Using Questionnaire Surveys and Occupancy Modelling to Identify Conservation Priorities for the Critically Endangered Balkan Lynx Lynx Lynx Balcanicus

Home , Balkan lynx

Using questionnaire surveys and occupancy modelling to identify conservation priorities for the Critically Endangered Balkan lynx Lynx lynx balcanicus

D IME M ELOVSKI,MANUELA VON A RX,VASKO A VUKATOV C HRISTINE B REITENMOSER-WÜRSTEN,MARINA Đ UROVIĆ ,RAFET E LEZI O LIVIER G IMENEZ,BLEDI H OXHA,SLAVCHO H RISTOVSKI,GJORGJE I VANOV A LEXANDROS A. KARAMANLIDIS,TABEA L ANZ,KUJTIM M ERSINI A LEKSANDAR P EROVIĆ ,AZEM R AMADANI,BARDH S ANAJA,PARSIM S ANAJA G ABRIEL S CHWADERER,ANNETTE S PANGENBERG,ALEKSANDAR S TOJANOV A LEKSANDËR T RAJÇE and U RS B REITENMOSER

Abstract With an estimated ,  adult individuals remain- poaching of lynx, and damage to livestock by lynx. ing, the Critically Endangered Balkan lynx Lynx lynx balca- Poaching was intense throughout the potential range of nicus is one of the rarest, most threatened and least-studied the subspecies, apparently having affected –% of the large carnivores. To identify priority conservation areas and total estimated extant population. Damage to livestock actions for the subspecies, during – we conducted was recorded only in relation to sheep, mainly in the south- , questionnaire surveys throughout the potential range ern part of the lynx’s potential range. Occupancy modelling of the Balkan lynx to () evaluate human–lynx interactions indicated  grid cells with high probability of site use, and identify potential threats, and () determine the prob- which was affected mainly by increased terrain ruggedness ability of site use in  grid cells through occupancy mod- and reduced forest cover. Based on the combined results of elling. Human–lynx interactions were related mainly to our study we identified five priority areas for conservation, as well as in situ habitat protection, community participation in the conservation of the subspecies, and the improve- DIME MELOVSKI* (Corresponding author), VASKO AVUKATOV,SLAVCHO ment and implementation of the existing legal framework as Hristovski†,GJORGJE IVANOV and ALEKSANDAR STOJANOV Macedonian Ecological Society, Skopje, Macedonia. E-mail [email protected] the priority conservation actions for the Balkan lynx. MANUELA VON ARX,CHRISTINE BREITENMOSER-WÜRSTEN,TABEA LANZ and URS Keywords Balkan lynx, conservation, Critically Endan- BREITENMOSER‡ KORA, Carnivore Ecology and Wildlife Management, Bern, Switzerland gered subspecies, distribution range, local ecological knowledge, occupancy modelling, questionnaire surveys MARINA ĐUROVIĆ Public Enterprise for National Parks of Montenegro, Podgorica, Montenegro The data and R code for the analyses are available at RAFET ELEZI and AZEM RAMADANI Finch, Prizren, Kosovo https://github.com/oliviergimenez/occ_balkanlynx OLIVIER GIMENEZ Centre d’Écologie Fonctionnelle et Évolutive, CNRS, Université Montpellier, Montpellier, France

BLEDI HOXHA,KUJTIM MERSINI and ALEKSANDËR TRAJÇE Protection and Preservation of Natural Environment in Albania, Tirana, Albania Introduction ALEXANDROS A. KARAMANLIDIS Faculty of Environmental Sciences and Natural Resource Management, Norwegian University of Life Sciences, Ås, Norway n the midst of a global biodiversity crisis (Butchart et al.,  ALEKSANDAR PEROVIĆ Centre for Protection and Research of Birds, Podgorica, I ) Europe has been witnessing a resurgence in its Montenegro wildlife, with several species showing signs of population re- BARDH SANAJA and PARSIM SANAJA Environmentally Responsible Action Group, covery (Deinet et al., ). Because of the charismatic na- Peja, Kosovo ture of large carnivores, and the challenges associated with GABRIEL SCHWADERER and ANNETTE SPANGENBERG EuroNatur Foundation, effectively managing and protecting them, their recovery Radolfzell, Germany has received considerable public attention in Europe *Also at: Faculty of Forest Sciences, Wildlife Sciences, University of Göttingen,  Göttingen, Germany (Chapron et al., ). Large carnivores have been recover- †Also at: Institute of Biology, Faculty of Natural Sciences and Mathematics, Ss. ing across the continent, including the Balkan Peninsula Cyrill and Methodius University, Skopje, Macedonia (Karamanlidis et al., ; Ivanov et al., ); however, an ‡Also at: Centre for Fish and Wildlife Health, University of Bern, Bern, Switzerland exception to this positive development is the Critically Received  December . Revision requested  March . Endangered Balkan lynx Lynx lynx balcanicus (Melovski Accepted  March . First published online  December . et al., ).

Oryx, 2020, 54(5), 706–714 © 2018 Fauna & Flora International doi:10.1017/S0030605318000492 Downloaded from https://www.cambridge.org/core. IP address: 170.106.33.19, on 29 Sep 2021 at 05:48:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605318000492 Conservation priorities for the lynx 707

The Balkan lynx was described in the th century (Lahoz‐Monfort et al., ). Therefore, site-occupancy (Bureš, ) but it was not until the early st century models have been developed specifically to distinguish be- that genetic evidence indicated its taxonomic status as a sub- tween non-detection and absence by modelling the imperfect, species of the Eurasian lynx Lynx lynx (Breitenmoser- possibly heterogeneous observation process (MacKenzie, Würsten & Obexer-Ruff, ; Gugolz et al., ). ). This modelling framework has been used successfully Although the phylogenetic relationship of the Balkan lynx to analyse data from multiple interviewees reporting the de- to other subspecies of Eurasian lynx is still unclear, its mor- tection or non-detection of large carnivores and to infer their phological distinctiveness and isolation from the nearest distribution, accounting for imperfect detection (e.g. Petracca lynx subspecies in the Carpathian Mountains (Mirić, ) et al., ; Taubmann et al., ). justifies its recognition as a distinct evolutionary significant The main aims of this study were to use a questionnaire unit for conservation (Vogler & Desalle, ). survey to () evaluate human–lynx interactions to identify Historically, the Balkan lynx suffered a fate similar to that potential threats to the survival of the Balkan lynx, and of most other large carnivores in Europe, where increased () determine the probability of site use. The results are used habitat alteration and persecution led to its extermination to identify research and conservation priorities to safeguard from large parts of the continent (Breitenmoser, ). the survival of this subspecies. Following World War II legal and administrative actions re- sulted in the partial recovery of the subspecies; in  the  population was estimated to comprise individuals, lo- Study area cated mainly in the south-west of the Balkan Peninsula (Mirić, ). However, political unrest in the region at the The study was conducted in presumed distribution areas turn of the century combined with habitat deterioration and of the Balkan lynx (von Arx et al., ; Kaczensky et al., poaching are believed to have led to a sharp population de- a,b), including areas that were considered by Grubač cline and range constriction, which has brought the Balkan (, ) to be potentially within the subspecies’ lynx to the brink of extinction (Melovski, ): it is esti- range. We also included some localities outside the pre- mated that there are ,  adult individuals remaining in sumed area of distribution, for which circumstantial evi- the wild, hence the subspecies’ categorization as Critically dence indicated the presence of the subspecies (Grubač,  Endangered on the IUCN Red List (Melovski et al., ). ). Thus, the study area comprised , km of pre- The Balkan lynx is legally protected in Albania, Kosovo (re- dominantly mountainous terrain in Albania, Kosovo, ferred to according to U.N. Security Council Resolution Macedonia and Montenegro (Fig. ). The mean altitude of ) and Macedonia (for reasons of neutrality and brevity the villages where questionnaire surveys were carried out the name Macedonia is used for the country with the con- was , m and the mean human population density was  stitutional name Republic of Macedonia, admitted to the . people per km (Kosovo Agency of Statistics, ; UN under the provisional designation ‘the former Institute of Statistics, Albania, ; Statistical Office of Yugoslav Republic of Macedonia’), and a compensation sys- Montenegro, ; State Statistical Office, Macedonia, tem for damage to livestock by lynx has been established in ). The vegetation in the study area is predominantly for- Macedonia. Effective conservation of biodiversity requires est, with beech Fagus sylvatica, fir Abies borisii-regis and detailed, quantitative scientific information, which for various types of oak Quercus spp. The region is character- many wildlife populations is not available (Gilchrist et al., ized by high-mountain pastures, river valleys and rural an- ). There has therefore been increasing consensus thropogenic landscapes. among conservationists that alternative sources of information are necessary for protecting nature and that local ecological knowledge can provide important information on Methods the status of wildlife populations and should be integrated  in environmental management (Drew, ; Anadón Sampling approach and data collection et al., ). Local ecological knowledge has been used to provide in- Following the design of similar surveys of large carnivores in formation on species’ distributions and status over large Europe (e.g. Kaczensky et al., a,b; Chapron et al., ) landscapes, with moderate effort (Taubmann et al., ). and taking into account the home ranges of the European However, there are limitations and caveats regarding its (Breitenmoser-Würsten et al., ) and Balkan lynx use (Caruso et al., ), especially when dealing with popu- (D. Melovski, unpubl. data), the study area was overlaid lations of large carnivores for which detectability is almost with the  ×  km Universal Transverse Mercator refer- certainly ,  (Louvrier et al., ), which can lead to false ence grid (UTMN/WGS) and  grid cells were se- negatives (Kéry, ). Falsely assuming perfect detection lected that included the entire potential range of the can lead to an underestimation of the species’ distribution Balkan lynx (Fig. ).

Data analysis

Human–lynx interactions were evaluated using descriptive statistics. To estimate site use of the Balkan lynx we implemented a single-season occupancy model (MacKenzie et al., ). Such models use replicates in time and space to distinguish absence from imperfect detection. Grid cells were used as spatial replicates, and each interviewee was considered to be a replicate for one or several cells (Taubmann et al., ). We added three buffers related to the profiles of the survey participants (see Louvrier et al., , for a similar approach), depending on their area of knowledge (Zeller et al., ). We considered hunters and sheep herders who owned .  sheep to have an area of knowledge of  km radius around their place of residence; sheep herders who owned ,  sheep were considered to have an area of knowledge of  km radius; and people who did not belong to the aforementioned categories were considered to have an area of knowledge of  km radius. We estimated the probability of site use, accounting for imperfect detection. Although the home ranges of female Balkan lynx match the size of our sampling unit, males tend to have larger home ranges. As a consequence, and assuming that move- ments are random, we interpreted occupancy as site use ra- ther than the proportion of the area occupied by the subspecies (MacKenzie & Nichols, ), usually referred to as third-order selection (Johnson, ). As people’s various levels of experience in the field can lead to differences in FIG. 1 The survey area in the south-western Balkans, showing their ability to detect a species (Davis & Wagner, ), we the   ×  km grid cells in which we conducted considered a covariate for the level of expertise (i.e. a con- questionnaire surveys, with the probability of site use of the – Balkan lynx Lynx lynx balcanicus as estimated through tinuous variable with values of , from the most to the occupancy modelling. least experienced observer) on the detection probability, to address this issue. The level of expertise was evaluated sub- jectively while conducting the interviews. To account for During – we conducted systematic, face-to- heterogeneity in the sampling effort we also considered face questionnaire surveys, using local ecological knowl- the number of interviewees per cell as a covariate potentially edge to collect information on site use by large carnivores, affecting the detection probability. With regard to the prob- and their interactions with people. In each of the grid cells ability of site use we used the proportion of forest cover (i.e. we visited at least one village and interviewed up to eight forest cover was estimated using a selection of natural and people of various professional backgrounds (e.g. hunters, semi-natural land-cover categories of the Corine Land foresters, livestock breeders, beekeepers, farmers). We Cover  vector data); terrain ruggedness generated asked  quantitative and qualitative questions regarding from a Shuttle Radar Topography Mission (SRTM) digital the presence of large carnivores in the area, human–carni- elevation model using the ‘ruggedness index’ tool in QGIS vore interactions, husbandry practices, and conservation v... (Quantum GIS Development Team, ); and alti- issues; of the  questions, we used six related to tude (SRTM digital elevation model; USGS, ) as site- human–lynx interactions, seven to lynx presence and one specific covariates potentially affecting site use and detec- to site use, in the data analysis. We interviewed , peo- tion (Long et al., ; Karanth et al., ; Martínez- ple in total (Albania, ; Kosovo, ; Macedonia, ; Martí et al., ). Site covariates were averaged for the Montenegro, ; mean . per grid cell). Hunters entire sampling unit, standardized by centring on their (.%) comprised the largest group of interviewees, fol- mean value, scaled by the standard deviation and incorpo- lowed by livestock breeders (.%), farmers (.%), fores- rated in the model through logistic-linear relationships, ex- ters (.%) and shop owners (.%). Local residents cept for altitude, for which we also considered a potential comprised .% of interviewees, the remainder being sea- optimum through a logistic-quadratic relationship. We con- sonal residents in the study area. sidered  models, which were fitted using the package

TABLE 1 Model selection results, ranked by Akaike information criterion corrected for small sample size (AICc), for occupancy modelling to identify conservation priorities for the Balkan lynx Lynx lynx balcanicus. Of the  models fitted, only those with ΔAICc ,  are reported.

Model1 AICc ΔAICc2 p(forest + expert + rug + sampeff) ψ(forest + rug) 4157.8 0 p(forest + expert + rug + sampeff) ψ(forest + rug + alt) 4158.5 0.70 p(forest + expert + rug + sampeff) ψ(forest + rug + alt + alt2) 4159.4 1.62

 p, detection probability; ψ, occupancy probability. Covariates: forest, proportion of forest cover; expert, level of expertise; rug, terrain ruggedness; sampeff, number of respondents; alt, altitude.  ΔAICc, difference between the AICc of the current model and the AICc of the model with the lowest AICc.

unmarked in Rv... (Fiske & Chandler, ) and com- mainly in western and southern Macedonia (Fig. b), within pared using the Akaike information criterion corrected for  months prior to the interview. Attacks were recorded for small sample size (AICc; Burnham & Anderson, ). To ac- sheep that were protected by guarding dogs (in  cases), shep- count for model selection uncertainty we used model aver- herds () or a combination of shepherds and dogs (), and aging as implemented in the R package MuMIn (Bartoń, sheep that were penned at night (). Only two cases were re- ), considering all models with ΔAIC , ,whereΔAIC is ported to the relevant authorities and no compensation was the difference between the AICc of a given model and the received. To mitigate damage to livestock by lynx, shepherds AICc of the model with the lowest AICc. We built an occu- used various measures, including moving herds to other graz- pancy map by calculating the estimated probability of site ing areas ( cases), intensifying herding (), and hunting or use for a given grid cell using the model-averaged parameter poisoning lynx (). Two collisions of lynx with vehicles on estimates and the value of the covariates for this cell. The re- secondary roads in western Macedonia were also recorded. sulting map depicts lynx presence as reflected in the questionnaires, and we assume that interviewees’ perception of lynx presence indicates actual lynx presence. Probability of site Site use use values were used to group the cells into three categories The top-ranked single-season occupancy models  using the Jenks optimization method (Jenks, ): low (ΔAICc , , cumulative AIC weight .) included an #     (probability of site use . ), medium ( . effect of the number of interviewees (model-averaged esti- , #   probability of site use . ) and high (probability mate = . ± SE .), terrain ruggedness (. ± SE .), $    of site use . )(Fig. ). forest cover (−. ± SE .) and level of expertise (. ± SE .) on the detection probability, and an effect of forest cover (−. ± SE .) and terrain ruggedness Results (. ± SE .) on the site use probability (Table ). There was some uncertainty regarding the altitude covariate and – Human lynx interactions whether or not it had an effect on the probability of site −  ±   The questionnaires recorded quantitative and qualitative use (model-averaged estimate linear term = . SE . , −  ±   data on various types of human–lynx interactions; however, quadratic term = . SE . ). the data facilitated evaluation of only the following three The site-use map resulting from the model-averaged pa-  types of interactions: poaching of lynx, damage to livestock rameter estimates included areas with low probability of by lynx, and lynx–vehicle collisions. site use, mainly in south-western Montenegro and central  Lynx poaching was reported by  interviewees (Albania, Kosovo, areas with medium probability of site use, mainly ; Kosovo, ; Macedonia, ; Montenegro, )in locations in south-western parts of Macedonia and southern Kosovo,  in the study area. Eleven interviewees reported killing lynx, and areas with high probabilities of site use, mainly in the – – and  reported indirect knowledge of lynx being killed. border region of Albania Kosovo and Albania Macedonia, Most poaching cases () were reported to have taken central and central-south Albania, and in the border region  place in the border areas of Macedonia and Albania, and of Albania and Montenegro (Fig. ). Albania and Montenegro (Fig. a) after .On occa-     sions (Albania, ; Kosovo, ; Macedonia, ; Montenegro, ) Discussion the survey teams were able to verify lynx poaching either by inspecting pelts, trophies and stuffed animals or by evaluat- The Balkan lynx is the least studied and known large carni- ing photographs. vore in Europe. With an estimated total population of ,  On  occasions interviewees reported that lynx had at- adult individuals in the wild, identifying human–lynx inter- tacked their livestock. The attacks occurred in  villages, actions and estimating site use should be considered

important research priorities for the subspecies (Inskip & As identified in the IUCN Red List assessment (Melovski Zimmermann, ). Our study provides the first wide- et al., ), habitat degradation and prey depletion are sig- scale assessment of the subspecies’ interactions with people, nificant factors contributing to the decline of the Balkan and its potential occupancy throughout its range. lynx. Forest degradation, which has been reported as one Poaching is a common and globally recognized challenge of the main conservation threats for lynx throughout to the conservation of wildlife, especially for large carnivores Europe (Breitenmoser et al., ), appeared to be intensive (Ripple et al., ). Although it is difficult, given its illegal throughout the entire potential range of the Balkan lynx, in- nature, to evaluate accurately its effect on population dynam- cluding the areas of high probability of site use; this is in ac- ics (Liberg et al., ), poaching appears to have been an im- cordance with information from the area (Albania, Stahl, portant factor in the mortality of lynx in Europe, and a threat ; Macedonia, Kolevska et al., ). Prey depletion is a to the species’ conservation (Červený et al., ). This ap- conservation threat for lynx in Eastern Europe (Schmidt, pears also to be the case for the Balkan lynx, with the survey a,b; Yom-Tov et al., ) and could also play a negative results indicating that since the beginning of this century the role in the survival of the Balkan lynx. As no causative effect equivalent of –% of the total estimated, currently sur- for prey depletion could be established in this study, dedi- viving population could have fallen victim to poaching. cated research is necessary to evaluate the impact of this However, this statement should be considered cautiously, threat on the survival of the subspecies; this should be one as during our questionnaire we were unable to determine if of the research priorities for the Balkan lynx. the  reported cases of lynx poaching referred to unique Interview-based surveys provide a cost-effective and lo- cases or if there was a degree of overlap. gistically practical alternative to time-consuming and Damage to livestock by large carnivores is one of the capacity-demanding field surveys, especially for large carni- most common causes of conflict between people and carnivores that are rare and difficult to detect (Gros et al., ). vores worldwide (Kaczensky, ) and poses an important However, evaluation of local ecological knowledge is not an conservation challenge in the recovery of several lynx popu- objective way to assess people’s generic reports (Huntington, lations in Europe (Trouwborst, ). However, the number ). We discovered that rare and/or spectacular events, of incidents of depredation recorded in our study was small, such as the sighting of a lynx, are remembered for a long probably as a result of the relatively good herding practices time, and this could have influenced the site occupancy ana- in the area (i.e. use of guarding dogs and shepherds, and lysis. The shortcoming of the low detectability of the lynx, a penning at night); however, this could also be regarded as result of their cryptic nature and low abundance, was ac- a measure to protect against other, more damaging large counted for through the occupancy framework, which ad- carnivores, such as wolves Canis lupus and bears Ursus arc- dresses imperfect detection (Taubmann et al., ). tos. The fact that the majority of sheep herders did not seek Using this methodology we provide the first map of the compensation for damage because of distrust in the com- probability of site use of the Balkan lynx in the south- pensation system or because they did not know that such western Balkans. Lynx detectability was positively affected a system existed indicates that the design of the damage com- by the number of interviewees, their level of expertise, and pensation system needs improvement. Improvements to the terrain ruggedness, and negatively affected by the percent- damage compensation system in Macedonia should include age of forest cover. These results are consistent with our ex- more rigorous damage evaluation procedures, resulting in pectations. However, our occupancy modelling results are the swifter delivery of compensation (Karamanlidis et al., more difficult to interpret, as they were positively affected ), and information campaigns to make the compensation by terrain ruggedness, but negatively affected by forest system more accessible. Modern compensation schemes can cover. Lynx presence is generally associated with rugged, be designed in such a way that they alleviate the financial forested areas (Lone et al., ). However, research has losses from livestock depredation but also aim to reduce shown that in human-modified areas (such as those poaching of large carnivores (Treves et al., ). throughout our study area) lynx may exhibit nuanced be- Road mortality can affect wildlife through habitat frag- havioural responses and select areas with medium levels of mentation and reduced gene flow (Clevenger et al., ). human modification, avoiding areas with high or low levels During the study only two road mortalities were recorded of modification (Bouyer et al., ). Considering that our throughout the entire potential range of the subspecies, results depict only the human perception of lynx habitat and therefore road mortality should not currently be re- use, and given the lack of relevant telemetry data, we believe garded as a major threat to its survival. However, consider- that studying Balkan lynx habitat use should be a research ing the conservation status of the Balkan lynx and the priority. economic development currently taking place throughout Information provided by local people in interview-based its range, care should be taken in the spatial planning and surveys may include errors (e.g. false positive detections) construction of the road network (Kusak et al., )to that can produce biased results in occupancy studies avoid the development of a new conservation threat. (Miller et al., ). To mitigate such errors we recommend

FIG. 2 Survey grid cells in which interviewees indicated Balkan lynx (a) had been poached, and (b) had caused damage to their livestock.

the use of site-occupancy models accounting for false posi- recovery of the Balkan lynx (Fig. ) and protected effectively tives, which have been applied to data collected via (Schwaderer et al., ) are the Macedonian part of the interview-based surveys (Pillay et al., ). If a similar sur- Shar Planina Mountains (Fig. , area ), the Shebenik– vey were to be conducted in  years time, we recommend Jablanica Mountains (Fig. , area ) and the Albanian Alps using the same questionnaire, to assess trends in the dynam- (Fig. ,area). ics of occupancy by considering local extinction and colo- nization events with dynamic occupancy models (e.g. Taubmann et al., ). Conservation priorities Our results on lynx–human interactions (Fig. ) and on With well-planned conservation action, carnivore populations the probability of site use (Fig. ) together indicate that the should continue to survive despite human population growth Balkan lynx is most likely to occur in the border region be- and modification of habitat. Based on the results and experi- tween the four countries in the study area, as well as in some ences of our interview-based survey, and taking into account parts of central, northern and southern Albania. Evaluating conservation threats, priorities and solutions from other, the results of our study in conjunction with information better-studied lynx populations in Europe (Breitenmoser from dedicated, localized studies on lynx presence and be- et al., ), we identify three main conservation priorities haviour using various methodologies (Melovski et al., , for the Balkan lynx that should be implemented on both a ; PPNEA, ; Stojanov et al., ) we identify five pri- local and wider geographical scale: ority conservation areas for the Balkan lynx (Fig. ). Two of these areas emerge as the core areas of the surviving Balkan (1) Conservation efforts should focus on protecting the re- lynx population: Mavrovo National Park in Macedonia maining populations in situ, their critical habitat and (Fig. ,area) and the Munella Mountains in central-north their prey base (Breitenmoser et al., ). Actions Albania (Fig. , area ) (Melovski et al., ). The other should focus primarily on the priority conservation three areas that should be considered important for the areas identified and should include law enforcement

Acknowledgements The questionnaire surveys were carried out within the framework of the Balkan Lynx Recovery Programme funded by the MAVA foundation, Switzerland. All research activities were carried out under the research permits 11-10190/2/20.11.2009, 11-2186/ 2/19.02.2010 and 11-546/2/29.01.2013 issued by the Macedonian Ministry of Environment and Physical Planning; and protocol number 2777/22.12.2009 of the Albanian Ministry of Environment, Forests and Water Administration, Directorate of Nature Protection Policies.

Author contributions Research conceptualization and design: DM, MvA, AAK, SH, GS, ASP, UB; data collection: DM, MĐ, RE, BH, GjI, KM, AP, AR, BS, PS, AST, AT; statistical analysis: DM, VA, OG, AAK; preparation of maps and creation of the common database: VA; writing: all authors.

Conflicts of interest None.

Ethical standards All interviews in this study were conducted with prior consent of the interviewees, who had the option of remaining anonymous.

References

ANADÓN, D.J., G IMÉNEZ,A.,BALLESTAR,R.&PÉREZ,I.() Evaluation of local ecological knowledge as a method for collecting extensive data on animal abundance. Conservation Biology, , –. BARTOŃ,K.() MuMIn: Multi-Model Inference. R package version ... BOUYER, Y., SAN MARTIN, G., PONCIN, P., BEUDELS-JAMAR, R.C., ODDEN,J.&LINNELL, J.D.C. () Eurasian lynx habitat selection in human-modified landscape in Norway: effects of different human habitat modifications and behavioral states. Biological Conservation, FIG. 3 Five important areas for the conservation of the Balkan , –. lynx, identified based on occupancy modelling (Fig. ) and BREITENMOSER,U.() Large predators in the Alps: the fall and rise questionnaire surveys (Fig. ): , Shebenik–Jablanica and of man’s competitors. Biological Conservation, , –. surroundings; , Mavrovo National Park and surroundings; BREITENMOSER, U., BREITENMOSER-WÜRSTEN, C., OKARMA, H., , Shar Planina Mountain; , Munella Mountains and KAPHEGYI, T., KAPHEGYI-WALLMANN,U.&MÜLLER, U.M.  surroundings; , Albanian Alps. () Action Plan for the Conservation of the Eurasian Lynx in Europe. Council of Europe, Strasbourg, France. and management actions (i.e. scientific monitoring of BREITENMOSER, U., VON ARX, M., BEGO, F., IVANOV, G., KEÇI, E., MELOVSKI,D.etal.() Strategic planning for the conservation of local lynx populations, forest management, restoration the Balkan lynx. In Proceedings of the III Congress of Ecologists of the of habitats and improvement of compensation sys- Republic of Macedonia with International Participation, pp. –. tems) in the areas that already have a legally protected Macedonian Ecological Society, Struga, FYR Macedonia. status (Mavrovo National Park), and the designation BREITENMOSER-WÜRSTEN,C.&OBEXER-RUFF,G.() Population of new protected areas for the subspecies (Munella and conservation genetics of two re-introduced lynx (Lynx lynx) populations in Switzerland – a molecular evaluation  years after Mountains in central-north Albania). translocation. In Proceedings of the nd Conference on the Status and (2) Active participation and involvement of various stake- Conservation of the Alpine Lynx Population (SCALP), pp. –. holder groups in the conservation of this subspecies is Amden, Switzerland. important. In the case of the Balkan lynx it is crucial to BREITENMOSER-WÜRSTEN, C., ZIMMERMANN, F., STAHL,P., involve hunters in the conservation process, as they VANDEL, J.-M., MOLINARI-JOBIN, A., MOLINARI, P. et al. () can play an important role in reducing poaching and Spatial and social stability of a Eurasian lynx Lynx lynx population: an assessment of  years of observations in the Jura Mountains. prey depletion, and in participating in monitoring. Wildlife Biology, , –. (3) Despite the Balkan lynx being protected throughout its BUREŠ,I.() Risove v Makedonija (Lynx in Macedonia). Priroda, range, trophies can still be found in cafeterias in rural , –. settlements throughout its range, yet cases of poaching BURNHAM, K.P. & ANDERSON, D.R. () Model Selection and of the subspecies have never been prosecuted. It is ne- Inference: A Practical Information Theoretic Approach. Springer, New York, USA. cessary to improve and enforce the legal framework for BUTCHART, S.H.M., WALPOLE, M., COLLEN, B., VAN STRIEN, A., the protection of the Balkan lynx effectively, to prevent SCHARLEMANN, J.P.W., ALMOND, R.E.A. et al. () Global the extinction of this rare subspecies. biodiversity: indicators of recent declines. Science, , –.

CARUSO, N., VIDAL, E.L., GUERISOLI,M.&LUCHERINI,M.() Report prepared with the assistance of Istituto di Ecologia Applicata Carnivore occurrence: do interview-based surveys produce and with the contributions of the IUCN/SSC Large Carnivore unreliable results? Oryx, , –. Initiative for Europe under contract N///SER/B. ČERVENÝ, J., KOUBEK,P.&BUFKA,L.() Eurasian lynx (Lynx KARAMANLIDIS, A.A., SANOPOULOS, A., GEORGIADIS,L.& lynx) and its chance for survival in Central Europe: the case of the ZEDROSSER,A.() Structural and economic aspects of human– Czech Republic. Acta Zoologica Lituanica, , –. bear conflicts in Greece. Ursus, , –. CHAPRON, G., KACZENSKY, P., LINNELL, J.D.C., VON ARX, M., HUBER, KARAMANLIDIS, A.A., DE GABRIEL HERNANDO, M., KRAMBOKOUKIS, D., ANDRÉN, H. et al. () Recovery of large carnivores in Europe’s L. & GIMENEZ,O.() Evidence of a large carnivore population modern human-dominated landscapes. Science, , –. recovery: counting bears in Greece. Journal for Nature Conservation, CLEVENGER, A.P., CHRUSZCZ,B.&GUNSON, K.E. () Highway , –. mitigation fencing reduces wildlife–vehicle collisions. Wildlife KARANTH, K.K., NICHOLS, J.D., HINES, J.E., KARANTH, K.U. & Society Bulletin, , –. CHRISTENSEN, N.L. () Patterns and determinants of mammal DAVIS,A.&WAGNER, J.R. () Who knows? On the importance of species occurrence in India. Journal of Applied Ecology, , – identifying “experts” when researching local ecological knowledge. . Human Ecology, , –. KÉRY,M.() Towards the modelling of true species distributions. DEINET,S.,IERONYMIDOU,C.,MCRAE, L., BURFIELD, I.J., FOPPEN,R. Journal of Biogeography, , –. P., C OLLEN,B.&BÖHM,M.() Wildlife Comeback in Europe: The KOLEVSKA, D.D., BLINKOV, I., TRAJKOV,P.&MALETIĆ,V.() Recovery of Selected Mammal and Bird Species. ZSL, London, UK. Reforestation in Macedonia: history, current practice and future DREW, J.A. () Use of traditional ecological knowledge in marine perspectives. Reforesta, , –. conservation. Conservation Biology, , –. KOSOVO AGENCY OF STATISTICS () Http://ask.rks-gov.net/en/ FISKE,I.&CHANDLER, R.B. () unmarked:AnR package for fitting kosovo-agency-of-statistics [accessed November ]. hierarchical models of wildlife occurrence and abundance. Journal KUSAK, J., HUBER, D., GOMERČIĆ, T., SCHWADERER,G.&GUŽVICA, of Statistical Software, , –. G. () The permeability of highway in Gorski kotar (Croatia) for GILCHRIST, G., MALLORY,M.&MERKEL,F.() Can local large mammals. European Journal of Wildlife Research, , –. ecological knowledge contribute to wildlife management? Case LAHOZ‐MONFORT, J.J., GUILLERA‐ARROITA,G.&WINTLE, B.A. studies of migratory birds. Ecology and Society, (), . () Imperfect detection impacts the performance of species GROS, P.M., KELLY, M.J. & CARO, T.M. () Estimating carnivore distribution models. Global Ecology and Biogeography, , –. densities for conservation purposes: indirect methods compared to LIBERG, O., CHAPRON, G., WABAKKEN, P., PEDERSEN, H.C., HOBBS, baseline demographic data. Oikos, , –. N.T. & SAND,H.() Shoot, shovel and shut up: cryptic poaching GRUBAČ,B.() The lynx (Lynx lynx) in Serbia. Zaštita prirode , slows restoration of a large carnivore in Europe. Proceedings of the –. Royal Society B, , –. GRUBAČ,B.() Contribution on the Balkan lynx Lynx lynx LONE, K., LOE, L.E., GOBAKKEN, T., LINNELL, J.D.C., ODDEN, J., martinoi (Mirić, ) in Macedonia and Montenegro. Zaštita REMMEN,J.&MYSTERUD,A.() Living and dying in a multi‐ prirode, , –. predator landscape of fear: roe deer are squeezed by contrasting GUGOLZ, D., BERNASCONI, M.V., BREITENMOSER-WÜRSTEN,C.& pattern of predation risk imposed by lynx and humans. Oikos, , WAMDELER,P.() Historical DNA reveals the phylogenetic –. position of the extinct Alpine lynx. Journal of Zoology, , –. LONG, R.A., DONOVAN, T.M., MACKAY, P., ZIELINSKI, W.J. & BUZAS, HUNTINGTON, H.P. () Using traditional ecological knowledge in J.S. () Effectiveness of scat detection dogs for detecting forest science: methods and applications. Ecological Applications, , – carnivores. Journal of Wildlife Management, , –. . LOUVRIER, J., DUCHAMP, C., LAURET, V., MARBOUTIN, E., CUBAYNES, INSKIP,C.&ZIMMERMANN,A.() Human–felid conflict: a review S., CHOQUET, R. et al. () Mapping and explaining wolf of patterns and priorities worldwide. Oryx, , –. recolonization in France using dynamic occupancy models and INSTITUTE OF STATISTICS,ALBANIA () Http://www.instat.gov.al opportunistic data. Ecography, , –. [accessed November ]. MACKENZIE, D.I. () Occupancy Estimation and Modeling: IVANOV, G., KARAMANLIDIS, A.A., STOJANOV, A., MELOVSKI,D.& Inferring Patterns and Dynamics of species Occurrence. Academic AVUKATOV,V.() The re-establishment of the golden jackal Press, London, UK. (Canis aureus) in FYR Macedonia: implications for conservation. MACKENZIE, D.I. & NICHOLS, J.D. () Occupancy as a surrogate Mammalian Biology, , –. for abundance estimation. Animal Biodiversity and Conservation, JENKS, G.F. () The data model concept in statistical mapping. , –. International Yearbook of Cartography, , –. MACKENZIE, D.I., NICHOLS, J.D., LACHMAN, G.B., DROEGE, S., JOHNSON, D.H. () The comparison of usage and availability ANDREW ROYLE,J.&LANGTIMM, C.A. () Estimating site measurements for evaluating resource preference. Ecology, , –. occupancy rates when detection probabilities are less than one. KACZENSKY,P.() Large carnivore depredation on livestock in Ecology, , –. Europe. Ursus, , –. MARTÍNEZ-MARTÍ, C., JIMÉNEZ-FRANCO, M.V., ROYLE, J.A., KACZENSKY, P., CHAPRON, G., VON ARX, M., HUBER, Đ., ANDRÉN,H. PALAZÓN, J.A. & CALVO, J.F. () Integrating occurrence and &LINNELL,J.(a) Status, Management and Distribution of Large detectability patterns based on interview data: a case study for Carnivores – Bear, Lynx, Wolf & Wolverine – in Europe. Part threatened mammals in Equatorial Guinea. Scientific Reports, , . Report prepared with the assistance of Istituto di Ecologia . Applicata and with the contributions of the IUCN/SSC Large MELOVSKI,D.() Status and distribution of the Balkan lynx (Lynx Carnivore Initiative for Europe under contract N// lynx martinoi Mirić ) and its prey. MSc thesis. University of /SER/B. Montenegro, Montenegro. KACZENSKY, P., CHAPRON, G., VON ARX, M., HUBER, Đ., ANDRÉN,H. MELOVSKI, D., IVANOV, G., STOJANOV, A., TRAJÇE, A., &LINNELL,J.(b) Status, Management and Distribution of Large ZIMMERMANN,F.&VON ARX,M.() First camera-trap survey Carnivores – Bear, Lynx, Wolf & Wolverine – in Europe. Part . in the National Park Mavrovo, Macedonia. In International

Conference on Biological and Environmental Sciences, pp. –. SCHMIDT,K.(b) Factors shaping the Eurasian lynx (Lynx lynx) University of Tirana, Tirana, Albania. population in the north-eastern Poland. Nature Conservation, , MELOVSKI, D., STOJANOV,A.&IVANOV,G.() Status, Ecology and –. Land Tenure System of the Critically Endangered Balkan Lynx in SCHWADERER, G., SPANGENBERG, A., MELOVSKI, D., TRAJÇE,A.& Macedonia and Albania. Radio-telemetry annual report, capture BEGO,F.() Protected area in species conservation – the session . Unpublished report. MES, Skopje, Republic of Macedonia. protected area component within the frame of the Balkan Lynx MELOVSKI, D., IVANOV, G., STOJANOV, A., TRAJCE, A., HOXHA, B., Recovery Programme. In Proceedings of the III Congress of Ecologists VON ARX, M. et al. () Distribution and conservation status of the of Macedonia with International Participation, pp. –. Balkan lynx (Lynx lynx balcanicus Bureš, ). In Proceedings of the Macedonian Ecological Society, Skopje. IV Congress of Ecologists of the Republic of Macedonia with STAHL,J.() The rents of illegal logging: the mechanisms behind International Participation, pp. –. Macedonian Ecological the rush on forest resources in southeast Albania. Conservation & Society, Ohrid, Macedonia. Society, , –. MELOVSKI, D., BREITENMOSER, U., VON ARX, M., STATE STATISTICAL OFFICE,MACEDONIA () Http://www.stat. BREITENMOSER-WÜRSTEN,C.&LANZ,T.() Lynx lynx ssp. gov.mk [accessed November ]. balcanicus. The IUCN Red List of Threatened Species. Http://dx.doi. STATISTICAL OFFICE OF MONTENEGRO () Http://www.monstat. org/./IUCN.UK.-.RLTS.TA.en org [accessed November ]. [accessed  August ]. STOJANOV, A., MELOVSKI, D., SAROV, A., IVANOV,G.&AVUKATOV, MILLER, D.A.W., NICHOLS, J.D., GUDE, J.A., RICH, L.N., PODRUZNY, V. () Deterministic camera-trapping study in Mavrovo National K.M., HINES,J.E.&MITCHELL, M.S. () Determining occurrence Park, Macedonia in . Unpublished report. MES, Skopje, dynamics when false positives occur: estimating the range dynamics Republic of Macedonia. of wolves from public survey data. PLoS ONE, (), e. TAUBMANN,J.,SHARMA,K.,UULU, K.Z., HINES,J.E.&MISHRA,C. MIRIĆ, Đ.() Lynx lynx martinoi ssp. nova (Carnivora, Mammalia) () Status assessment of the Endangered snow leopard Panthera – neue Luchsunterart von der Balkanhalbinsel. Bulletin of the uncia and other large mammals in the Kyrgyz Alay, using community Museum of Natural History of Belgrade,B, –. knowledge corrected for imperfect detection. Oryx, , –. MIRIĆ, Đ.() The lynx populations of the Balkan Peninsula (Mirić, TREVES, A., WALLACE, R.B., NAUGHTON-TREVES,L.&MORALES,A. ). Pos. izd. SANU , Odel prir.-mat. nauka, , –. () Co-managing human–wildlife conflicts: a review. Human PETRACCA, L.S., RAMÍREZ-BRAVO, O.E. & HERNÁNDEZ-SANTÍN,L. Dimensions of Wildlife, , –. () Occupancy estimation of jaguar Panthera onca to assess the TROUWBORST,A.() Managing the carnivore comeback: international value of east-central Mexico as a jaguar corridor. Oryx, , –. and EU species protection law and the return of lynx, wolf and bear to PILLAY, R., MILLER, D.A.W., HINES, J.E., JOSHI, A.A. & Western Europe. Journal of Environmental Law, , –. MADHUSUDAN, M.D. () Accounting for false positives USGS (UNITED STATES GEOLOGICAL SURVEY)() Shuttle Radar improves estimates of occupancy from key informant interviews. Topography Mission,  Arc Second Multiple Digital Elevation Model Diversity and Distributions, , –. Scenes Covering the Balkan Peninsula. Global Land Cover Facility, PPNEA (PROTECTION AND PRESERVATION OF NATURAL University of Maryland, College Park, Maryland. ENVIRONMENT IN ALBANIA)() Report on the Munella VOGLER, P.A. & DESALLE,R.() Diagnosing units of conservation mountain, summary of the findings from the Balkan lynx recovery management. Conservation Biology, , –. programme. Unpublished report. PPNEA, Tirana, Albania. VON ARX,M.,BREITENMOSER-WÜRSTEN,C.,ZIMMERMANN,F.& QUANTUM GIS DEVELOPMENT TEAM () Quantum GIS BREITENMOSER,U.() Status and Conservation of the Geographical Information System. Open Source Geospatial Eurasian Lynx (Lynx lynx) in Europe in .KORA,Bern,Switzerland. Foundation Project, Beaverton, USA. YOM-TOV, Y., KVAM,T.&WIIG,Ø.() Lynx body size in Norway RIPPLE, W.J., ESTES, J.A., BESCHTA, R.L., WILMERS, C.C., RITCHIE,E. is related to its main prey (roe deer) density, climate, and latitude. G., HEBBLEWHITE, M. et al. () Status and ecological effects of Ambio, , –. the world’s largest carnivores. Science, , . ZELLER, K.A., NIJHAWAN, S., SALOM-PÉREZ, R., POTOSME, S.H. & SCHMIDT,K.(a) Behavioural and spatial adaptation of the HINES, J.E. () Integrating occupancy modeling and interview Eurasian lynx to a decline in prey availability. Acta Theriologica, , data for corridor identification: a case study for jaguars in –. Nicaragua. Biological Conservation, , –.