Are motorway wildlife passages worth building? Vertebrate use of road-crossing structures on a Spanish motorway

ARTICLE IN PRESS

0301-4797/$ - se

doi:10.1016/j.je

�CorrespondE-mail addr

Journal of Environmental Management 88 (2008) 407–415

www.elsevier.com/locate/jenvman

Are motorway wildlife passages worth building? Vertebrate use ofroad-crossing structures on a Spanish motorway

C. Mata�, I. Hervas, J. Herranz, F. Suarez, J.E. Malo

Departamento de Ecologıa, Facultad de Ciencias, Universidad Autonoma de Madrid, E-28049 Madrid, Spain

Received 21 June 2006; received in revised form 31 January 2007; accepted 10 March 2007

Available online 30 April 2007

Abstract

Numerous road and railway construction projects include costly mitigation measures to offset the barrier effect produced on local

fauna, despite the scarcity of data on the effectiveness of such mitigation measures. In this study, we evaluate the utility of different types

of crossing structures. Vertebrate use of 43 transverse crossing structures along the A-52 motorway (north-western Spain) was studied

during spring 2001. Research centered on wildlife passages (9), wildlife-adapted box culverts (7), functional passages (6 overpasses, 7

underpasses) and culverts (14), with marble dust being used to record animal tracks. A total of 424 track-days were recorded, with most

of the larger vertebrate groups present in the area being detected. All crossing structure types were used by animals, although the

intensity of use varied significantly among them (Kruskal–Wallis test, po0.05); culverts were used less frequently than other structures.

Crossing structure type and width were identified as the most important factors in their selection for use. Wildlife passages and adapted

culverts allowed crossing by certain species (wild boar, roe deer, Eurasian badger), which do not tend to cross elsewhere. These results

highlight the importance of using both mixed-type structures and wildlife passages in reducing the barrier effect of roads.

r 2007 Elsevier Ltd. All rights reserved.

Keywords: Barrier effect; Habitat fragmentation; Road ecology; Vertebrate; Wildlife crossing structures

1. Introduction

Animal population isolation produced by habitatfragmentation constitutes one of the commonest causesof local extinctions (Hunt et al., 1987; Clarke et al., 1998;Lode, 2000; Fahrig, 2003). Linear infrastructures not onlyphysically destroy habitats but also become barriers, whichconsiderably limit movement and dispersion of terrestrialvertebrates (Oxley et al., 1974; Mader, 1984; Camby andMaizeret, 1987, Walker et al., 2003). As a result, over thelast few decades, the inclusion of faunal passages or themodification of existing culverts to serve as crossingstructures has frequently been recommended in Environ-mental Impact Assessments and Statements (Beier andLoe, 1992; De Santo and Smith, 1993; van Bohemen,1998). Wildlife crossing structures aim to re-establish themovement of animals between both sides of a road orrailway line, helping to connect the areas affected by the

e front matter r 2007 Elsevier Ltd. All rights reserved.

nvman.2007.03.014

ing author. Tel./fax: +34 914978011.

ess: [email protected] (C. Mata).

transport corridor (Saunders and Hobbs, 1991; Clergeau,1993; Rodrıguez et al., 1996, Putman, 1997).Knowledge of the effectiveness of wildlife crossing

structures is still scarce, although increasing quickly. Theregular use of different wildlife passage types has beendemonstrated for some vertebrate groups (Singer andDoherty, 1985; Foster and Humphrey, 1995; Bekker andCanters, 1997; Gloyne and Clevenger, 2001; Puky andVogel, 2003; Taylor and Goldingay, 2003; Dodd et al.,2004) but others, like ungulates, are noteworthy for theirreluctance to use most crossing structures (Reed, 1981;Vassant et al., 1993; Mata et al., 2005). The key elementsfor the design of effective mitigation measures are not yetwell understood and different solutions may be needed fordifferent faunal groups. The modification of constructionprojects to adapt culverts for vertebrates or to includewildlife passages imposes costs on new projects. Conse-quently, it is important to evaluate which characteristics ofexisting crossing structures affect their use by differentspecies, in order to design cost-effective passages for targetspecies in new projects. It is therefore important to

ARTICLE IN PRESSC. Mata et al. / Journal of Environmental Management 88 (2008) 407–415408

determine the actual use of over- and underpassesspecifically designed for wildlife, due to the high construc-tion costs they represent (Rosell et al., 2003).

The present study had two objectives. Firstly, to discoverwhich vertebrate groups use road crossing structures. Thiswas determined through an analysis of structures specifi-cally designed for wildlife, as well as normal structuralfeatures of the highway, such as culverts, which aredesigned for other purposes but which could also beimportant for wildlife (Hunt et al., 1987; Yanes et al., 1995;Rodrıguez et al., 1996). Secondly, we looked at the relativeimportance of crossing structure design since structuralparameters of passages (e.g. their width and whether theyare over-or underpasses) could determine their effective useby wildlife (Olbrich, 1984; Clevenger et al., 2001). Of the 43crossing structures considered here only 16 were includedin our earlier study (Mata et al. 2005) and the present datawas obtained 1 year earlier.

2. Materials and methods

2.1. Study area

The study was undertaken along a stretch of the SpanishA-52 motorway between Camarzana de Tera (Km 34,Zamora province) and Orense (Km 217, Fig. 1). This fourlane highway is fenced along its entire length. It wasopened to traffic in 1998 and carries approximately 4500vehicles daily, 23% of them heavy vehicles.

The first few kilometers pass through the flat countrysideof the northern Spanish plain. The relief increasesgradually up to the town of Puebla de Sanabria, afterwhich the landscape is mountainous (720–960ma.s.l.). Theincrease in elevation coincides with a decrease in tempera-ture and an increase in rainfall (from 400 to 950mm). Thevegetation along the first 10 km is dominated by

Orense

A-52

Verín

PORTUGAL

20 Km

N-122

N-631

A-52

Ginzo de

Limia

SPAIN

Cama

de T

Puebla de

SanabriaRequejo

Fig. 1. Location of the study area and the monitored sections

continental Holm oak (Quercus rotundifolia), accompaniedby gum cistus (Cistus ladanifer), broom spp. (Cytisus

scoparius and Cytisus multiflorus), low shrubs (Cistus

laurifolius, Halimium umbellatum, Thymus mastichina,

Thymus zygis, Lavandula stoechas subsp. pedunculata)and pastures of Agrostis castellana. Cultivations of maizeand non-irrigated cereals are also present but restricted tovalley bottoms. There are also vineyards of limited extent.The next 45 km passes through patches of Pyrenean oak(Quercus pyrenaica), broom spp. and other low scrub(Genista tridentata, Halimium ocymoides, Halimium la-

sianthum) and moist meadows. The remaining section (inOrense province) is flanked by forests dominated byPyrenean oak woodlands, some more heavily disturbedthan others.The progressive abandonment of fields and rural

countryside, which is especially noticeable in Zamoraprovince, is having an important impact on the fauna ofthe area. It has led to an intense regeneration of naturalvegetation, leading in turn to an increase in the populationof large herbivores, such as red deer (Cervus elaphus), roedeer (Capreolus capreolus) and wild boar (Sus scrofa).Additionally, the increase in herbivores has led to a densewolf (Canis lupus) population (Blanco, 1998). These eventsled to the construction of specific crossing structures, andthe modification of culverts along the highway, to reducewildlife mortality from traffic.

2.2. Monitoring of crossing structures

A total of 39 crossing structures were studied along57 km of road between Camarzana de Tera and Requejo(Zamora province). We also included four structuresspecifically designed or modified for wildlife (two over-passes, one underpass and one wildlife-adapted culvert), allof which were in Orense province (177.5–216.3 km). Our

Zamora

N-120

N-630

A-6

A-6N-630

rzana

era Benavente

(in bold) of the A-52 motorway in the Iberian Peninsula.

ARTICLE IN PRESSC. Mata et al. / Journal of Environmental Management 88 (2008) 407–415 409

selection of crossing structures aimed to include all thosespecifically designed for wildlife and to obtain a represen-tative sample of the remainder, including all design types(see Table 1 and Fig. 2 for design details). The maindifference between over- and underpasses designed forwildlife and other types is that the former have no vehicularaccess . Fauna-adapted culverts are characterised by theirflat bases and enlarged cross-sectional areas.

Crossing structure use was monitored late March andearly June 2001. Marble dust, a scentless substance, whichis an excellent substrate both for recording high-qualityimpressions and for its persistence, was used to obtaintracks and signs (Yanes et al., 1995). A band of dust 1mwide and 3–10mm thick was spread across and perpendi-cular to the line of each passage in its midpoint. Animaltracks present in each passage were identified and recordeddaily before being smoothed flat. Each passage was

Table 1

Characteristics of the monitored crossing structures: type, number, dimension

n Dimensions (m)

Width H

Circular culverts 14 (6) + 1.80

Adapted culverts 7 (0) 1.7–4 1

Open span underpasses 7 (2) 4–9 4

Wildlife underpasses 5 (4) 14–20 5

Overpasses 6 (2) 7–8 —

Wildlife overpasses 4 (2) 14.7–20 —

The number of them used in a previous study (Mata et al., 2005 ) is includedaOne case of 72 m.bOne case of 96 m.

1.80 m

19 m

5 m

Fig. 2. Basic design of the different crossing structure types under study: circula

overpass (bottom right). Non-wildlife engineered passages of the two last types

monitored until ten days worth of reliable records wereobtained, rejecting days when meteorological conditionsprevented the reliable identification of tracks. Althoughsome species may not be detected, a 10-day monitoringperiod has been found to be sufficient to detect a significantportion of the fauna using a passage (Malo et al., 2006).Tracks and signs were identified following Strachan

(1995), Bang and Dahlstrom (1997) and Blanco (1998).Data for those groups where identification to species wasimpossible were pooled. Consequently, apart from theindividually identified species, the following groups wererecorded: lizards (Lacerta spp., Psammodromus spp. andPodarcis spp.), ophidians (snakes), small mammals (mice,voles and shrews), water voles (Arvicola sapidus andArvicola terrestris), rats (Rattus rattus and Rattus norvegi-

cus), lagomorphs (Oryctolagus cuniculus and Lepus grana-

tensis), small mustelids (Mustela nivalis and Mustela

s and function

Main function

eight Length

35–62 Drainage

.7–3 36–45 Drainage, adapted for wildlife

-6 32–46a Rural tracks and livestock paths

–8 30–32.5b Wildlife, closed to vehicles

58–65 Rural tracks

60–62 Wildlife, closed to vehicles

in brackets.

2.0 m

r culvert (top left), adapted culvert (top right), underpass (bottom left) and

differ from wildlife passages in the absence of vehicular access to the latter.

ARTICLE IN PRESSC. Mata et al. / Journal of Environmental Management 88 (2008) 407–415410

erminea), large mustelids (Martes foina and Martes

martes), cats (Felis catus and Felis silvestris), and canids(Canis familiaris and C. lupus).

2.3. Data analysis

The number of days on which tracks of a species (orfaunal group) were registered has been considered as thebasic unit for analysis, in order to minimize the problemsof pseudoreplication deriving from counting severalrecords of a particular species from the same passage onthe same day.

The use of each crossing structure type by each species/group employed the following use index (UI):

UI ¼ ðnij=csjÞ=ðNi=CSÞ,

where nij is the number of observations (animal crossing-days) of a species/group i in one type of crossing structurej, csj is the number of structures of type j, Ni is the numberof observation days of a species (or group) i in the totalstructures, and CS is the total number of structures studiedThis index allows the frequency of tracks observed in eachpassage type and that expected taking into account allpassage types to be compared without bias due to samplesize. Therefore, it is a selection index for animal use of apassage-type independent of use frequency, with a UI valueof 1 when actual use is equal to that expected. The UI wascompared between passage types using a Kruskal–Wallistest since the assumptions for an ANOVAR were not met(Quinn and Keough, 2002).

General use patterns of the crossing structures wereanalyzed using multidimensional scaling (STATISTICA,1998) to obtain an ordination of the crossing structures as afunction of the frequencies of use by each species (animalcrossing-days). Euclidean distance was used as the index ofsimilarity between samples and ordination was carried outwith the standard procedure provided by the statisticalprogram. Given the adjustment values (stresso0.1), athree-dimensional solution was chosen.

Once the ordination had been undertaken, the presenceof separable groups of observations in the MDS diagramwas tested using a MANCOVA applied to their axialpositions (x, y, z). In this analysis, crossing structure typewas the independent variable and the position of thepassage along the road (kilometer location) the covariate.This allowed us to test whether differences in crossingstructure use were related to passage type or to thedistribution of vertebrate species along the geographicalgradient traversed by the road. Differences between thefactor levels have been tested a posteriori by testing pairs offactors in successive MANCOVAs and correcting theprobability obtained by applying the Bonferroni sequentialprobability correction (Rice, 1989). Additionally, thespecies explaining most of the variance along the three-dimensional MDS axes (x, y, z) were detected using theSpearman rank correlations, selecting those species withvalues of po0.05 after applying the Bonferroni correction.

Finally, patterns of crossing structure use were analyzedby species. The crossing frequency of the different crossingstructures by each species was analyzed using theKruskal–Wallis test, given the non-normal distribution ofthe data. This analysis was carried out for species with atleast ten observations and, since multiple comparisons wereinvolved, Bonferroni corrections were applied to correctsignificance values.The influence of structural variables on use patterns

was also analyzed with the openness index (OI),which embraces cross-sectional area and passage length(Olbrich, 1984):

OI ¼ width� height=length:

As it is not possible to give a meaningful height value tooverpasses (see, however, Clevenger and Waltho, 2005), theOI was adapted for these cases as (OI ¼ width/length).Therefore, data from both wildlife- and other underpasseswere analyzed separately from those from overpasses.The relationship between the OI of the structures and

their use (N of track-days) by different species wasmeasured through the Spearman rank correlations due todata lacking normality. In this analysis, the Bonferronicorrection was also applied.

3. Results

A total of 424 species track-days of animals wasrecorded, equivalent to 0.99 species passing per crossingstructure per day. A total of 17 different faunal species andspecies-groups used the road crossing structures (Table 2),with the red fox (Vulpes vulpes) being the most commonspecies recorded (0.27 records/day). Other canids (dogs andwolves) and Eurasian badgers (Meles meles) also had ahigh crossing frequency (0.19 detections/day and 0.15detections/day, respectively). Other species using the cross-ing structures with medium frequency were cats (0.10detections/day), lagomorphs and small mammals (each0.07 detections/day). The remaining species had lowerfrequencies (o0.04 detections/day). Records related tohuman activities showed little use by humans of moststructure types (Table 2).All crossing structure types were used by wildlife, though

there were significant differences in the levels of use ofdifferent types (Fig. 3; H5,43 ¼ 12.07; p ¼ 0.034). Over-passes had the highest use index (UI ¼ 1.5), whereascircular culverts and those culverts modified forwildlife were used less than expected (UI ¼ 0.6 and 0.96,respectively).Positions obtained in the MDS allowed us to distinguish

several groups in the distribution of the crossing structures(Fig. 3). This grouping was significantly determinedby crossing structure type (MANCOVA test, po0.01;Table 3). In contrast, the geographical location ofthese (covariate km) was not significant to the suite ofspecies using the crossing structures throughout thelength of the section of motorway. A posteriori analysis

ARTICLE IN PRESS

Table 2

Mean number of day-detections per crossing structure for species and species groups detected throughout the monitoring period (10 days for each crossing

structure)

Circular

culverts

Adapted

culverts

Open span

underpasses

Wildlife

underpasses

Overpasses Wildlife

overpasses

Mean

n ¼ 14 n ¼ 7 n ¼ 7 n ¼ 5 n ¼ 6 n ¼ 4

Species

Western hedgehog (Erinaceus

europaeus)

— — — — 0.17 — 0.02

Polecat (Mustela putorius) — 0.14 — 0.20 — — 0.05

European badger (Meles meles) 1.00 3.14 2.71 1.60 0.17 — 1.49

Small-spotted genet (Genetta genetta) — — 0.29 — — — 0.05

Red fox (Vulpes vulpes) 1.07 2.43 4.71 3.60 3.00 3.50 2.67

Roe deer (Capreolus capreolus) — — — 0.40 — — 0.05

Red deer (Cervus elaphus) — — — 0.40 0.17 — 0.07

Wild boar (Sus scrofa) — — — 0.60 0.67 2.25 0.37

Groups of species

Lizards 0.36 — 0.14 — — — 0.14

Ophidians — — — — — 0.25 0.02

Small mammals 1.57 0.14 — — 1.00 0.25 0.70

Rats 0.07 0.43 0.14 — 1.33 — 0.30

Lagomorphs — — 0.43 1.20 1.83 3.00 0.74

Small mustelids 0.29 — 0.14 0.40 0.17 — 0.19

Large mustelids — 0.14 — 0.20 — — 0.05

Cats 0.93 0.57 1.29 0.60 1.67 1.00 1.00

Canids (Canis spp.) 0.64 2.43 2.29 1.40 4.50 2.00 1.95

Human activity 0.02

People 0.93 1.43 1.00 1.20 6.33 0.75 0.05

Livestock 0.14 0.86 0.57 0.20 2.50 — 1.49

Vehicles — 0.86 1.00 — 3.50 0.50 0.05

2.0

1.5

Use Index

1.0

0.5

0.0

CC

(n=14)

AC

(n=7)

UP

(n=7)

WUP

(n=5)

OP

(n=6)

WOP

(n=4)

Fig. 3. Use index (mean+SE) recorded in the six types of crossing

structures differentiated in the study. CC: circular culverts, AC: adapted

culverts, U: underpasses; WU, wildlife underpasses, O: overpasses, and

WOP: wildlife overpasses. n: number of monitored structures.

Table 3

Results of the MANCOVA on the effect of crossing structure type with

the location (km post) as covariate on the position of the structures on

axes x, y, z of the multidimensional scaling

Factors Wilk’s l F d.f. 1 d.f. 2 p

Structure type 0.431 2.239 15 94 0.0098

Covariate (km post) 0.843 2.107 3 34 0.1175

C. Mata et al. / Journal of Environmental Management 88 (2008) 407–415 411

shows that the difference between the diverse crossingstructure types was only significant after probabilitycorrection for wildlife overpasses vs. circular culverts(Table 4).

This differential location of the crossing structure typesin the Multidimensional Scaling can be interpreted throughthe correlations between species and axes of the analysis(Table 5). Thus, the negative values on the x axis aresignificantly associated (the Spearman rank correlation,po0.05 after Bonferroni correction) with the presence ofcanids, Eurasian badger and red fox where non-wildlifeengineered underpasses occurred, whereas along thepositive part of the axis where circular culverts werelocated there were no significant associations with any ofthe species. y-Axis is positively correlated with wild boar,lagomorphs and red fox, with overpasses and wildlifeoverpasses located along the positive part of the axis. Incontrast, the two types of culverts are located along thenegative values and give no significant relationship.Finally, the positive values along the z-axis are associatedwith the presence of Eurasian badger, and adapted culverts

ARTICLE IN PRESS

Table 4

Results of a posteriori comparisons of differences between passage types and the fauna crossing them, simplified by their position on the x, y, z axes of the

multidimensional scaling

Wildlife overpasses Overpasses Wildlife underpasses Open span underpasses Adapted culverts

Circular culverts Wilk’s l 0.254 0.526 0.636 0.618 0.503

p 0.0004 0.0192 0.0878 0.0476 0.0192

Adapted culverts Wilk’s l 0.401 0.517 0.683 0.894

p 0.1180 0.1344 0.4162 0.7863

Open span underpasses Wilk’s l 0.255 0.804 0.803

p 0.0327 0.5258 0.6508

Wildlife underpasses Wilk’s l 0.701 0.594

p 0.6649 0.3396

Overpasses Wilk’s l 0.577

p 0.3935

The Wilk’s l value is given (upper line) and its probability value (lower line) corresponding to the MANCOVAs for pair comparisons. Significant results

(po0.05) after Bonferroni sequential probability correction are shown in bold.

Table 5

Mean position along the x-, y- and z-axis of the multidimensional scaling

of crossing structure types

x y z

Circular culverts 0.380 �0.284 �0.044

Adapted culverts �0.253 �0.252 0.172

Open span underpasses �0.471 0.126 0.062

Wildlife underpasses �0.083 0.155 0.164

Overpasses �0.062 0.250 �0.376

Wildlife overpasses 0.135 0.644 0.104

Table 6

Results of the comparison of different crossing structure use by species

along the A-52

Species N H43, 5 p

European badger 64 8.650 0.1246

Red fox 115 11.577 0.0638

Wild boar 16 19.350 0.0017

Species groupings

Small mammals 30 14.700 0.0121

Rats 13 4.383 0.6030

Lagomorphs 32 17.009 0.0043

Cats 43 0.915 0.9465

Canids (Canis spp.) 84 13.422 0.0264

Heterogeneity (H) values and Kruskal–Wallis probability (p) values are

shown for species with at least ten detections. N ¼ number of records of

each species for all crossing structures studied. Significant results (po0.05)

after Bonferroni sequential probability correction are shown in bold.

C. Mata et al. / Journal of Environmental Management 88 (2008) 407–415412

and wildlife underpasses principally appear along this partof the axis. However, the negative part of the axis, alongwhich overpasses are located, is not significantly related toany particular species.

The analysis by species of the use of the crossingstructures has only demonstrated a differential use by wildboar and lagomorphs (Kruskal–Wallis test, po0.05 afterBonferroni correction; Table 6 and Fig. 4). Wild boar andlagomorphs mainly crossed on wildlife overpasses (0.22and 0.30 detections/day, respectively), overpasses (0.07 and0.18 detections/day) and wildlife underpasses (0.06 and0.12 detections/day). None of them were recorded in thetwo culvert types, and only lagomorphs made some use ofunderpasses (0.04 detections/day).

Regarding the dimensions of underpasses, only smallmammals showed a significant negative relationship withthe OI (rS ¼ �0.506, po0.05 after Bonferroni correction).Accordingly, small mammals (shrews, mice and voles),preferentially used circular culverts to cross (UI ¼ 2.25).Passage use by lagomorphs, fox, wild boar, roe deer andcanids showed positive correlations with OI (Table 7)although they became non-significant after the applicationof the Bonferroni correction. Red foxes crossed principallyusing wider passages, showing a similar level of use in all ofthem (mean7standard deviation: 1.3970.28). Roe deer,with two records, exclusively used wildlife underpasses.

Similarly, other canids also had a preference for using widecrossing structures, but they also had a higher frequency ofusing multipurpose passages, independently of their posi-tion relative to the road. No significant correlations withOI were found for the species using passages over the road(Table 7).No significant relationships were found for the remain-

ing species but it is worth noting the case of the Eurasianbadger. This species almost exclusively used underpasses,and mainly wildlife adapted culverts (UI ¼ 2.24).

4. Discussion

Our results highlight the broad spectrum of specieswhich use the highway’s crossing structures and therelevance of structure adaptation for wildlife. With theexception of amphibians, garden dormice (Elyomis querci-

nus), water voles (A. sapidus and A. terrestris), moles (Talpa

occidentalis), terrapins (Emys orbicularis), red squirrels

ARTICLE IN PRESS

CC

n=14

AC

n=6

UP

n=8

WUP

n=5

OP

n=6

WOP

n=4

0

2

4

6

8

CC

n=14

AC

n=6

UP

n=8

WUP

n=5

OP

n=6

WOP

n=4

Cro

ssin

g fre

quency

Mean Lagomorphs (N=32)

ES

DS

Wildboar (N=16)

Fig. 4. Crossing frequency (N observations/10 days) of lagomorphs and wild boar in the different structure types (mean, SE, SD). CC: circular culverts,

AC: adapted culverts, UP: underpasses, WUP: wildlife underpasses, OP: overpasses, WOP: wildlife-overpasses. n: number of crossing structures.

Table 7

Results of the Spearman correlations between Openness index of each

passage and crossing frequency by species along the A-52

Species Underpasses Overpasses

rS p rS p

Western hedgehog — — �0.121 0.7402

Polecat 0.214 0.2327 — —

European badger 0.208 0.2455 �0.121 0.7402

Small-spotted genet 0.186 0.3005 — —

Red fox 0.467 0.0062 0.475 0.1651

Roe deer 0.387 0.0260 — —

Red deer 0.279 0.1162 —0.121 0.7402

Wild boar 0.401 0.0207 0.482 0.1586

Species groupings

Lizards �0.056 0.7578 — —

Ophidians — — 0.241 0.5023

Small mammals �0.506 0.0026 0.000 1.0000

Rats 0.109 0.5472 �0.422 0.2247

Lagomorphs 0.472 0.0056 0.247 0.4920

Small mustelids 0.044 0.8094 �0.121 0.7402

Large mustelids 0.227 0.2040 — —

Cats 0.073 0.6873 0.050 0.8920

Canids (Canis spp.) 0.380 0.0292 �0.465 0.1759

Openess indices for under- and over-passages are computed separately as

explained in the text. The Spearman r value (rs) is given together with its

probability value (p) and the significant results (po0.05) after Bonferroni

sequential probability correction are shown in bold.

C. Mata et al. / Journal of Environmental Management 88 (2008) 407–415 413

(Sciurus vulgaris) and Eurasian otters (Lutra lutra), tracksshowing passage of all vertebrate groups potentiallypresent in the study area were recorded (Palomo andGisbert, 2002). The absence of data for amphibians, watervoles, terrapins and Eurasian otter is biased by the surveymethod used since it does not monitor movements throughflooded passages. Most of these species have been recordedduring other studies in underpasses of different types(Rosell and Velasco, 1999; Mata et al., 2005; C. Mata,unpublished data). Cold weather conditions during thepresent study could be responsible for the absence ofrecords of amphibians and garden dormice in comparisonwith other data gathered in the area (Mata et al., 2005).

All crossing structure types were used by wildlife, boththose specifically designed for that purpose and those

intended for other uses. The use indices demonstrated arelative use above that expected for all crossing structuretypes, except for culverts. Despite this, the potentialimportance of culverts, as the commonest crossingstructures of roads and railways, should not be forgotten(e.g. a 91 km stretch of the A-52 under study has 106culverts, C. Mata, unpublished data). Such structures donot represent an additional construction cost given theirstructural function and they allow the movement of smalland medium-sized mammals from one side to the other (seealso Huijser et al., 1999; Clevenger et al., 2001; Dodd et al.2004). Moreover, culverts can be adapted at little extra costto favour wildlife, as shown by the intensive use of theadapted culverts (mainly 2m� 2m boxes) by the Eurasianbadger: this is noteworthy since badgers suffer highmortality from traffic (Clarke et al., 1998).Over-and underpasses had high use indices and high

crossing frequencies for a wide range of species, includingthe red fox and lagomorphs. In general terms, multi-purpose crossing structures are clearly important forenhancing the permeability/crossability of roads and rail-ways (Camby and Maizeret, 1987; Yanes et al., 1995; Ng etal., 2004). However, it is important to note that crossingstructures of mixed human–faunal use have been demon-strated to be effective for fauna where there is little humanactivity (Rodrıguez et al., 1996; Clevenger and Waltho,2005). In our case, the maximum value for human use (thatof overpasses) equals 1.2 pedestrians, vehicles or herds aday and the use of multipurpose crossing structures bywildlife could be radically different in areas with a higherintensity of human use.The crossing structure design is shown to be a key factor

for its use or selection by some species, thus determining itseffectiveness for barrier effect mitigation (Ballon, 1985;Forman et al., 2003). This could justify the investment onwildlife crossing structures provided that target species relymostly or exclusively on them. In our study, the differencein wildlife use of the various crossing structure types wasgreatest between culverts and wildlife overpasses, butdifferences were notable in the majority of comparisons.In simple terms, the width of the crossing structure was animportant feature in their selection by wildlife, as revealedby lagomorphs, wild boar, deer, red foxes and other canids.

ARTICLE IN PRESSC. Mata et al. / Journal of Environmental Management 88 (2008) 407–415414

The relevance of passage width has been highlighted inother studies (Foster and Humphrey, 1995; Yanes et al.,1995; Rosell et al., 1997; Veenbaas and Brandjes, 1999),but our results also show that the position of the passagewith reference to the road and its design for wildlife are keyfactors.

The use of crossing structures by rabbits and Iberianhares was related to passage design, width and by theposition of the passage with respect to the road. Thus,overpasses were more frequently used than underpassesand no rabbit or hare crossed through a culvert, inagreement with the results presented by Rosell and Velasco(1999). In addition, the frequency of use of passages builtfor wildlife was between 1.5 and 3 times greater than thatof non-wildlife engineered passages in the same positionrelative to the road. This tendency was already present inthe previously published results (Mata et al., 2005), butdifferences were not as clear-cut.

The results obtained for wild boar and for ungulates alsosupport the importance of building large wildlife passages.It is worth noting that the crossing frequencies of wild boarwere higher in this study than in any other to date(Rodrıguez et al., 1996; Rosell et al., 1997; Mata et al.,2005). The width and position of the crossing structures inrelation to the road determined their selection by wild boarin a similar fashion to that discussed above for lagomorphs(see also Vassant et al., 1993). The use of wildlifeoverpasses by wild boar was three times greater than thatfor overpasses with mixed use, and the boars also usedwildlife underpasses but not open span underpasses. Theseresults contrast with those of Rosell and Velasco (1999) forroads in Catalonia (NE Spain), where the species has beenobserved using wide underpasses of mixed use. Althoughwith fewer observations, data available for red and roe deerpoint in the same direction. Roe deer used only wildlifeunderpasses and had not previously been detected crossingthis motorway. During this study, red deer crossed mainlythrough wildlife underpasses, and previous data showedthe relevance of wildlife overpasses for them (Mata et al.,2005).

The results for the remaining species also showed atendency to use wider crossing structures, although thisfinding was less clear-cut. The high frequency of use by redfoxes of these passage types was notable, with an intensiveuse of both over- and underpasses (see also Trewhella andHarris, 1990; Rodrıguez et al., 1997). Tendencies shown byother canids are difficult to interpret due to the presence ofdomestic and feral animals as well as wild individuals.Footprints do not allow the differentiation of wolves butphotographic monitoring of passages carried out in thearea after this study (Mata et al., 2005; C. Mata,unpublished data) showed them using mainly wildlifeoverpasses. In fact, the frequency of wildlife overpass useis 30–40 times higher than that of non-wildlife engineeredover- and underpasses. However, since photographicsystems show that domestic and feral dogs are at least sixtimes more frequent on passages of the area than wolves,

the bulk of results from other canids presumably corre-spond to dogs crossing through most structure types.

5. Conclusions

Three main conclusions arise from our study: (i) mostterrestrial vertebrate species use the different types ofcrossing structures present on motorways, with more orless frequency, and their correct design and maintenancecould thus mitigate the barrier effect. However, the widthand specific design of passages play a key role fordetermining their use by at least some species of concern,such as ungulates. Therefore, (ii) special attention shouldbe paid to culvert adaptation and/or enlargement due totheir significance for certain species (e.g. badgers) andrelatively low cost, and (iii) the high construction costs ofspecific wildlife passages is justified in environmental terms.Moreover, it is important to note that wildlife passages canbe kept free of human activity while the long-termeffectiveness of mixed-use over- and underpasses can bethreatened by an intensification of their use by humans.

Acknowledgments

This study was funded by a research agreement betweenthe UAM, the Ministerio de Medio Ambiente and theCentro de Estudios y Experimentacion (CEDEX), and byan FPI grant from the Comunidad de Madrid to CristinaMata. Thanks are due to Juan Manuel Varela and JavierCachon from CEDEX and to Antonio and Teresa fromHostal Las Ventas for their help at different stages of theproject. Comments and suggestions by R. Goldingay andone anonymous referee improved the final text.

References

Ballon, P., 1985. Bilan technique des amenagements realises en France

pour reduire les impacts des grandes infrastructures lineaires sur les

ongules gibiers. In: 17e Congres de l’Union Internationale des

Biologistes du Gibier. Brussels, pp. 679–689.

Bang, P., Dahlstrom, P., 1997. Huellas y senales de los animales de

Europa. Ediciones Omega, S.A., Barcelona.

Beier, P., Loe, S., 1992. A checklist for evaluating impacts to wildlife

movement corridors. Wildlife Society Bulletin 20, 434–440.

Bekker, H.G.J., Canters, K., 1997. The continuing story of badgers and

their tunnels. In: Canters, K. (Ed.), Habitat Fragmentation and

Infrastructure. Maastricht, The Hague, Ministry of Transport, Public

Works and Water Management, Delft, The Netherlands, pp. 344–353.

Blanco, J.C., 1998. Mamıferos de Espana.

Camby, A., Maizeret, C., 1987. Permeabilite des routes et autoroutes vis-

a-vis des mammiferes carnivores: exemple des etudes menees dans les

landes de Gascogne. In: Bernard, J.M., Lansiart, M., Tille, M. (Eds.),

Routes Et Faune Sauvage. SETRA, Strasbourg, pp. 173–181.

Clarke, G.P., White, P.C.L., Harris, S., 1998. Effects of roads on badger

Meles meles populations in south-west England. Biological Conserva-

tion 86, 117–124.

Clergeau, P., 1993. Utilisation des concepts de l’ecologie du paysage pour

l’elaboration d’un nouveau type de passage a faune. Revue Gibier et

Faune Sauvage 10, 47–57.

ARTICLE IN PRESSC. Mata et al. / Journal of Environmental Management 88 (2008) 407–415 415

Clevenger, A.P., Waltho, N., 2005. Performance indices to identify

attributes of highway crossing structures facilitating movement of

large mammals. Biological Conservation 121, 453–464.

Clevenger, A.P., Chruszcz, B., Gunson, K., 2001. Drainage culverts as

habitat linkages and factors affecting passage by mammals. Journal of

Applied Ecology 38, 1340–1349.

De Santo, R.S., Smith, D.G., 1993. Environmental auditing: an

lntroduction to issues of habitat fragmentation relative to transporta-

tion corridors with special reference to high-speed rail (HSR).

Environmental Management 17, 111–114.

Dodd, C.K., Barichivich, W.J., Smith, L.L., 2004. Effectiveness of a

barrier wall and culverts in reducing wildlife mortality on a heavily

traveled highway in Florida. Biological Conservation 118, 619–631.

Fahrig, L., 2003. Effects of habitat fragmentation on biodiversity. Annual

Review of Ecology and Systematics 34, 487–515.

Forman, R., Sperling, D., Bissonette, J.A., Clevenger, A.P., Cutshall,

C.D., Dale, V.H., Fahrig, L., France, R., Goldman, C.R., Heanue, K.,

2003. Road Ecology: Science and Solutions. Island Press, Washington,

DC, USA, 481pp.

Foster, M.L., Humphrey, S.R., 1995. Use of highway underpasses by

Florida panthers and other wildlife. Wildlife Society Bulletin 23,

95–100.

Gloyne, C.C., Clevenger, A.P., 2001. Cougar Puma concolor use of wildlife

crossing structures on the Trans-Canada highway in Banff national

park, Alberta. Wildlife Biology 7, 117–124.

Huijser, M.P., Van Der Grift, E.A., Bekker, G.J., 1999. Habitat

fragmentation and infrastructure: a review of the proceedings of a

European and North American conference. Lutra 41, 43–54.

Hunt, A., Dickens, H.J., Whelan, R.J., 1987. Movements of mammals

through tunnels under railway lines. Australian Zoologist 24, 89–93.

Lode, T., 2000. Effect of a motorway on mortality and isolation of wildlife

populations. AMBIO 29, 163–166.

Mader, H.J., 1984. Animal habitat isolation by roads and agricultural

fields. Biological Conservation 29, 81–96.

Malo, J.E., Hervas, I., Mata, C., Herranz, J., Suarez, F., 2006. How many

days to monitor a wildlife passage? Species detection patterns and the

estimation of vertebrate fauna using crossing structures in a motorway.

In: Leroy, C., Garret, P., McDermott, K.P. (Eds.), On the Road to

Stewardship Proceedings of the International Conference on Ecology

& Transportation (ICOET-05). Center for Transportation and the

Environment, NC State University, Raleigh, NC, USA, pp. 406–413.

Mata, C., Hervas, I., Herranz, J., Suarez, F., Malo, J.E., 2005.

Complementary use by vertebrates of crossing structures along a

fenced Spanish motorway. Biological Conservation 124, 397–405.

Ng, S.J., Dole, J.W., Sauvajot, R.M., Riley, S.P.D., Valone, T.J., 2004.

Use of highway undercrossings by wildlife in Southern California.

Biological Conservation 115, 499–507.

Olbrich, P., 1984. Study of the effectiveness of game warning reflectors

and the suitability of game passages. Zeitschrift fur Jagdwissenschaft

30, 101–116.

Oxley, D.J., Fenton, M.B., Carmody, G.R., 1974. The effects of roads on

populations of small mammals. Journal of Applied Ecology 11, 51–59.

Palomo, L.J., Gisbert, J., 2002. Atlas de los mamıferos terrestres de

Espana. Direccion General de Conservacion de la Naturaleza-

SECEM-SECEMU, Madrid.

Puky, M., Vogel, Z., 2003. Amphibian Mitigation Measures on Hungarian

Roads: Design, Efficiency, Problems and Possible Improvement, Need

for a Co-ordinated European Environmental Education Strategy.

Habitat Fragmentation due to Transportation Infrastrucutre. IENE,

Brussels.

Putman, R.J., 1997. Deer and road traffic accidents: options for

management. Journal of Environmental Management 51, 43–57.

Quinn, G.P., Keough, M.J., 2002. Experimental Design and Data

Analysis for Biologists. Cambridge University Press, Cambridge, UK.

Reed, D.F., 1981. Mule deer behavior at a highway underpass exit.

Journal of Wildlife Management 45, 542–543.

Rice, W.R., 1989. Analyzing tables of statistical tests. Evolution 43,

223–225.

Rodrıguez, A., Crema, G., Delibes, M., 1996. Use of non-wildlife passages

across a high speed railway by terrestrial vertebrates. Journal of

Applied Ecology 33, 1527–1540.

Rodrıguez, A., Crema, G., Delibes, M., 1997. Factors affecting crossing of

red foxes and wildcats through non-wildlife passages across a high-

speed railway. Ecography 20, 287–294.

Rosell, C., Velasco, J.M., 1999. Manual de prevencio i correccio dels

impactes de les infraestructuras viaries sobre la fauna. Departament de

Medi Ambient, Generalitat de Catalunya, Barcelona.

Rosell, C., Parpal, J., Campeny, R., Jove, S., Pasquina, A., Velasco, J.M.,

1997. Mitigation of barrier effect of linear infrastructures to wildlife.

In: Canters, K., (Ed.), Habitat Fragmentation and Infrastructure.

Maastricht, The Hague, Ministry of Transport, Public Works and

Water Management, Delft, The Netherlands, pp. 367–372.

Rosell, C., Alvarez, G., Cahil, S., Campeny, R., Rodrıguez, A., Seiler, A.,

2003. COST 341. La fragmentacion del habitat en relacion con las

infraestructuras de transporte en Espana. O.A. Parques Nacionales,

Ministerio de Medio Ambiente, Madrid.

Saunders, D.A., Hobbs, R.J. (Eds.), 1991. Nature Conservation 2: The

Role of the Corridors. Surrey, Beatty, Chipping Norton, New South

Wales, Australia.

Singer, F.J., Doherty, J.L., 1985. Managing mountain goats at a highway

crossing. Wildlife Society Bulletin 13, 469–477.

StatSoft, Inc. 1998. STATISTICA, (data analysis software) version 6.0.

Tulsa, OK, USA.

Strachan, R., 1995. Mammal Detective. Whittet Books Limited, London.

Taylor, B.D.T., Goldingay, R.L., 2003. Cutting the carnage: wildlife usage

of road culverts in north-eastern new south Wales. Wildlife Research

30, 529–537.

Trewhella, W.J., Harris, S., 1990. The effect of railway lines on urban fox

(Vulpes vulpes) numbers and dispersal movements. Journal of Zoology

221, 321–326.

Van Bohemen, H.D., 1998. Habitat fragmentation, infrastructure and

ecological engineering. Ecological Engineering 11, 199–207.

Vassant, J., Brandt, S., Jullien, J.M., 1993. Effects of the wildlife passage

over motorway A5 on the populations of red deer and wild boar in the

Massif d’Arc-en-Barrois. part 2: case of the population of wild boar.

Bulletin Mensuel de l’Office National de la Chasse 184, 24–33.

Veenbaas, G., Brandjes, G.J., 1999. The use of fauna passages along

waterways under motorways. ICOWET III. The International Con-

ference on Wildlife Ecology and Transportation, Florida, 7pp.

Walker, R.S., Novaro, A.J., Branch, L.C., 2003. Effects of patch

attributes, barriers, and distance between patches on the distribution

of a rock-dwelling rodent (Lagidium viscacia). Landscape Ecology 18,

187–194.

Yanes, M., Velasco, J.M., Suarez, F., 1995. Permeability of roads and

railways to vertebrates: the importance of culverts. Biological

Conservation 71, 217–222.