The Lynx in the Italian South-Eastern Alps

Hystriu. (n.s.) 10 (I) (19%): 55-64

THE LYNX IN THE ITALIAN SOUTH-EASTERN ALPS

PAOLO MOLINARI

Uiiivcr-srty .f Parlua. Forest Zoology Institute, Strada Ronica n"16 - Polo 09 Lrgntiro (PD) ltuly

ABSTRACT - From 1986 to May 1995 I collected records of signs of lynx presence in the region of Friuli Venezia Giulia. A series of regularly checked trail transects, explorative excursions, and the collection ol second-hand observations led to a total of 150 records. They confirm lynx prcscnce and allow an assessmenl of the situation. The first lynx are believed 10 have immigrated to the northcrn part of the study area from Austria. An increase and the distribution in the signs of presence show a south-westerly cxpansion. The trend in the Juliaii Alps and Prc-Alps is also increasing. Some inter- pretations of the status of this new population are made. The study area is in Ihe far south-easteni Alps. This area is important as a corridor between the Alps and thc Balkans, where a reintroduced lynx population exists which would be able to support the Alpine lynx population through dispers- ing lynx. Key-worcls: Lyxu Ixm,status. dislribution, prey, Tarvisiano, Italy, Alps

INTRODUCTION STUDY AREA The lynx (Lynx lynx) probably disappeared The study area is located in the Eastern Alps in the middle of the 19th century in the and includes an area of about 5000 km' eastern Alps (Toschi 1968). 11 came back at (Fig. 1). It is made up of the Alpine region the beginning of the 1990s (Molinari 199 1 ) and ils foothills (the Pre-Alps) in Friuli as a consequence of re-introduction pro- Venezia Giulia and the adjoining areas of jects in Austria (Festetics et al. 1980) and Veneto, Austria and Slovenia. It is divided Slovenia (Cop 1977). Because of the lack into two study sites, the first intensive and of specific research into the arrival of this the second extensive. predator, I started an inquiry in the Intensive study area: Conventionally Tarvisiano in 1986. This area is important called Tarvisiano, it is placed in the north- as a corridor between the Alps and the eastern part of the region of Friuli Venezia Balkans, and is a geographical key spot for Giulia (Fig. l), bordering on Austria spehes with long-distance migration or (Carinthiaj and Slovenia. It is an area of dispersal. It is in the centre of three areas about 500 kin' whose centre is the Tarvisio of importance to lynx. In two of them re- Forest. The main valley, Valcanale, is divid- introductions had taken place: in Austria in ed by two drainage systems; in the east by 1977 at a distance of SO kin from the the Slizza-Danube, in the west by the Fella- Tarvisiano, and in Slovenia in 1973, 120 Taglianiento. The geology, the land mor- kin away (Fig. 1). The third area is 140 kni phology, the slope gradient, and the expo- to the west, where a lynx population of un- sure from the west to the east of the main known origin exists (Ragni 1991). The pur- valley create a remarkable variety of forest pose of this study was to confirm lynx and landscape forms. The mountain chains presence, to collate the heterogeneous consist of the Julian Alps (altitude up to records and to estimate the number of in- 3000 in), Carnic Alps (2200 m) and the Car- dividuals. avanche (1800 m). The climate is a transi- 56 f! Molinari

,/7 Austria B

Figure 1. Location of the sttidy area and of lynx observations adjaccnt to the inlen5ive area. IYWY5. For details of the intcnsive area see Fig. 2. A is the centre of a lynx population of unknown origin. B and C are the sites of re-introductions in Austria and Slovenia, respectivcly. Shaded area: intensive study area. Points: ‘confirmed’ signs of presence. Circles: ’unconfirmed’ signs of presence. Bold lines are highways (e.g. A23). bold-dashed lines are international boundaries, and a dashed line represents a district boundary.

Lion between the continental and the occan- rejuvenation, and conserves this consider- ic rCgime. particularly favourable to a di- able biodiversity. The exploitation is lower verse vegetation. 65% of the area is covercd than the natural increase. Agriculture is lim- with woods (Di Bernardo et al. 1995), ited to the valley bottoms. Livestock-farm- which are in a state of preservation rarcly ing is not very common and confined to on- met in the whole Alpine arc. These woods ly a few areas. give way to grazing lands above the tiinber- Thc fauna is diverse with good sized popu- line in the Carnic Alps, and in the Julian Iations. The red deer (Ceriws clnphus) is Alps they rise up to the base of the rocky prcscnt with about 2000 individuals, the roe cliffs. Extended like a mosaic through the dccr (C‘crpiz.olus ccrpreolusj with 2500-3000 whole forest are pastures (13%)).As a con- individuals, and the chamois (Rupicapra sequeiicc the area has an excellent biodiver- mpi~~qmr)with about 1200- 1500 individu- sity. Sylviculture has a long tradition and the als (which is the historical minimum, as the system adopted can be defined as ‘group chamois at present is strongly affected by a shelterwood cutting’. This favours the main- Saimptcs scuhiei epidetnic). The ibex tenance of a permanent forest and natural (Cqn-uihcs) is represented by a small population of 130 individuals, and the wild-boar accurate study of their population sizes. The ) appears only sporadically. ibex, the moufflon (Ol'is nii4s7inoii) and the Brown hare (Lepus eui-opaeiis) and alpine fallow-deer (Dun70 dunra) are also present hare (Lep7is fimidirs vai.rorii.7) arc present in in small nuclei. small numbers. The avifauna is rich. with a good number of grouse. All the common Alpine species of rodents and small carni- METHODS vorcs occur. The presence of the brown bear 111 19x6 1 started to collect records of lynx (Ursus arcm.~),the golden eagle (Ayrrila observations. At that time the study area was chrisuetos), the griffon-vulture (Gyx fir/- restricted to For-cstu di 7ur-iisio (300 km'). IW), thc appearance, even if sporadically. of In addition to collecting records of lynx ob- the bearded-vulture (Gyps harhatus); and servations. I started a preliminary survey by thc now permanent presence of the lynx means of transects based on a cartographi- show how the food chains at present con- cal analysis of the area of highest lynx pres- tinue throughout. ence recorded as well as potential lynx habi- Extensive study area: This zone includes tat. 1 always followed trails because a much 4500 km' of mountainous and hilly areas greater distancc could be checked per day, (Fig. 1). In addition to the mountain ranges and I observed that lynx followed trails as mentioned before, it includes the Carnic well. On these transects 1 recordcd all signs Pre-Alps, the Julian Pre-Alps. and in the of presence (see Table 1). To find further south-east the Karst. These mountainous signs of prcscnce I added explorative excur- zones descend into high lowlands, formed sions into areas where othcr people had re- by rough river drifts. ported lynx signs. The region's climate is cxtremely varied, In 19x8-89. second-hand observations fi-on1 presenting typically inediten-anem aspects the Julian and Carnic Pre-Alps. at a distance (south-east). that become more and more of 25-69 km from the Tarvisiano, started to Alpine to the north. With the declinc of the increase and made an expansion of the ini- sea's influence, the precipitation level in- tial study area necessary. As a consequence, creases. The vegetation is variable as a con- the whole study area had to be divided into sequence of the climate. going from typi- an intensive (SO0 km') and an extensive cally mediterranean characteristics of the (4500 kin') study area. Karst (pine-woods and oak groves) to the Trail transects were periodically checked alpine zones with mountain forests (spruce- every IS days and the explorative excur- beech woods, larch-woods) through bands sions were carried out even more often. At of hornbeam woods, beech-woods and the beginning, most field work was concen- chestnut woods. There is a high degree of trated in the non-winter months. Later this cover and forest continuity. At present, a practice was altered so that by the end of process of considerable hunian depopula- 1991, at least 60% of the field work was tion is taking place, so that as time passes carried out in winter, thus taking advantage nature is allowed to regain precious habitat ol' the snow. Because of the size of the study for Fauna. area. cooperation with foresters. game-war- The fauna is represented by the same dens, hunters and amatcur naturalists be- species listed for the intensive study area. came necessary. All people involvcd had at- Ungulates important to lynx diet are abun- tended training sessions enabling them to dant in this region. although with hctcroge- identify signs ol' lynx presence. From 1988- ncous densities. Roe deer, red deer, chamois 92, a Bavarian hound (Bayrischer Gebirgs- and wild-boar are common. but there is no schweisshund) was used; this breed is par- 58 P. Molinari

Table 1. The number and kind of all confirmed signs of lynx presence found in the intensive study area. Search effort was also recorded, and a standardized index of observations per 100 days was calculated.

Year Direct ob- Tracks Vocali- Excre- Kills Total Search Index servations ration ment effort (days)

1986 23 2 70 2.8 1987 I I 5 156 3.2 1988 6 I I 8 188 4.2 1989 1 8 2 2 13 I98 6.5 1990 10 1 1 12 212 5.6 1991 7 -7 3 12 204 5.8 1992 3 8 3 14 1x0 7.7 1993 5 1 1 7 132 5.3 IYY4 1 4 1 4 1 0 124 X.0 1995 4 5 9 I08 8.3

Total 5 51 I 8 21 92 1572 5.9 ticularly effective in searching for carcasses whereas all records were used in the exten- and potential lynx kills, especially in situa- sive study area. An annual observation in- tions where humans find tracking difficult, dex was calculated as the number of signs e.g. in the absence of snow. of lynx presence per 100 days of search ef- The signs of presence were divided into the fort. categories ‘colzfil-tncd’ and ‘uticonfitmed’ , confir-nwd meaning that they were found by myself or collaborators, or were sccond- hand obscrvations that had been double- RESLLrS checked. Second-hand observations not con- A total of 150 signs of lynx presence were firmed were put into the category i.4mv17- collected up to May 1995, 42 of them in the ji’rmrd. The category confi’med was further extensive and 108 in the intensive study divided into ‘cerfairi’ and ‘pr-ohahle’. area. In the extensive study area 82%’of the From 1988 to 1990 I collected only second- signs of presence fall in the reliability cate- hand records of lynx presence from the ex- gory unconfirmed (Fig. 1). To give a more tensive study area. A series of investigations confident picture of the situation, in the in- started in 1990; it5 aim was to collect more tensive study area the records belonging to dctailcd information about lynx presence the reliability category uncorzfil-nzrd (1 4% of and also about suitablc potential habitats. all records) wcrc cxcludcd from the analy- Cooperation was established during this sis. The number of signs of presence found phase with additional people, but they were in the intensive study area are shown in not as highly trained as the staff of the in- Table I. The first lynx observation in the Ju- tensive study area. Promising reports were lian Alps was made as early as 1979 (Fig. checked by the author. The data analysis 2a). The next record dated only from 1986 was also different for the intensive and ex- (Fig. 2a). In 1989 a series of pictures of a tensive study areas. In the intensive study lynx chasing a marmot was taken (C. area, only confirmed records were analysed Vuerich, pers. comm.) and thus confirmed Figure 2. Confirmed signs of lynx presence in the intensive study area; (a) 1986-X9 (star: first and isolated record in 1979); (b) 1990-92; (c) 1993 - May 1995. Lines: rivers; broken line: state boundaries; bold line: highway A23: Iriangles: major mountain peaks (with altitude). 60 I? Molinari

Table 2. The number and species of kills found per ycar in thc intensive study area.

Year Roe deer Chamois Rcd dccr Hare Marmot Capercaillie Sheep Total

1987 1 1 1988 1 1

1989 I I i7 I990 1 1 1991 1 1 1 3 1992 1 2 3 1993 1 1 1994 1 3 4 1995 I 4 5

Tolal 6 2 8 1 1 1 7 71

lynx presence. From this period to 1991, probably become more important in the signs of presence increased (Table I), but nineties (Table 2). Seven out of eight red only in the Carnic Alps (Fig. 2a). deer kills were found only in the past two In the early 1990s, the signs of presence in years; 88% of the kills in this period were the south-west of the intensive study area red deer. increased. In 1992 the first records were found south of the highway in the Julian Some anecdotal observations allow conclu- Alps and the Caravanche (Fig. 2b). In thc sions about the number of lynx in the inten- period from 1993 to May 1995 the signs of sive study area: in spring 1994 in the slope presence south of the highway and in the of the Carnic Alps of the Tarvisio Forest. it south-west increased further (Fig. 2c).These was possible to hear two distinct individual trends existed in the extensive study area, lynx during the mating season (confirmed al- too. The south-westerly expansion already so by tracks). In the meantime, tracks were noticed in the inteiisive study area was also also found in a different area of the Julian observed in the extensive one. In 1993 the Alps. confirming that at least three individu- first sign of presence was found in the re- als were present in the intensive study area. gion of Veneto and a remarkable number of records collected during the last two years came from the Julian Alps and Pre-Alps. DISC US S 1 ON Correlating the search effort pcr year with The surveys were very variable because of the collected signs of presence according to the different search effort each year, the time, we can scc a slightly increasing trend presence or absence of a dog, and the ex- (Fig. 3). Despite a reduced search effort in tremely varying weather conditions (snow). recent years, the number of signs of pres- Together with the searcher's increasing ex- ence that were found remained high. This perience, this had an influence on finding trend also seemed to apply for the extensive signs of presence. Nevertheless the first aim study area. but was not shown here in detail of this study - to confirm lynx presence - because of weak data. has been fulfilled. The number of kills found has increased The first record of a lynx observation in since 1991: the predation on red deer has 1979 in the Julian Alps remained isolated 10

0 85 86 87 88 89 90 gl 92 93 94 96 96 Year

Figure 3. Observation index (ratio of signs of lynx presencc pcr 100 days of search effort; Table 1) per year. The regression (y = -44.93 + 0.56 x: R = 0.875, P = 0.001) indicates an increasing trend of lynx presence in time for years. The origin of this animal is not in Slovenia, or from the northern population known. The highway A23 (Udine-Villach) already established in the Carnic Alps. I had not been built at that time. and this lynx postulate that these individuals immigrated may have emigrated from the population rc- from Slovenia because the highway separat- introduced in Slovenia in 1973 (Cop 1977) ing the Carnic from the Julian Alps may rep- or perhaps from the one re-introduced in resent barrier. Observations from the neigh- Austria in 1977 (Festetics et al. 1980). bouring Slovenian tcrrilory also support this In the 1990s, lynx are believed to have im- interpretation (J. Cop. pers. comm.). migrated to the northern part of the inten- The increase and the location of the signs of sive study area from Austria (Fig. 2a). Data presence that extend westwards to thc from the neighbouring regions support this Dolomites as well as southwards to the Car- hypothesis. In Austrian Carinthia, lynx pres- nic Pre-Alps (Fig. 1) indicate a south-west- ence had been confirmed (Huber IWS), erly expansion. In the Julian Alps and Pre- whereas in the south in the Julian Alps and Alps the trend is increasing, too. The conti- also in the adjacent Slovenian region. no ev- nuity of suitable habitat without a biogeo- idence had been found by then. The first graphical barrier between the Alpine regions signs of lynx presence south of the highway of Italy and Slovenia suggcst increasing im- originated from 1992. It is difficult to judge migration from the high-density population whether these individuals immigrated from in Slovenia (J.Cop, pers. comm.). This spec- the south, from the re-introduced population ulation is further supported by the fact that 62 P. Molinari hunting pressure has decreased in Slovenia 1993). The use of radiotelemetry could re- (in particular near the borders) since 1993 veal more about the spatial organisation of (Cop and Frkovic 1996). this secretive cat and enable the finding of The interpretation of the expansion of the more kills in our region. area occupied by lynx raises many ques- Attitudes towards the arrival of the lynx tions. Did individual lynx increase their vary among the people of the region. Italian home ranges or did the number of individu- hunting officials remained rather indifferent. als increase'? Did the lynx in the study area they simply noted the return of this predator reproduce (as yet there is no evidence) or without taking a position. They seem to ac- were there always new immigrations'? Are cept a 'natural' return, and the lynx has been we still facing a colonisation phase (hence well protected by law. Unfortunately. local with its own dynamics) or a thriving popu- hunters consider lynx a rival and are in lation? To answer these questions we need favour of hunting it. Livestock predation has to know the spatial distribution or density of fortunately been low so far: and compensa- lynx, their age structure and sex ratio. Brei- tion will be paid for losses. Another prob- tenmoser and Haller ( 1993) postulated that lem that has to be mentioned here is habitat in the first phase of lynx expansion, a high fragmentation. The Julian Alps are separat- number of predators would feed on a dense ed from the Carnic Alps by both a major population of nai've prey that were easy to highway (A23) and a railway line running kill. This high predation results in decreased parallel. Therefore. it is important to con- prey availability and induces a sccond scrvc thc last remaining corridors. phase, with predators hunting over a larger Large contiguous habitat and its conscrva- area and consequently having larger home tion remain a key factor for lynx survival. ranges. Before expanding their home The re-introduced Slovenian population ranges, lynx included unusual prey such as could support both the threatened au- domestic sheep into their diet, due to the re- tochthonous population in the south-western duced availability of roe deer and chamois. Balkans and the feeble population in the The decrease in observations in the inten- Alps. This underlines the importance of the sive study area on the one hand, as well as Tarvisio region as a corridor between the the surprisingly high predation on red deer. Balkans and the Alps, as already shown for since 1994 indicate that in the intensive the brown bear (Schroder 1992, Molinari study area the lynx could already be in the 1994. Kusak and Gutleb, unpiibl.). second phase. The small number of sheep and the high number of red deer in the intensive study area would explain why the ACKNOWLEDGEMENTS. Several people have con- lynx switched to red deer. A temporary in- tributed to (his study. Thanks to C. Vuerich lor crease in predation on red deer was also ob- the collaboration with field work and personal served in Austria some years after the re-in- niatcrial put at disposition, to T. Huber and J. troduction of lynx (Gossow and Honsig-Er- Cop for the long-lasting and important cross- lenburg 1986). Another reason why lynx border cooperation, and to L. Autischer, L. Boi- switched to red deer might be the reduced tani. M. Bottazzo, G. Battistutti. A. Buzzi. R. availability ol' roe deer and because the best De Battisti, P. De Martin, S. Di Bernardo. T. alternative prey, the chamois, suffered a dc- Ferro. H. Haring, L. Masutti, F. Perco, B. Rag- crease through Sarroprrs scahiri. There is ni, S. Roy, 11. Schnabl, M. Tosoni, R. Venturi- only little evidence for both hypotheses so no. Special lhanks also to the collaborating in- far. To get an idea of lynx density in a cer- stilu(ions: Istilulo tli Zoologia Foreslale - LJni- tain area, the spatial oganisation of individ- versiti di Padova. Corpo Forestale dello Stalo- uals has to be known (Breitenmoser et al. Foresta di Tarvisio, Servizio Caccia e Pesca della Regione Friuli V.G., Osservatorio Faunistico conservazione della steysa. distribuzione e di Pordenone, Swiss Lynx Project, Wildbiolo- consistenza degli ungulati), ci Sono buone gische Gesellschaft Miinchen. and also to A. ragioni per credere che la lince sia destina- Jobin for valuable coinmcnts and the writing of ta ad aumentare in un prossimo futuro. the manuscript. Thanks especially to Ch. and LJ. Lindagine 6 tutt’ora in corso. vi C I’inten- Breilenmoser for the important advice on zione di approfondirla. nxthodology during the recent years of research and their comments on an earlier version of this paper. REFERENCES

Breitenmoser, U. and Haller, H.. 1993. Pat- RIASSUNTO terns of Predation by Reintroduced Euro- Nci primi anni 1980 la lince ricornpare nelle pean Lynx in the Swiss Alps. J. Wildl. estreme Alpi sud orientali italiane, Regione Manage., 57( 1): 135- 144. Friuli Venezia Giulia (Fig. 1). La situazione Breitenmoscr, U.. Kaczensky; P., Dotterer, inedita e la mancanza di ricerche specifiche M., Breitenmoser-Wursten, Ch., Capt, S., hanno indotto, nel 1986, ad avviare il pre- Bernhart, F. and Liberek, M., 1993. Spa- sentc lavoro. A partire dal 1988 C stata cre- tial organization and recruitment of lynx ata un‘area di studio intensivo (500 km2. (Lytzu 1yn.x) in a re-introduced population Fig. 2a), convenzionalmente chiainata in the Swiss Jura Mountains. J. Zool., Tarvisiano, ed una di studio estensivo (4500 Lond.. 231: 1-15 km’). Sono stati individuati una serie di sen- Cop, J.. 1977. Die Ansiedlung des Luchses tieri che vengono percorsi a scadenze rego- (Lynx /yz.x-) in Slowenien. Z. Jagdwiss.. lari in tutte e quattro le stagioni, con mag- 23( 1): 30-40. gior frequenza nel periodo di innewmento. Di Bemardo, S., Terlicher, P. and Travan, L., Vengono rilevate orme e piste su nevc e ter- 1995. Forstsystem im Kanaltal - Intema- reno plastico. residui di attiviti varie, tionales Forstsyrnposiuin 1995. Karntner percezione di richianii sonori ed avuista- Forstverein InTormation. 26: 60-62. menti. Fcsletics, A., Von Berg, F.-C. and Sonimer- Fino ad ora Sono stati raccolti complessiva- latte, M., 1980. Die Wiedereinburgerung mente 150 scgni di presenza (Tab. 1) che des Luchses in Osterreich. In: Festetics, hanno consentito di confcrmare la presenza A. (ed), Der Luchs in Europa. Kilda Ver- e di elaborarc un quadro generale dclla situ- lag, Greven, pp. 268-284. azione del felide. Si ritiene che le prime lin- Gossow, H. and Honsig-Erlenburg, P.. 1986. ci siano immigrate a nord dell’area di studio Management problems with re-intro- dall’Austria (Fig. 221). L‘espansione dei scg- duced lynx in Austria. In: S.D. Miller and ni di presenza mostra un’espansione verso D.D. Everett (eds) Cats of the World: Bi- sud ovest (Fig. 2b, 2c). A partire dal 1992 ology, Conservation. and Management. compaiono le prime linci nelle Alpi Giulie, National Wildlife Federation, Washing- a sud dell’autostrada. L‘indice di osser- ton, D.C.. pp. 77-84. vazioni nel tempo (numero segni di presen- Huber, T., 1995. Neucs vom Luchs - Einc za in relazione allo sforzo di campionamen- Fahrtensuche. Natur und Land 81(1): 28- to * 100 giorni, Fig. 3), mostra inoltre un in- 3s. cremento della presenza di lincc. Quest‘area Molinari, P., 1991. La Lince nel e un importante corridoio faunistico tra le Tarvisiano (Alpi Sud Orientali). Suppl. Alpi e i Balcani. Considerata la situazione Kic. Biol. Selvaggina, Bologna, XJX: ambientalc dell’area. (morfologia e stato di 589-594. 64 P. Molinari

Molinari, P., 1994. Gestione e Conscrvazione Schroder. W., 1992. Biirenxchutz in den dei Grandi Mammiferi Predatori. Pianifi- Alpen. Wildbiologische Gesellschaft cazione e Gestione delle Risorse Faunis- Miinchcn, Oberamniergau, 12pp. tiche - Ministero dell’Arnbiente - Azienda Toschi, A.. 1968. Rapport sur la Disparition Regionale Foreste Veneto. pp. 103-1 16. du Lynx en Italie. In: Kratochvil et al.: Ragni, B., 1991. I1 Ritorno della Lince. Le History of the lynx in Europe. Acta Sc. Scienzc, 273: 54-63. Nat. Brno, 2(4): 17-24.