, EM EastMidlands, L London, NE Northeast, NW Northwest, S Scotland, SESoutheast, SW Southwest, W Wales, WM West Midlands, Y Yorkshireand The Humber

Fig. 2 The locations of the respondents (n=1054) to the question-naire. Black dots (n=580) represent farmers who saw hedgehogs, greydots (n=474) represent farmers who did not see hedgehogs.Questionnaires were not sent to offshore islands except the Isle ofWight

Table 2 Table displaying in how many models a particular variableappeared as a positive explaining factor, as a negative explainingfactor and the respective average partial r2

Variable Numberof modelspositive

Numberof modelsnegative

Averagepartial r2

positive

Averagepartial r2

negative

Upland 3 0 0.024 n.a.

Hedgerow 3 0 0.006 n.a.

Improved grass 2 0 0.033 n.a.

Farm habitat: arable 1 0 0.017 n.a.

Farm type: horticulture 1 0 0.007 n.a.

Farm habitat: grass 1 0 0.006 n.a.

Residential urban 1 0 0.005 n.a.

Farm type: poultry 1 0 0.002 n.a.

Badger 0 4 n.a. 0.114

Farm habitat: wood 0 3 n.a. 0.029

Farm type: livestock 0 2 n.a. 0.148

Soil type: clayey 0 2 n.a. 0.018

Neutral grass 0 1 n.a. 0.004

Arable and horticulture 3 2 0.048 0.431

Minor roads 2 2 0.392 0.009

Broadleaf 2 1 0.199 0.060

Farm type: arable 2 1 0.035 0.073

Soil type: peaty 2 1 0.013 0.086

Farm type: dairy 2 1 0.005 0.008

Major roads 1 3 0.113 0.055

Conifer 1 1 0.002 0.044

Semi-natural 1 1 0.014 0.005

n.a. not applicable

Acta Theriol (2012) 57:79–88 83

variables upland and hedgerow appeared in three models andwere consistently positive. Of these two, hedgerow onlyseemed to have a minor positive effect. When we alsoinclude the ambiguous variables, the variable minor roadshad on average the strongest positive effect on the presenceof hedgehogs on farms. However, minor roads were only inthe uplands of England, Wales and Scotland positivelyrelated to hedgehog presence (Tables 3 and 4). The relativeabundance of badgers appeared most frequently in themodels followed by the amount of woodland on the farmand consistently as negative variables. When we also includethe ambiguous variables, the variable area of arable andhorticulture had on average the strongest negative effect onthe presence of hedgehogs on farms. However, this variableappeared in five models, and it showed a positive effect inthree of those. The area of arable and horticulture was onlynegatively related to the presence of hedgehogs in the (true)upland areas and islands of Scotland (see below and Tables 3and 4). Furthermore, the variable major roads appeared as anegative variable in three models, although it showed astrong positive relation to hedgehog presence in one model;the uplands of England and Wales.

Table 3 shows the summary of the GLM for GreatBritain. For the whole of Great Britain, only 35% of thevariance in the presence of hedgehogs was explained. Thevariable area of upland followed by area of arable andhorticulture had the strongest positive impact. The strongestnegative effect on the presence of hedgehogs was explainedby the variable coverage of major roads followed bylivestock farms and relative badger abundance.

Table 4 shows the summaries of the GLM for the sixenvironmental zones. The GLMs for the environmentalzones explained on average 49% of the variance inhedgehog presence. In the arable-dominated lowlands ofEngland and Wales, the strongest negative variable was

livestock farms. The variable farm habitat arable showedon the other hand the strongest positive effect. In thepasture-dominated lowlands of England and Wales, thestrongest negative variable was sandy soils, whilst the areaof arable and horticulture was the strongest positivevariable. In the uplands of England and Wales, 60% ofthe variance was explained by three variables: the coverageof minor roads, coverage of major roads and the relativeabundance of badgers. Only the variable relative abun-dance of badgers was negatively related to the presence ofhedgehogs. In the lowlands of Scotland, the variables areaof broadleaved woodland and coniferous woodland showedthe strongest positive relations and coverage of major roadsthe strongest negative relation to hedgehog sightings. Intotal, 60% of the variation in hedgehog presence wasexplained in the marginal uplands and islands of Scotlandby coverage of minor roads, area of broadleaved woodland,area of arable and horticulture and the coverage of majorroads. The first two variables were positively correlatedwith the presence of hedgehogs, whilst the latter two hadnegative impacts. In the true uplands of Scotland, thevariable area of arable and horticulture was the sole factorin the minimum adequate model explaining over 57% ofthe variance in hedgehog presence. Hedgehogs were seenless often in areas that had a high amount of arable landand/or horticulture.

Discussion

A questionnaire was used to collect data on the distributionof hedgehogs in agricultural landscapes. The target group,landowners, was not trained at identifying hedgehogs anddid not especially aim to find them. Therefore, the presenceof hedgehogs might have been inaccurate. Hedgehogs,

Model summary Variable Estimate Partial r2 p

Explained, 35%, n=1050, p<0.001 Constant 1.768 <0.001

Major roads −0.622 0.098 <0.001

Farm type: livestock −1.642 0.056 <0.001

Badger −0.642 0.041 <0.001

Upland 2.539 0.015 <0.001

Arable and horticulture 0.552 0.013 <0.001

Farm habitat: wood −0.887 0.013 <0.001

Soils: clayey −1.812 0.012 <0.001

Soils: peaty 2.598 0.012 <0.001

Hedgerow 0.444 0.009 <0.001

Farm type: arable 0.617 0.007 <0.001

Farm type: dairy 0.843 0.006 <0.001

Conifer −0.546 0.002 0.014

Farm type: poultry 0.701 0.002 0.023

Table 3 Summary of the GLMof hedgehog presence on farmsin Great Britain

84 Acta Theriol (2012) 57:79–88

however, are easily recognized, and the assumption wasthat landowners do in general display knowledge aboutpresence of wildlife on their property. Thus, it was felt that

the respondents were able to give a reasonable estimation ofthe presence of hedgehogs on their holdings and valuabledata could come from a questionnaire survey. Nevertheless,

Model summary Variable Estimate Partial r2 p

Arable-dominated lowlands ofEngland and Wales explained,42%; n=294; p<0.001

Constant 5.024 <0.001

Farm type: livestock −3.490 0.239 <0.001

Farm type: arable −2.528 0.073 <0.001

Broadleaf −0.897 0.060 <0.001

Badger −0.662 0.041 <0.001

Soils: clayey −1.983 0.023 <0.001

Farm habitat: arable 1.477 0.017 <0.001

Soils: sandy 1.691 0.014 <0.001

Minor roads −0.487 0.013 <0.001

Farm type: dairy −0.994 0.008 0.002

Farm habitat: wood −0.490 0.008 0.004

Farm type: horticulture 1.052 0.007 0.006

Upland 1.366 0.007 0.008

Semi-natural −0.483 0.005 0.014

Pasture-dominated lowlands ofEngland and Wales explained,54%; n=274; p<0.001

Constant 0.745 0.036

Arable and horticulture 1.698 0.113 <0.001

Soils: sandy −6.379 0.086 <0.001

Farm habitat: wood −4.340 0.066 <0.001

Farm type: arable 2.582 0.064 <0.001

Badger −1.320 0.057 <0.001

Upland 6.625 0.050 <0.001

Improved grass 0.884 0.032 <0.001

Minor roads −0.243 0.006 0.005

Farm type: dairy 0.916 0.004 0.023

Hedgerow 0.885 0.004 0.030

Uplands of England and Walesexplained, 60%; n=73; p<0.001

Constant −5.640 0.046

Minor roads 2.040 0.348 0.006

Badger −10.77 0.317 0.025

Major roads 2.600 0.113 0.017

Lowlands of Scotland explained,22%; n=252; p<0.001

Constant 2.569 <0.001

Conifer 1.397 0.073 <0.001

Broadleaf 1.377 0.071 <0.001

Improved grass 1.231 0.046 <0.001

Arable and horticulture 0.857 0.017 <0.001

Semi-natural 0.555 0.016 <0.001

Major roads −0.330 0.011 <0.001

Neutral grass −0.644 0.009 0.003

Uplands and islands of Scotlandexplained, 60%; n=90; p<0.001

Residential urban 0.410 0.006 0.015

Hedgerow 1.729 0.005 0.032

Constant −1.547 0.031

Minor roads 5.580 0.436 <0.001

Broadleaf 6.370 0.330 <0.001

Arable and horticulture −7.760 0.289 <0.001

Major roads −1.468 0.060 0.014

The true uplands of Scotland explained,57%; n=47; p<0.001

Constant 4.560 0.006

Arable and horticulture −0.264 0.572 0.044

Table 4 Summary of the GLMsof hedgehog presence ofhedgehogs on farms perenvironmental zone

Acta Theriol (2012) 57:79–88 85

differences in visibility and in the amount of time alandowner spent on the field could not be accounted forand might have biased the dataset. The respondents werespread throughout the country, not geographically biasedper environmental zone, and respondents that had seenhedgehogs were not highly overrepresented (55%). It wastherefore assumed that respondents with a particular interestin, or those that had hedgehogs on their farm, were notparticularly more inclined to return questionnaires, and thusthat the dataset was a fair indication of the presence ofhedgehogs in British agricultural landscapes.

The presence of hedgehogs was highest in the northernand eastern regions of England and lowest in the southernand western regions, which is in agreement with resultsfound by Hof and Bright (2009). We were further able toconclude from this study that a large range of variables hadan impact on the presence of hedgehogs on farms.Moreover, the impact of several variables was ratherambiguous, displaying positive effects in some environ-mental zones and negative effects in other. This was likelycaused by geographic variations and environmental variationswithin the variable, e.g. the relative abundance of badgers wasconsistently negatively related to the presence of hedgehogs,which concurs with previous findings (Doncaster 1992;Young et al. 2006). In areas where badgers are lessabundant, the presence of hedgehogs might be driven byother environmental variables. In this the geographicscale on which the study took place is important.Specific local landscape features might be highlyimportant in determining the presence of hedgehogs ona small scale but be irrelevant on the countrywide scale.It is also likely that factors that were not taken intoaccount, such as differences in rainfall, temperature,pollutants in the soil and habitat fragmentation, wereinteracting with the variables used in the model and/orinfluencing the presence of hedgehogs directly. Althoughvariables did not show strong collinearity (r≤0.60,Graham 2003), hedgehogs might respond to a feature inthe landscape depending on the surrounding landscape.Nevertheless, a small number of variables frequently, orstrongly, had an either positive (upland, hedgerow andminor roads) or negative (relative abundance of badgers,major roads and the amount of woodland on the farm)impact on the presence of hedgehogs. We discuss thosevariables below.

Hedgehogs are frequently associated with edge habitat(e.g. Huijser 2000; Morris 1986), and Hof and Bright(2010) found that hedgehogs were frequently found nearhedgerows in arable landscapes. Although hedgerows werepositively associated with the presence of hedgehogs inseveral models, we could not find a strong benefit.Hedgerows only had a slight positive relation withhedgehog presence on farms. The role of woodland in

other hedgehog studies is not clear (Berthoud 1982;Doncaster 1992; Riber 2006). The results derived from thisstudy were ambiguous as well; although the amount ofwoodland on the farm was consistently negatively related tothe presence of hedgehogs, both the areas of coniferous andof broadleaved woodland in the wider landscape were not.The presence or absence of hedgehogs within forested areasmight be explained by differences in factors such as size ofthe habitat patch and clearings within forested areas, andthus presence of edge habitat (Huijser 2000; Riber 2006).Large woodlands with a low number of clearings offer littleedge habitat and are often avoided by hedgehogs, whilstsmall patches of woodlands on the other hand will increasethe amount of edge habitat and are perhaps more attractive.Other factors that might affect the avoidance or preferenceof woodlands are undergrowth and predator and preydensities (Doncaster 1992; Micol et al. 1994; Young et al.2006). It was surprising that upland related positively tohedgehog presence, since hedgehogs are frequently foundto avoid landscapes with a high percentage of upland inradio-tracking studies (e.g. Huijser 2000; Reeve 1994).Other factors might have driven hedgehogs to areas that areusually not preferred by them. The presence of competitorsor predators may for instance drive species to less suitablehabitat (Hilton et al. 1999; Sweitzer 1996).

The fact that a high coverage of minor roads was relatedto hedgehog presence in the uplands might be partly causedby the high visibility of hedgehogs on roads. Additionally,minor roads are abundant in small-scale landscapes that areoften characterised by high densities of hedgerows andwoodlands; habitat types which are frequently thought to bepreferred by hedgehogs (Huijser 2000; Morris 1986; Riber2006). Furthermore, roads also create habitat edges, whichmay attract hedgehogs. However, major roads, which weremostly negatively related to hedgehog presence in severalenvironmental zones and in Great Britain as a whole, forma greater barrier for hedgehogs (Bontadina 1991; Rondininiand Doncaster 2002). Furthermore, the amount of traffic isexpected to be higher on major roads, which might result inmore traffic victims amongst hedgehogs. It is unexpectedthat major roads, together with minor roads, were positivelyrelated to hedgehog presence in the uplands of England andWales. The negative impact of badgers might have beenparticularly strong in this environmental zone, which wasalso indicated by our results.

It is not surprising that badgers, being a predator andcompetitor of hedgehogs (Reeve 1994; Young et al. 2006),had a negative impact on hedgehogs. Several studies showstrong local impacts of badgers upon hedgehogs (Doncaster1992; Young et al. 2006). The results from this studyconcur and suggest that the abundance of badgers had anegative impact on the presence of hedgehogs on a widerscale as well.

86 Acta Theriol (2012) 57:79–88

Conservation strategies

The abundance of badgers, major roads, and, to a lower extent,minor roads, negatively affected hedgehog presence; conser-vation strategies should therefore be aimed at lessening thoseimpacts. Fencing major roads to reduce road mortality hadadverse effects on hedgehog populations due to increasedhabitat fragmentation in The Netherlands (Huijser 2000), andis therefore not a sustainable option. Wildlife passages on theother hand, frequently used in various countries, but so farlargely unexploited in Great Britain (Jackson and Griffin2000), may provide hedgehogs, and other taxa, with safepassages across roads. Badgers may negatively affect the useof tunnels by hedgehogs due to the latter avoiding responseto badger odour (Doncaster 1999; Ward et al. 1997). It mighthowever be possible to adjust the size of tunnels to reducetheir use by badgers.

Badgers frequently showed a relatively strong negativeimpact on hedgehog presence. However, both practical, thebadger is protected by law in the UK, and ethical issues arelikely to arise in culling one species in order to protectanother. Furthermore, it is generally found that althoughpredation might be an important limiting factor in preypopulations, it is unlikely to drive populations to extinctionon a wide scale (MacDonald et al. 1999). Moderate or lowlevels of predation generally have little influence ifpopulation growth rates remain high (MacDonald et al.1999), and predator control is not a guaranteed successfulmitigation measure to stabilize/increase numbers of preyspecies on the long term (Harding et al. 2001). It thereforeseems imperative to seek more effective and non-lethalmethods to preserve hedgehogs. Habitat management inorder to reduce the impact of predators can be beneficial forprey species (Finke and Denno 2002; Janssen et al. 2007)and should therefore be considered as a conservationstrategy for the hedgehog. Increasing the number andavailability of sites that offer coverage by increasing thecomplexity of the habitat structure may be beneficial tohedgehogs. This can be done by establishing, for instance,more and denser hedgerows, habitat features that werepositively associated with hedgehogs in this study. Further-more, increasing the connectivity between suitable habitatsmight also lessen the negative impact of badgers onhedgehogs. However, the likelihood of success of increas-ing the habitat structure is uncertain. It is currently notknown if complex and dense habitat structures providehedgehogs with the level of protection necessary to escapethe risk of predation.

Acknowledgements We wish to thank all respondents to thequestionnaire. The entire project was funded jointly by the People’sTrust of Endangered Species and the British Hedgehog PreservationSociety to whom we are greatly indebted. We would also like to thankthree anonymous referees for their useful comments.

References

Berthoud G (1982) Contribution à la biologie du hérisson (Erinaceuseuropaeus) et application à sa protection. PhD Thesis, UniversityNeuchâtel

Bontadina F (1991) Straßenüberquerungen von Igeln (Erinaceuseuropaeus). PhD Thesis, University of Zurich

Bunce RGH, Barr CJ, Clarke RT, Howard DC, Lane AMJ (1996) ITEMerlewood Land Classification of Great Britain. J Biogeogr23:625–634. doi:10.1111/j.1365-2699.1996.tb00023.x

Clark B, Clark B, Johnson L, Haynie M (2001) Influence of roads onmovements of small mammals. Southwest Nat 46:338–344

Davey P, Aebischer N (2006) Contract no: F90-01-708 Participationof the National Gamebag Census in the Mammal SurveillanceNetwork. A report to Joint Nature Conservation Committee forthe year 2005/06

Doncaster C (1992) Testing the role of intraguild predation inregulating hedgehog populations. Proc R Soc Lond B Biol Sci249:113–117. doi:10.1098/rspb.1992.0092

Doncaster CP (1999) Can badgers affect the use of tunnels byhedgehogs? A review of the literature. Lutra 42:59–64

Finke DL, Denno RF (2002) Intraguild predation diminished incomplex-structured vegetation: implications for prey suppression.Ecology 83:643–652. doi:10.1890/0012-9658(2002)083[0643:IPDICS]2.0.CO;2

Forman RTT, Alexander LE (1998) Roads and their major ecologicaleffects. Annu Rev Ecol Syst 29:207–231. doi:0066-4162/98/1120-0207$08.00

Gaston K, Fuller R (2008) Commonness, population depletion andconservation biology. Trends Ecol Evol 23:14–19. doi:10.1016/j.tree.2007.11.001

Graham MH (2003) Confronting multicollinearity in ecological multipleregression. Ecology 84:2809–2815. doi:10.1890/02-3114

Harding EK, Doak DF, Albertson JD (2001) Evaluating theeffectiveness of predator control: the non-native red fox as acase study. Conserv Biol 15:1114–1122. doi:10.1046/j.1523-1739.2001.0150041114.x

Hilton G, Ruxton G, Cresswell W (1999) Choice of foraging area withrespect to predation risk in redshanks: the effects of weather andpredator activity. Oikos 87:295–302

Hof A (2009) A study of the current status of the hedgehog (Erinaceuseuropaeus), and its decline in Great Britain since 1960. PhDthesis, Royal Holloway, University of London

Hof AR, Bright P (2009) The value of green-spaces in built-up areasfor hedgehogs. Lutra 52:69–82

Hof A, Bright P (2010) The value of agri-environment schemes formacro-invertebrate feeders: hedgehogs on arable farms in Britain.Anim Conserv 13:467–473. doi:10.1111/j.1469-1795.2010.00359.x

Huijser MP (2000) Life on the edge. Hedgehog traffic victims andmitigation strategies in an anthropogenic landscape. PhD thesis,Wageningen University

Jackson SD, Griffin CR (2000) A strategy for mitigating highwayimpacts on wildlife. In: Messmer TA, West B (eds) Wildlifeand highways: seeking solutions to an ecological and socio-economic dilemma. The Wildlife Society, Washington, D.C.,USA, pp 143–159

Janssen A, Sabelis MW, Magalhães S, Montserrat M, van derHammen T (2007) Habitat structure affects intraguild predation.Ecology 88:2713–2719. doi:10.1890/06-1408.1

Krebs J, Wilson J, Bradbury R, Siriwardena G (1999) The secondsilent spring? Nature 400:611–612. doi:10.1038/23127

MacDonald DW, Mace GM, Barretto GR (1999) The effects ofpredators on fragmented prey populations: a case study for theconservation of endangered prey. J Zool 247:487–506.doi:10.1111/j.1469-7998.1999.tb01012.x

Acta Theriol (2012) 57:79–88 87

McNeely JA, Gadgil C, Leveque C, Padoch C, Redford K (1995)Human influences on biodiversity. In: Heywood VH (ed)Global biodiversity assessment. United Nations environmentprogramme. Cambridge University Press, Cambridge, pp 711–823

Micol T, Doncaster C, Mackinlay L (1994) Correlates of localvariation in the abundance of hedgehogs Erinaceus europaeus.J Anim Ecol 63:851–860. doi:10.1111/j.1469-7998.2006.00078.x

Morris PA (1986) Nightly movements of hedgehogs (Erinaceuseuropaeus) in forest edge habitat. Mammalia 50:395–398

Neal EG, Cheeseman CL (1996) Badgers. T & A D Poyser Ltd,London

Pimm SL, Russell GJ, Gittleman JL, Brooks TM (1995) The futureof biodiversity. Science 269:347–350. doi:10.1126/science.269.5222.347

Reeve NJ (1994) Hedgehogs. T & A D Poyser Ltd, LondonRiber A (2006) Habitat use and behaviour of European hedgehog

Erinaceus europaeus in a Danish rural area. Acta Theriol51:363–371

Robinson R, Sutherland W (2002) Post-war changes in arable farmingand biodiversity in Great Britain. J Appl Ecol 39:157–176.doi:10.1046/j.1365-2664.2002.00695.x

Rondinini C, Doncaster C (2002) Roads as barriers to movement forhedgehogs. Funct Ecol 16:504–509. doi:10.1046/j.1365-2435.2002.00651.x

Sala OE, Chapin Iii FS, Armesto JJ, Berlow E, Bloomfield J, Dirzo R,Huber-Sanwald E, Huenneke LF, Jackson RB, Kinzig A,Leemans R, Lodge DM, Mooney HA, Oesterheld M, LeRoyPN, Sykes MT, Walker BH, Walker M, Wall DH (2000) Globalbiodiversity scenarios for the year 2100. Science 287:1770–1774.doi:10.1126/science.287.5459.1770

Sweitzer R (1996) Predation or starvation: consequences of foragingdecisions by porcupines (Erethizon dorsatum). J Mammal77:1068–1077

Ward JF, MacDonald DW, Doncaster CP (1997) Responses offoraging hedgehogs to badger odour. Anim Behav 53:709–720.doi:10.1006/anbe.1996.0307

White P, Jennings N, Renwick A, Barker N (2005) Questionnaires inecology: a review of past use and recommendations for best practice.J Appl Ecol 42:421–430. doi:10.1111/j.1365-2664.2005.01032.x

Young R, Davison J, Trewby I, Wilson G, Delahay R, Doncaster C(2006) Abundance of hedgehogs (Erinaceus europaeus) inrelation to the density and distribution of badgers (Meles meles).J Zool 269:349–356. doi:10.1111/j.1469-7998.2006.00078.x

88 Acta Theriol (2012) 57:79–88