A large-scale assessment of European rabbit damage to agriculture in Spain

Depósi o de in es igación de la Uni e sidad de Se illa h ps://idus.us.es/ “This is he pee e iewed e sion o he ollowing a icle: Delibes-Ma eos, M., Fa án, M.Á., Rouco, C., Oli e o, J., Má quez, A.L., Fa, J.E., Va gas, J.M. and Villa ue e, R. (2018), A la ge-scale assessmen o Eu opean abbi damage o ag icul u e in Spain. Pes . Manag. Sci, 74: 111-119, which has been published in inal o m a h ps://doi.o g/10.1002/ps.4658 . This a icle may be used o non-comme cial pu poses in acco dance wi h Wiley Te ms and Condi ions o Use o Sel - A chi ed Ve sions. This a icle may no be enhanced, en iched o o he wise ans o med in o a de i a i e wo k, wi hou exp ess pe mission om Wiley o by s a u o y igh s unde applicable legisla ion. Copy igh no ices mus no be emo ed, obscu ed o modi ied. The a icle mus be linked o Wiley’s e sion o eco d on Wiley Online Lib a y and any embedding, aming o o he wise making a ailable he a icle o pages he eo by hi d pa ies om pla o ms, se ices and websi es o he han Wiley Online Lib a y mus be p ohibi ed." Accep ed A icle A la ge-scale assessmen o Eu opean abbi damage o ag icul u e in Spain Running i le: Ag icul u al damage by abbi s in Spain Miguel Delibes-Ma eos1,2*, Miguel Ángel Fa án3,4, Ca los Rouco5, Jesús Oli e o3, Ana Luz Má quez3, John E. Fa6, Juan Ma io Va gas3, Ra ael Villa ue e2 1 Depa amen o de Biología Vege al y Ecología, Facul ad de Biología, Uni e sidad de Se illa, Apa ado 1095, 41080, Se illa, Spain 2 Ins i u o de Es udios Sociales A anzados (CSIC), Campo San o de los Má i es 7, 14004, Có doba, Spain 3 Depa amen o de Biología Animal, Facul ad de Ciencias, Uni e sidad de Málaga, Campus de Tea inos, 29071, Málaga, Spain 4 Biogea Consul o es, Calle Na a o Ledesma 243, 29010, Málaga, Spain 5 Depa amen o de Zoología, Campus de Rabanales, Uni e sidad de Có doba, 14071, Có doba, Spain 6 Di ision o Biology and Conse a ion Ecology, School o Science and he En i onmen , Manches e Me opoli an Uni e si y, Manches e M1 5GD, UK E-mail add esses: mdelibesma [email protected] (Miguel Delibes-Ma eos); [email p o ec ed]om (M.A. Fa án); [email p o ec ed] (C. Rouco); jesusoli e [email protected] (J. Oli e o); [email p o ec ed] (A.L. Má quez); j [email p o ec ed] (J.E. Fa); [email p o ec ed] (J.M. Va gas); a ael. illa ue [email protected] (R. Villa ue e) This a icle is p o ec ed by copy igh . All igh s ese ed. This a icle has been accep ed o publica ion and unde gone ull pee e iew bu has no been h ough he copyedi ing, ypese ing, pagina ion and p oo eading p ocess, which may lead o di e ences be ween his e sion and he Ve sion o Reco d. Please ci e his a icle as doi: 10.1002/ps.4658 Accep ed A icle *Co esponding au ho : Miguel Delibes-Ma eos. Depa amen o de Biología Vege al y Ecología, Facul ad de Biología, Uni e sidad de Se illa, Apa ado 1095, 41080, Se illa, Spain. Phone numbe : 0034 954557070. E-mail add ess: [email p o ec ed] This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle Abs ac BACKGROUND: Nume ous small and medium-sized mammal pes s cause widesp ead and economically signi ican damage o c ops all o e he globe. Howe e , mos esea ch on pes species has ocused on accoun s o he le el o damage. The e a e ewe s udies concen a ing on he desc ip ion o c op damage caused by pes s a la ge geog aphical scales, o on analysing he ecological and an h opogenic ac o s co ela ed wi h hese obse ed pa e ns. We in es iga ed he ela ionship be ween ag icul u al damage by he Eu opean abbi (O yc olagus cuniculus) and en i onmen al and an h opogenic a iables h oughou Spain. RESULTS: Rabbi damage was mainly concen a ed wi hin he cen al-sou he n egions o Spain. We ound ha abbi damage has inc eased signi ican ly be ween he ea ly 2000s and 2013. G ea e losses we e ypical o hose a eas whe e a ming domina ed and na u al ege a ion was sca ce, whe e main ailways and highways we e p esen , and whe e en i onmen al condi ions we e gene ally a ou able o abbi popula ions o p oli e a e. CONCLUSIONS: F om ou analysis we sugges ha oads and ailway lines ac as po en ial co ido s along which abbi s can sp ead. The ecen inc ease in Spain o such in as uc u e may explain he ise in abbi damage epo ed in his s udy. Ou app oach is aluable as a me hod o assessing d i e s o wildli e pes damage a la ge spa ial scales, and can be used o p opose me hods o educe human-wildli e con lic . Keywo ds: Google sea ch; Human-wildli e con lic ; Landscape change; O yc olagus cuniculus; Pes species; Small mammals This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle 1. INTRODUCTION Nega i e impac s o wildli e on humans o hei esou ces ha e occu ed h oughou p ehis o y and eco ded his o y. Consequences o human-wildli e con lic can be bo h di ec , including disease, inju y and dea h om encoun e s wi h dange ous animals, and indi ec , such as loss o c ops and li es ock and damaged in as uc u e. Measu es o p e en ing o educing human-wildli e con lic a e inc easingly being de eloped.1 Success ul con lic esolu ion plans a e o en he esul o he assimila ion o p o en me hods o con ol o de e ence and hei clea adap a ion o he na u e o he p oblem a hand.2 The la e equi es an accu a e assessmen o he human-wildli e con lic si ua ion in es iga ed, and a clea unde s anding o he e icacy, cos -e ec i eness and social accep abili y o he me hods o be applied. Al hough human-animal con lic in ol ing la ge mammals is o en highligh ed in he media, human li elihoods a e a guably much mo e impac ed by small and medium-sized mammals.3,4 These species o en cause subs an ial damage o pas u es as well as g ound and ee c ops h ough g azing o b owsing. Fo example, ou b eaks o oden s in apple g owing a eas in Ge many can cause up o 25 million € pe yea o damage.4 Close g azing by la ge numbe s o he Eu opean abbi (O yc olagus cuniculus) can weaken o kill e en pe sis en lea y c ops, and sc a ching and bu owing can deg ade pas u e s ill u he by encou aging he es ablishmen o weeds.5 In he UK, abbi damage was eco ded as being € 115 - € 150 million in 1986,6 and a ound € 130 million pe yea in Aus alia.7 Coun y-wide assessmen s o he damage caused by abbi s in he UK and Aus alia8,9 o coypu (Myocas o coypus) in I aly10 ha e been unde aken. Howe e , analyses o he spa ial (whe e damage happens) and empo al (when damage occu s) ac o s de e mining he c op damage in lic ed by a pa icula pes species ha e been la gely missing om la ge-scale assessmen s. Such app oaches allow o he applica ion o unde s anding wha a iables may This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle co ela e wi h a species’ p opensi y o cause damage and hus can pe mi manage s o de ise mo e e ec i e measu es o con ol especially since wha may be success ul a a smalle demons a ion scale can ail when applied a he b oade landscape le el.11 Pes ou b eaks can be linked o speci ic en i onmen al condi ions. In he case o he Douglas i ussock mo h (O gyia pseudo suga a) e up ions co ela ed p ima ily wi h o es ype and clima e.12 Howe e , o some species, pa icula ly insec pes s, anspo in as uc u e ( oads and ailway lines) can also pe mi hei dispe sal and es ablishmen .13 The e a e ew s udies on he ole o oads and ailways as po en ial co ido s along which pes species may dispe se. An excep ion is a s udy o he b own ma mo a ed s ink bug (Halyomo pha halys) in he US, which showed ha u ban de elopmen and ail oads acili a ed he species’ sp ead.14 Gi en he ubiqui y o anspo in as uc u e in many pa s o he wo ld, i is likely ha pes species will each emo e a eas as hey es ablish hemsel es along oads o ailway lines whe e hey can become abundan , and om which hey can p oli e a e.15 Assessing c op damage caused by indigenous small mammal pes s, as well as inc easing ou unde s anding o wha de e mines his, is impo an when he species in ques ion also plays a key ole in a na i e ecosys em. This is he case o he Eu opean abbi in he Ibe ian Peninsula, i s na i e ange.16 The con ol o he abbi he e needs o be e ec ed in such a way ha does no cause i s un es ained decline o ex inc ion; his could p oduce cascading e ec s in he ecosys em.16 Rabbi s in Ibe ian Medi e anean ecosys ems ac as a mul i unc ional keys one species, no only because hey a e p ey o o e 30 p eda o s, bu also because o hei ole as an ecosys em enginee .17 By con as , abbi s cause subs an ial damage o ag icul u e in Spain and Po ugal,18,19,20 and a e ega ded as a se e e pes species in some egions. Mo eo e , abbi s a e one o he mos impo an small-game species, hun ed in hei This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle millions in he wo Ibe ian coun ies. These mul iple oles o en lead o equen con lic s be ween a me s, hun e s and conse a ionis s on how o manage he species.18 Knowing whe e and when abbi damage occu s in Spain and which ac o s may d i e i can help mi iga e con lic s be ween he a ious s akeholde s. In his s udy, we i s desc ibe abbi damage h oughou Spain, un il now un eco ded, by: 1) documen ing a eas (i.e. municipali ies) whe e abbi damage has been epo ed o occu ; 2) eco ding which c ops a e a ec ed; and 3) e alua ing he empo al e olu ion o abbi damage in he coun y. Ac oss mul iple si es we hen assess he p obabili y o occu ence o c op damage caused by abbi s, and i s ela ionship wi h local condi ions and en i onmen al ac o s, such as land use, na u al ege a ion o in as uc u e. To do his, we employ la ge-scale da abases o en i onmen al and bio ic a iables, and use spa ial modelling.21 Ou esul s a e used o de e mine which a eas a e mo e p one o abbi damage and assess which ac o s a e co ela ed wi h his. 2. MATERIALS AND METHODS 2.1 S udy a ea Peninsula Spain co e s 493,518 km2 (nea ly 85% o he Ibe ian Peninsula). The coun y is di ided in o 8,125 adminis a i e municipali ies (h p://www.ine.es) he e i o ial uni s used in he cu en analyses. Clima e a ies widely h oughou Spain, wi h p ecipi a ion dec easing eas wa ds and sou hwa ds, and empe a u e inc easing sou hwa ds.22 Moun ains bo de he no he n and Medi e anean coas s (max. al . 3,478 m) bu he e a e also eas -wes acing anges in he cen e o he Peninsula. The na u al ege a ion in he moun ains is pine and oak o es s and sc ubland, bu lowlands a e mos ly domina ed by he baceous c ops and i e e aces, and hill slopes by woody c ops such as oli e. Acco ding o he 2010 Eu opean Union Fa m S uc u e Su ey This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle A la ge-scale assessmen o Eu opean abbi damage o ag icul u e in Spain Running i le: Ag icul u al damage by abbi s in Spain Miguel Delibes-Ma eos1,2*, Miguel Ángel Fa án3,4, Ca los Rouco5, Jesús Oli e o3, Ana Luz Má quez3, John E. Fa6, Juan Ma io Va gas3, Ra ael Villa ue e2 1 Depa amen o de Biología Vege al y Ecología, Facul ad de Biología, Uni e sidad de Se illa, Apa ado 1095, 41080, Se illa, Spain 2 Ins i u o de Es udios Sociales A anzados (CSIC), Campo San o de los Má i es 7, 14004, Có doba, Spain 3 Depa amen o de Biología Animal, Facul ad de Ciencias, Uni e sidad de Málaga, Campus de Tea inos, 29071, Málaga, Spain 4 Biogea Consul o es, Calle Na a o Ledesma 243, 29010, Málaga, Spain 5 Depa amen o de Zoología, Campus de Rabanales, Uni e sidad de Có doba, 14071, Có doba, Spain 6 Di ision o Biology and Conse a ion Ecology, School o Science and he En i onmen , Manches e Me opoli an Uni e si y, Manches e M1 5GD, UK E-mail add esses: mdelibesma [email protected] (Miguel Delibes-Ma eos); [email p o ec ed]om (M.A. Fa án); [email p o ec ed] (C. Rouco); jesusoli e [email protected] (J. Oli e o); [email p o ec ed] (A.L. Má quez); j [email p o ec ed] (J.E. Fa); [email p o ec ed] (J.M. Va gas); a ael. illa ue [email protected] (R. Villa ue e) This a icle is p o ec ed by copy igh . All igh s ese ed. This a icle has been accep ed o publica ion and unde gone ull pee e iew bu has no been h ough he copyedi ing, ypese ing, pagina ion and p oo eading p ocess, which may lead o di e ences be ween his e sion and he Ve sion o Reco d. Please ci e his a icle as doi: 10.1002/ps.4658 Accep ed A icle *Co esponding au ho : Miguel Delibes-Ma eos. Depa amen o de Biología Vege al y Ecología, Facul ad de Biología, Uni e sidad de Se illa, Apa ado 1095, 41080, Se illa, Spain. Phone numbe : 0034 954557070. E-mail add ess: [email p o ec ed] This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle O e all, ou model o Eu opean abbi damage used a o al o 136 p edic o a iables (Table S2). The a ou abili y model o he p esence o Eu opean abbi s included 79 ecosys em ype desc ip o s based on na u al ege a ion s uc u e, and 56 indica o s o an h opogenic ac i i y namely anspo in as uc u e (i.e. main ailways and highways), ag icul u e and li es ock. Spa ial analyses o calcula e he p esence o anspo in as uc u e we e pe o med using A cGIS 10.0 (ESRI, Redlands, Cali o nia, USA). We excluded nonlinea and in e ac ion e ec s om he model o keep i s ma hema ical o mula ion as simple as possible o explana o y pu poses. To accoun o Type-I e o s caused by he la ge numbe o a iables conside ed, we con olled he False Disco e y Ra e (FDR).29 Thus, using he p esence/absence o Eu opean abbi damage as he dependen a iable, we an a logis ic eg ession on each o he 136 p edic o a iables, and only signi ican (P< 0.05) a iables unde an FDR o q < 0.05 we e accep ed as pa o a mul i a ia e en i onmen al model. Mul icollinea i y was also minimised by excluding a iables ha we e co ela ed wi h Spea man > 0.8. Only hen did we pe o m he a ou abili y model using he ollowing equa ion26,27: F = [P/(1-P)]/[(n1/n0)+(P/[1-P])] whe e F is he a ou abili y alue in each municipali y; n1 and n0 he numbe o municipali ies wi h and wi hou damage epo ed in websi es, espec i ely; and P he p obabili y o abbi damage p esence in each municipali y. P was calcula ed wi h a mul iple logis ic eg ession, using he p esence and absence o damage in each municipali y acco ding o he analysed websi es as he dependen a iable. A o wa d s epwise p ocedu e was pe o med o a oid he inclusion o edundan a iables in a model. The pa ame e s in he This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle models we e i ed by i e a i ely maximizing he log-likelihood. (using IBM SPSS s a is ics 22). 2.4 Explana o y analysis o a iables in he model We employed a a ia ion pa i ioning p ocedu e o measu e he ela i e pa icipa ion o he ou main ac o s desc ibed in Table S2 ( a ou abili y o he p esence o Eu opean abbi , anspo in as uc u e, na u al ege a ion and land uses) on he model explana ion o a ou abili y o damage occu ence.30 In his way, we speci ied how much o he a ia ion in a ou abili y was accoun ed o by he pu e e ec o each ac o (i.e., a ia ion ha is no a ec ed by co a ia ion wi h ano he ac o ), and wha p opo ion was clea ly a ibu able o mo e han one ac o (i.e. sha ed e ec ). The signi icance o he in luence o all a iables in he model was assessed using he uni a ia e Wald es s a is ic.31 S epwise me hods selec a iables ac ing on a la ge scale in he i s s eps, and a subsequen s eps only add a iables ha a e signi ican ly ela ed o he esiduals no accoun ed o by p e iously inco po a ed a iables.32 The egional ele ance o e e y a iable was, hus, analysed using wo app oaches. Fi s ly, we measu ed he co ela ion (Spea man R) o each a iable wi h he a ou abili y ou pu , and compa ed he sign o R (which indica es global ela ionship wi hin he s udy a ea) wi h he sign o he a iable coe icien in he model equa ion (which indica es he sign o he a iable con ibu ion o explaining a ou abili y). Secondly, we isualised he egional con ibu ion o each a iable o he model by mapping he di e ence be ween a ou abili y alues ob ained in successi e s eps, along he s epwise a iable selec ion. 3. RESULTS 3.1 Tempo al and spa ial cha ac e isa ion o damage caused by abbi s This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle The gene al Google sea ch yielded a o al o 306 unique, alid websi es con aining in o ma ion on damage caused by Eu opean abbi s in Spain. The i s cases o abbi damage we e epo ed in 2004. The e we e ew cases du ing 2004-2006, bu hese inc eased sha ply un il 2013, a e which he e was a sha p decline (Fig. 1). In 2014, we eco ded he highes numbe o websi es con aining in o ma ion on abbi damage, eaching le els highe han in 2012. The e was a s ong signi ican co ela ion (R= 0.971, p<0.001) be ween he numbe o websi es con aining in o ma ion on abbi damage and yea when we s anda dised ou esul s using he Google sea ch by da e op ion. Nea ly all analysed websi es epo ed c op damage caused by abbi s (95.7%; n=301), wi h only a ew documen ing damage o in as uc u e a ound ailways o highways (10.3%; n=301), o human p ope ies such as ga dens o pa ks (9.6%; n=301). Mos abbi damage was epo ed o ineya ds and ce eal c ops (66.1% and 53.3%, espec i ely; n=180), al hough i was also ela i ely equen in oli e g o es and ui o cha ds (28.3% and 21.7%, espec i ely; n=180). G een ege ables and o he c ops we e less equen ly men ioned in he websi es (15.0% and 9.4%, espec i ely; n=180). Rabbi damage was epo ed in 437 (5.5%) municipali ies ac oss mainland Spain. These we e mos ly ound in: 1) Guadalqui i and Eb o alleys in sou he n and no h-eas e n Spain, espec i ely; 2) Cas illa-La Mancha and Cas illa y León pla eaus in cen al Spain; and 3) he coas al eas e n egions, pa icula ly he Comunidad Valenciana egion (Fig. 2). 3.2 Modelling a ou abili y o Eu opean abbi damage The a ou abili y model o he p esence o Eu opean abbi damage included 18 di e en a iables ela ed o anspo in as uc u e, land use, na u al ege a ion and en i onmen al a iables (Table 1). The model showed ha municipali ies a ou ing he p esence o Eu opean abbi damage ended o be ound nea each o he (Fig. 3), loca ed This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle mainly along he Guadalqui i and Eb o alleys, se e al a eas o he Guadiana alley, No h and Sou h pla eaus and some coas al a eas o Sou heas peninsula Spain. Va iables wi h he highes explana o y powe wi hin he model (Wald s a is ic > 25) we e p ima ily anspo in as uc u e and land use desc ip o s (Table 1). Humanised habi a s, such as c oplands and in as uc u e, and en i onmen ally a ou able a eas o he p esence o Eu opean abbi s we e posi i ely ela ed o damage. In con as , na u al ege a ion was nega i ely associa ed wi h abbi damage. Mo e speci ically, highways and main ailways, d y and i iga ed oli e c ops, and d y and i iga ed ineya ds posi i ely co ela ed wi h he occu ence o Eu opean abbi damage, i.e. hey showed iden ical signs wi hin he model equa ion and in he a iable co ela ions wi h he model, and hei en y in o he model p oduced an inc ease in a ou abili y (Fig. 4). On he o he hand, o es s (specially ma u e oak o es s), sc ublands and pas u es ep esen ed en i onmen al cons ain s o he occu ence o Eu opean abbi damage, i.e. bo h signs wi hin he model equa ion and a iable co ela ions wi hin he model we e nega i e, and hei en y in o he model p oduced a dec ease in a ou abili y (Fig. 4). The g ea e p opo ion o he spa ial a ia ion in a ou abili y was explained by habi a desc ip o s, ep esen ed by na u al ege a ion and land use ac o s. The sum o he pu e and sha ed e ec s o hese ac o s on a ou abili y a ia ion was 81.8% and 53.7%, espec i ely (Fig. 5). T anspo in as uc u e and en i onmen al a ou abili y o he p esence o Eu opean abbi s only explained 8.4% and 10.6% o a ia ion, espec i ely. We ound sha ed e ec s be ween he di e en ac o s (meaning ei he c oss o indis inguishable explana o y powe ). Land uses and in as uc u e sha ed 90.3% o hei in luence, and land uses and na u al ege a ion sha ed 40.8%. The o he sha ed e ec s be ween ac o s we e conside ably smalle . As a esul , he pu e e ec o na u al ege a ion explained 39% o a ou abili y o he p esence o damage, whils he pu e e ec o land uses, en i onmen al This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle a ou abili y o he p esence o Eu opean abbi s, and in as uc u e explained only 10.0%, 3.0% and 2.3%, espec i ely. 4. DISCUSSION We show ha abbi damage is mo e common in he cen al-sou he n egions o Spain, and in some no he n a eas (Fig. 2). We also demons a e ha he likelihood o damage by abbi s in an a ea depends on a combina ion o ac o s, which ou model has clea ly iden i ied (Fig. 5). Mos abbi damage occu s in municipali ies whe e a ming domina es (and he e o e na u al ege a ion is sca ce), whe e main ailways and highways a e p esen and whe e en i onmen al condi ions o abbi popula ions o succeed a e gene ally good (Table S2). Using ou app oach we we e able o highligh hose a eas po en ially a g ea e isk o abbi damage a a la ge na ional scale, which a e likely o be missed using o he means o e i ica ion. Fo example, ou model clea ly showed ha a eas such as Cas illa y León egion (no h-wes e n Spain) ha e ela i ely high a ou abili y alues e en hough he numbe o cases ound in ou Google sea ch was low (Figs. 2 & 3). Ou s udy demons a es ha he modelling app oach we use can be use ul in de e mining he ole o mul iple ac o s a la ge spa ial scales o explaining he damage caused by a pes species. In addi ion, we show ha online media epo s on pes damage can be an e ec i e and ela i ely apid op ion o unde s and he dis ibu ion o pes damage in a landscape as la ge as an en i e coun y. Howe e , a po en ial d awback o his in o ma ion is ha epo s in he media on pes damage may only eco d he s akeholde s’ pe cep ions, bu pe cei ed and ac ual damage caused by wildli e may no be co ela ed.33 Despi e his, media epo s ha e been success ully used p e iously o assess he impac o small mammal pes s.34 Thus, we a gue ha , e en i he e is some e o in ol ed, he in o ma ion ob ained in ou s udy can help iden i y a eas o po en ial s akeholde con lic ega ding he managemen o This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle pes species ( a me s and hun e s in ou s udy model),18,24 and can be used in he design o mo e ealis ic managemen plans. Acco ding o he analysed websi es in Spain, mos abbi damage occu ed in c ops. This is in line wi h he esul o ou model in which a ou able a eas o abbi damage we e posi i ely co ela ed wi h se e al land uses ela ed o ag icul u e; many di e en c op ypes we e among he mos impo an p edic o s o abbi damage (Table 1). Ou esul s sugges ha , in Spain, he abbi is a gene alis ag icul u al pes since a a ie y o c ops a e damaged by his species (see also 20). This con as s wi h o he Spanish small mammal pes s ha a e linked o speci ic c op ypes; e.g. oles and he baceous c ops.35 F om he analysed websi es, we showed ha ineya ds we e he mos equen ly eco ded c op ype a ec ed by abbi s, and bo h hese and oli e g o es had signi ican e ec s on abbi damage a ou abili y in he model (Table 1). No able educ ions in woody c op yields om abbi b owsing ha e been epo ed in some Spanish ag icul u al egions.21 Gi en ha a as p opo ion o ag icul u al land in some egions o Spain is de o ed o hese woody c ops – e.g. >98% o he a ea in Mon illa-Mo iles Wineg owing Region in sou he n Spain36 – losses caused by abbi s o oli e oil and wine p oduce s can be economically c ippling. Na u al ege a ion explained a la ge p opo ion o a ou abili y o abbi damage in ou model (Table 1). The ac ha mos na u al ege a ion ypes we e nega i ely associa ed wi h he po en ial occu ence o abbi damage is no su p ising, because he highe he p opo ion o land occupied by na u al ege a ion he lowe he a ailabili y o c ops o be g azed by abbi s. Fu he mo e, abbi s a e ound in low numbe s in a eas domina ed by o es s o dense sc ublands,37 hus dec easing he likelihood o damage o adjacen c ops. Simila ly, we obse ed ha abbi damage is mos likely in a eas en i onmen ally a ou able o he species (Table 1). In hese a eas high abbi densi ies a e possible in addi ion o being posi i ely associa ed wi h he p esence o woody c ops,38 ypically damaged by abbi s. This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle Al hough he numbe o hun e s in Spain has emained ela i ely cons an (h p://www.mag ama.gob.es), abbi s hun ed ha e inc eased alongside he sp ead o as speed ailways du ing he las decade (Fig.1, R=0.904, p<0,001). Acco ding o ou model, one o he main p edic o s o abbi damage was he p esence o ailways o mo o ways (Table 1; Fig. 4). The explana ion is ha abbi s can become mo e abundan close o hese in as uc u e,39,40 pa icula ly in a eas wi hin a mland egions whe e he a ailabili y o o he sui able habi a o hei success ul es ablishmen is sca ce.41 Ve ges and embankmen s along oads o ailways ( o as -speed ailways hese a e o en chain-mesh secu i y enced) p o ide abbi s wi h e uge agains p eda o s and hun e s, as p eda o abundance is lowe and hun ing o en banned in hese a eas.39,41 In addi ion, hese e ges and embankmen s o e ideal condi ions o abbi wa en building,40 essen ial o es ablish heal hy social g oups.42,43 Simila easons ha e been p oposed o explain he high numbe s o some small mammals ound in a eas wi h high oad densi y44 o he in ensi e use o a eas adjacen o oads by ju eniles o wo lagomo ph species.45 In a mland a eas, abbi s usually es ablish dense popula ions along he e ges o oads and ailway lines, whe e nea by c ops a e hei only eeding sou ce since na u al ege a ion has almos disappea ed as a consequence o in ensi e ag icul u al p ac ices.36 This may also explain he high impo ance o human in as uc u e explaining abbi damage obse ed in he model. Ou Google sea ches unco e ed e y ew si es con aining in o ma ion ela ed o abbi pes s in Spain du ing he i s hal o he 2000s. This was a ime when Spanish abbi popula ions we e a hei lowes numbe s in eco ded his o y because mos o hem had no eco e ed om he huge impac o abbi haemo hagic disease (RHD), i s de ec ed in Spain in 1988.46 The inc ease in he numbe o websi es epo ing damage be ween 2006 and 2012 ma ches he known eco e y o some popula ions du ing ha pe iod,47,48 pa icula ly in some ag icul u al a eas.18 I also ma ches he inc ease in he numbe o abbi con ol eques s This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle and he numbe o ha es ed abbi s eco ded in Po ugal18 and in Spain (Fig. 1). These indings closely esemble he si ua ion in Aus alia, whe e abbi numbe s we e hea ily supp essed by RHD be ween 1995 and 2002, bu inc eased 5-10 old be ween 2003 and 2010.49 Ou indings also show ha in 2013 he a ailable online in o ma ion ela ed o abbi damage ell sha ply, which seems o be only explicable by he popula ion c ash caused by a new a ian o he RHD i us.48,50,51 I is e y likely ha he decline in abbi s obse ed in ag icul u al a eas esul ed educed abbi damage, which was e lec ed in he lowe numbe o websi es add essing his issue. In e es ingly, websi es epo ing abbi damage inc eased conside ably be ween 2013 and 2014 (Fig. 1), which sugges s ha abbi popula ions migh ha e eco e ed a leas in some ag icul u al a eas, al hough in na u al a eas abbi numbe s we e s ill e y low.51 Ou indings indica e ha he e olu ion o online epo s on abbi damage migh be a good indica o o ends in abbi popula ion and hei damage a a la ge scale in Spain, al hough such da a ha e some po en ial limi a ions. Fo example, he ela ionship be ween abbi abundance and hei damage is complex, as damage o c ops by abbi s has been also obse ed in a eas wi h mode a e abbi abundance.18 On he o he hand, he conside able inc ease in he use o In e ne o e ecen yea s migh explain he g owing a ailabili y o online in o ma ion ega ding almos any opic. Howe e , he obse ed ise in abbi damage epo s when we con olled o sampling e o pe yea and he g owing numbe s o abbi s hun ed and abbi con ol pe mi s eques ed18,20 sugges ha abbi damage has in ac inc eased in ecen yea s in Spain. In any case, we do ecommend ha online da a be in e p e ed cau iously. The p esence o anspo in as uc u e was one o he main explana o y ac o s o abbi damage in ou model. Gi en he ex ensi e de elopmen o as -speed ailways and highways ha occu ed in Spain in ecen yea s, we sugges ha such landscape modi ica ion This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle may ha e been one o he main d i e s in he ise in abbi damage epo ed in ou s udy. This seems o be suppo ed by he signi ican o e lap obse ed be ween he inc ease in he numbe o websi es dealing wi h abbi pes s and inc ease in he leng h o as -speed ailways in Spain (Fig. 1; R= 0,869, p<0,001). Al e na i ely, he educ ion o o he ood sou ces o abbi s has led o inc eased c op damage.36 The ole o humans in he expansion o small mammal pes s and damage by hese has been also obse ed in o he a mland sys ems.35 Economic damage caused by mammal pes species can be high, pa icula ly in a mland a eas (e.g. 52). This may be he sou ce o con lic be ween s akeholde s whose li elihood is a ec ed by wildli e like a me s and o he pa ies ha make o he uses o hese species, such as hun e s o en i onmen alis s. Ou case s udy is a good example, bu he e a e o he s ac oss he wo ld.16,53 We sugges ha i is essen ial o assess wha d i es wildli e damage a la ge spa ial scales as well as o iden i y a eas whe e po en ial con lic s may eme ge. The app oach p esen ed in his pape can be ep oduced in o he sys ems, o no only educe wildli e pes damage, bu also o wo k agains smoo hing ou ensions be ween s akeholde s ega ding he managemen o wildli e. ACKNOWLEDGEMENTS We wish o hank 3 anonymous e iewe s o cons uc i e commen s. M. Delibes-Ma eos and C. Rouco we e suppo ed by he Talen ia Pos doc and Talen Hub P og am, espec i ely, launched by he Andalusian Knowledge Agency, and co- unded by he Eu opean Union’s Se en h F amewo k P og am, Ma ie Skłodowska-Cu ie ac ions (COFUND – G an Ag eemen s n267226 and 291780, espec i ely) and he Minis y o Economy, Inno a ion, Science and Employmen o he Jun a de Andalucía. M. Delibes-Ma eos is cu en ly unded by V Plan P opio de In es igación o he Uni e sidad de Se illa. This s udy is a con ibu ion o he p ojec s CGL2013-43197-R (I+D Na ional Plan) unded by he Spanish Minis y o This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle Science and Inno a ion and EU–FEDER unds, and 1098/2014 (O ganismo Au ónomo Pa ques Nacionales, Spain). REFERENCES 1 Wood o e R, Thi good S, Rabinowi z A, People and Wildli e, Con lic o Co-exis ence? Camb idge Uni e si y P ess, Camb idge, Uni ed Kingdom (2005). 2 Cono e MR, Resol ing human-wildli e con lic s: The science o wildli e damage managemen . CRC P ess, Boca Ra on, Uni ed S a es (2002). 3 Capizzi D, Be olino S and Mo elli i A, Ra ing he a s: global pa e n and esea ch p io i ies in impac s and managemen o oden pes s. Mammal Re iew 44:148–162 (2014). 4 Imhol C, Reil D, Plasil P, Rödinge K and Jacob J, Long- e m popula ion pa e ns o oden s and associa ed damage in Ge man o es y. Pes Managemen Science 73:332–340. (2017). 5 Mu ze G, Cooke B and Jennings S, Densi y-dependen g azing impac s o in oduced Eu opean abbi s and sympa ic kanga oos on Aus alian na i e pas u es. Biological In asions 18:2365–2376 (2016). 6 Mills S, Rabbi s b eed a g owing con o e sy. New Scien is 109:50–54 (1986). 7 Gong W, Sinden J, B ayshe M and Jones R, The economic impac s o e eb a e pes s in Aus alia. In asi e Animals Coope a i e Resea ch Cen e, Canbe a, Aus alia (2009). 8 Whi e PCL and Ha is S, Economic and en i onmen al cos s o alien e eb a e species in B i ain. In Biological in asions: economic and en i onmen al cos s o alien plan . Animal and mic obe species, ed. by Pimen el D, CRC P ess, Boca Ra on, USA, pp. 113–149 (2002). This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle Va iables desc ibing a ou able a eas o wild abbi damages DRY_OLIV 3 124.834 + + HIGHW_RAILW 1 119.444 + + IRR_OLIV 7 51.344 + + DRY_VIN 8 43.624 + + IRR_VIN 4 26.387 + + F_WILDRABBIT 11 15.961 + + ORG_CER 10 13.372 + + WOLIV_COP 9 9.238 + + WOLIV_FOR 13 7.006 + + P_UAA 14 4.157 + + NOAK_CLEFOR 6 9.686 + - SW Va iables appa en ly ou lining un a ou able a eas o wild abbi damages DEV_OAK 5 32.730 - - PAST 12 7.016 - - DIV_FOR 2 6.276 - - PINE_FOR 17 5.619 - - GALL_FOR 16 5.294 - - OAK_CLEFOR 18 4.075 - - PAST_SHRUB 15 9.360 - + SE, E, NE FIGURE LEGENDS Figu e 1. E olu ion o he numbe o websi es con aining in o ma ion abou Eu opean abbi damage acco ding o a Google sea ch (solid black line), he o al leng h (km) o as -speed ailway in Spain (g ey line; sou ce: h p://www.spanish ailwaysnews.com), and he millions o abbi s hun ed in Spain (do ed line; sou ce: h p://www.mag ama.gob.es and 46). Figu e 2. Municipali ies epo ing Eu opean abbi damage wi hin he s udy a ea (in whi e) acco ding o he analysed websi es (see ex o de ails). This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle Figu e 3. Fa ou abili y model o Eu opean abbi damage in each municipali y o mainland Spain. The small map shows he en i onmen al a ou abili y o he p esence o Eu opean abbi in Spain, on UTM 10 x 10 km squa es. Scale anges om 0 (black) o 1 (whi e). Figu e 4. Mapped con ibu ion o a iables o he model o a ou abili y o Eu opean abbi damage along he s epwise a iable selec ion. S ep: O de o en ance in he model. G een: nega i e con ibu ion o damage. Red: posi i e con ibu ion o damage. HIGHW_RAILW: Leng h o highway and as -speed ailway pe municipali y (m); DIV_FOR: Di e si y o o es uses and ege a ion s uc u e (Simpson di e si y index); DRY_OLIV: D y oli e g o es (% a ea); IRR_VIN: I iga ed ineya ds (% a ea); DEV_OAK: De elopmen phase o oaks (Heigh ); NOAK_CLEFOR: Ca ob o es (% a ea); IRR_OLIV: I iga ed oli e g o es (% a ea); DRY_VIN: D y ineya ds (% a ea); WOLIV_COP: Beech o es (% a ea); ORG_CER: O ganic ce eal (% a ea); F_WILDRABBIT: Fa ou abili y model o he p esence o Eu opean abbi s, based on clima e, opog aphy and human ac i i y (Table S1); PAST: Pas u es (% a ea); WOLIV_FOR: Wild oli e ee o es (% a ea); P_UAA: P opo ion o use ul ag icul u al a ea (% a ea); PAST_SHRUB: Pas u es and sh ubs (% a ea); GALL_FOR: Alignmen (% a ea); PINE_FOR: Bi ch o es (% a ea); OAK_CLEFOR: Boxwood o es (% a ea). Figu e 5. Va ia ion pa i ioning o he inal model. Values show he p opo ions o how much o he a ia ion in en i onmen al a ou abili y o Eu opean abbi damage occu ence is explained exclusi ely by in as uc u e, en i onmen al a ou abili y o Eu opean abbi p esence, na u al ege a ion and land uses ac o s, and which p opo ion was a ibu able o hei sha ed e ec s (in e sec ions). This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle Figu e 1 This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle Figu e 2 This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle Figu e 3 This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle Figu e 4 This a icle is p o ec ed by copy igh . All igh s ese ed. Accep ed A icle Figu e 5 This a icle is p o ec ed by copy igh . All igh s ese ed.