Brown Bear in Greece: Distribution,Present Status-Ecology of a Northernpindus Subpopulation

BROWN BEAR IN GREECE: DISTRIBUTION,PRESENT STATUS-ECOLOGY OF A NORTHERNPINDUS SUBPOPULATION

GEORGE A. MERTZANIS, Hellenic Society for the Protection of Nature 24, Nikiw Street, 10557, Athens, Greece

Abstract: Brown bear distributionrange in Greece comprises 2 distinct nuclei of unequal size, covering a total of about 11,000 km2, and seems to have stabilized for the last 20 years after dramaticregression in the 19th century. Extra-limitalsporadic occurrence of bears southwardsof the western population nucleus down to the 39th parallel, as well as unexplored sectors of potentialbear occurrence in the northernparts of the country, may add new data to the species chorology in Greece. There is a risk of further internal fragmentationof the western distribution nucleus. Human-causedmortality appears to be the main factor of populations' negative trends. Brown bear food habits were determined by investigations in a 900 km2 bear area located in the northernPindus range and scat analysis (N = 343). Only plant materialwas found in 77 % of the samples, whereas 17% contained both plant and animal material, and 6% only animal material (mostly insects-ants). Omnivory and opportunisticstrategy appearedas the main characteristicsof bears' feeding behavior. Brown bear annualactivity cycle was determinedby data on signs of presence and activity (N = 664). It appears in relation to trophic optimas and mesoclimatic conditions of the habitat. There is evidence of winter inactivity. Brown bear habitatpreferences determined by Marcum& Loftsgaarden'smethod (N = 289 bear locations) show seasonal influence of types of vegetation communities on habitatuse. Bear-humaninteractions level seems critical: poaching and logging are the main causes of habitat deteriorationand populationdecrease.

Int. Conf. Bear Res. and Manage. 9(1):187-197

Elements on the historical distributionof the brown Ministryof Agricultureand the Hellenic Society for the bear in Greece are mainly based on historical reports, Protection of Nature (Greek Ministry of Agriculture notes, and available texts. Report 1988). According to this information, brown bear range in Results and figures presented in this paper come Greece appears to have suffered dramatic regression essentially from the latter sources. throughthe centuries, as in other parts of Europewhere We thank the Greek Ministry of Agriculture, the the species was or is still in conflict with human Hellenic Society for the Protection of Nature, and the expansion. In the late 1960s (1969), brown bear were Royal Instituteof NaturalSciences of Belgium for their legally regarded as a protected species in Greece. It is collaborationin the data availability. We also thank the actuallyconsidered as an endangered-vulnerablespecies Veterinary School of Toulouse-France, the National (Greek Red Data Book-Universityof Athens 1992). A Instituteof Researchin Agronomy of Toulouse, France, review of the late 20th century brown bear distribution and the University of Agronomy-Laboratory of patterns and population status in Greece gives various Pedology, Athens, Greece for their collaboration and estimates for the first 2 decades (Couturier 1954, sound technical assistance. Hainard 1961, Curry-Lindahl1972). Apart from previous authors' fragmentary information, lack of systematicdata on the distribution STUDYAREA and ecology of the brown bear in Greece before the The study area for distributionestimate included all early 1980s resulted in an urgent need for systematic northernparts of the Pindus and Peristeriranges as well status and ecology studies. as the Rhodopimountain complex. Divided into 4 main More up-to-date information(Matsakis et al. 1981, apparentphysiographic/phytogeographic units, this area Mertzanis 1989) gives a brown bear distributionpattern comprises (Fig. 1): of 2 distinct population nuclei located respectively in Unit I: The Peristeri range with alpine meadows, the Pindus range (northwestern Greece) and the largebeech (Fagion sylvaticae) forests between 1,200 Rhodopes mountaincomplex (northeasternGreece), and and 2,100 m, and oak forests on lower altitudes, all considers the southwesternnucleus as the southernmost covering mostly graniticsoils (Quezel 1967, Debazac range of the species in Europe. These papers also and Mavromatis 1971). contain some preliminary information on bears' food Unit II: Large parts of northern Pindus range, habits. includingthe valleys of Aliakmon, Sarantaporos,and In 1985 I started 31/2years of researchwork, in the Aoos rivers with alpine meadows, large black pine framework of a doctoral thesis (Mertzanis 1992), in forests (Pinus nigra ssp. Pallasiana) as well as mixed which were incorporated data and results of the forests with black pine (Pinus nigra), fir (Abies preliminary stage of an EEC-funded survey project, borisii regis), beech (Fagus sylvatica), and white pine conducted in 1988 by the Royal Institute of Natural (Pinus heldreichiii) covering mostly limestone and Sciences of Belgium in collaboration with the Greek ophiolithic soils. At lower altitudes the vegetation 188 Int. Conf.Bear Res. and Manage.9(1) 1994

- I /, . J

Macedoine

r- I f/F

IV Thessalie

? Regularpresence N o 0 13.5 Km + Sporadicoccurance I . .

4- Transbordermigration phenomena

.- --. Studyarea

Fig. 1. Brown bear distribution range in Greece-western population nucleus.

zones of Quercion-frainettoand Ostryo-Carpinionare Climate of the area is temperate with cold winters. widely present. Annual precipitationranges from 1,270 to 1,450 mm Unit III: The river Acheloos high valleys with and vegetation zones comprise: oak forests with mainly large fir (Abies borisii regis) forests covering (Quercusfrainetto, Quercuspubescens, Quercus cerris, limestone soils and oak forests on lower altitudes. Ostryacarpinifolia, Carpinusorientalis) pure and mixed Unit IV: Rhodopi mountaincomplex: a vast granitic coniferous and deciduous forests with: black pine (Pinus mountain complex covered with large forests of nigra), beech (Fagus sylvatica), fir (Abies borisii regis), spruce (Picea excelsa), forest pine (Pinus sylvestris), and white pine (Pinus heldreichii) followed at the higher beech (Fagus orlentalis, Fagus mosaecus), and oak elevations by alpine meadows. (Quercusfrainetto, Quercus macedonica). The composition (percentage of occurrence) of the The study area for the ecological analysis extends main types of forest vegetation is: over a 900 km2 zone in northwestern Pindus with Oak forests 27% elevations ranging between 550 and 2,637 m. About Pine forests 27% 40% of the area presents a denivelation of 500 m per Beech forests 10% km2, and 7% of it is above 1,900 m of altitude. Fir forests 5% BROWNBEAR IN GREECE* Mertzanis 189

Table 1. Percent composition of the 11 vegetation 1989) only during1988. They were often backedby communities in the study area and bear use. second-handinformation of the same year. Cases of bearsightings located in 2 neighboringareas or regions Observeduse Expecteduse countedonce. An of Vegetationcommunities (n = 289) (n = 840) were only approach population trendsis attemptedthrough harvest data as well as cases 1. Oak forests 0.23 (66) 0.22 (185) of illegalkilling. 2. Other deciduous forest types 0.03 (10) 0.03 (21) 3. Orchardsand cultivated land 0.12 (35) 0.04 (35) Ecology Food habits. Collections of scats in the field 4. Black pine forests 0.18 (52) 0.22 (184) (N = 343) were used to determinethe quantitative 5. Fir forests 0.01 0.01 (4) (4) importanceof food itemsused by brownbears. Partof 6. Beech Forests 0.09 (25) 0.04 (32) the sample(n = 95) was analyzedin the laboratory, 7. White pine forests 0.01 (4) 0.02 (15) whereasthe otherpart (n = 248) was macroscopically in the field. scat contentswere 8. Mixed Forests 0.06 (18) 0.09 (77) analysed Generally, moreeasily identifiable when freshin the field. In the 9. Mixed forests and ecotones 0.10 (30) 0.03 (25) secondcase identificationof plantremains was easier 10. Ecotones 0.08 (24) 0.14 (117) when comparedwith nearby specimens. In several cases materials in the field were takento the 11. Subalpine and alpine pastures 0 07 (21) 0.19 (145) analyzed laboratoryfor furtherstudy. All scatswere individually Chi-square test demonstratedthat there was a highly significant identifiedaccording to location, altitude, vegetative difference between the expected utilization of the vegetation surroundings,and age. Analysisof bear scats in the communities and the observed frequency of use in the study area laboratoryfollowed the techniquesof Tisch (1961), df = P < (Chi-sq. test, 10; 0.001). Russell (1971), Sumnerand (1973), and Consequently, Bonferronisimultaneous confidence intervalswere Craighead used to determinewhich categories of habitattypes were utilized more Faliuet al. (1980). Basicsteps involved: or less than expected within an annual cycle. - rehydrationof fecal material - separationof materialinto homogeneous groups by Whitepine forests 6% use of screens(4.0, 2.0, 1.0, 0.5 mm mesh) Mixedforests 11% - identificationof contents Alpinemeadows 14% - recordingof identifiedmaterials. Identificationof species, throughmacroscopic and METHODS microscopic examination, was usually successful, presentingsome difficulties for grass. Animalmaterials Distributionand PopulationStatus weremainly identified through micro-techniques of hair Estimateof brown bear historicaldistribution in examinationusing reference collections as well as Greece was based on data from relevantliterature. throughexamination of bone remains. The occurrence Recent data, based on signs of bear activity, bear of each identifiedfood item was recordedas each scat sightings,or illegal killing, were gatheredthrough 2 was analysed. Visual estimatesof occurrencewere investigationcampaigns carried out in 1986 and 1988 recordedunder 5 categories. by meansof questionnairessent to localpopulations or 0: trace-5% filledduring interviews. Answered questionnaires were 1: 5-25% dividedinto 3 evaluationcategories according to quality 2: 25-50% and reliabilityof informationprovided: 3: 50-75% - direct information:checked by interviewand/or 4: 75-100% field inspection. Underevaluationof the use of some foods had to be - second-hand information: checked only by takeninto account because of differentialdigestibility of interview often coming from several different somefood items,Mealey (1980). Percentfrequency of sources. occurrenceand percent of diet for each food itemwere - uncertain: reliability of these reports remains calculated. doubtful. AnnualActivity and Habitat Use.-The 3 /2-yearfield The "probableminimum population size" estimate investigationwas based on the indirect method of was mainly based on direct informationconcerning systematicrecording of all observationsof bearsigns of sightingsof females with cubs-of-the-year(Servheen presenceand activity in the study area (N = 664). 190 Int. Conf.Bear Res. and Manage.9(1) 1994

Table 2. Utilization of the different vegetation communities by the brown bear within the study area.

Vegetationcommunities Bonferroniconfidence intervals Choiceof bears

1. Oak forests -0.06; +0.06 Proportionaluse 2. Other deciduous forests -0.04; +0.021 Proportionaluse 3. Orchards- Cultivated lands -0.133; -0.26 Preference

4. Black pine forests -0.03; 0.11 Proportionaluse 5. Fir forests -0.03; +0.01 Proportionaluse 6. Beech Forests -0.095; -0.002 Preference

7. White pine forests -0.017; +0.025 Proportionaluse 8. Mixed forests -0.016; +0.074 Proportionaluse 9. Mixed forests and ecotones -0.123; -0.025 Preference

10. Ecotones +0.005; +0.11 Avoidance

11. Subalpine and alpine pastures +0.05; +0.15 Avoidance

As we may see from the above, bears in our study area demonstratea strong preference for forested habitatswith a significant sub-use of open lands. Some forest or vegetation types such as: beech forests, mixed forests with ecotones, and orchardswith cultivated lands appear to be utilized in greater proportionthan their availability. This could be attributedto their seasonally high value as food resources. More specifically, significant preference for phytocenotic combinations comprising mixed forests with ecotones (Abieto-Fagetum, Fageto-Abietum, Pineto-Fagetum, Abieto-Pinetum)is probably related to the optimum ecological diversity of this type of forest ecosystem, satisfying ecological requirementsof the bears such as available food and good shelter.

Because of field ruggedness and heterogeneity, 86 RESULTS "piecewise"(nonlinear) type transects(Anderson et al. 1980) were chosen for field sampling, covering all Distributionand PopulationStatus major ecological units of the study area. Their length Historic sources as well as 105 localities with names (max length = 20 km; min length = 2 km; average containing the "bear"radical distributed all over the length = 7.7 km) and density were adjusted to the country, inform us about a possible brown bear ecological diversity of the investigatedsectors. The 3 distributionrange in Greece, in ancienttimes, extending major categories of sampling transects and their over almost the entire continental part of the country cumulatedlength were: (Xenophon, Pausanias). Bear presence in the - ancient or still in use paths and trails: 235.5 km mountainousparts of Greece including the peninsula of - forest roads (of various use levels): 242.3 km Peloponnisos seems to have been continuous in time - other transects independent of human use: until the 15th-16th centuries (Pizzicoli, Candiloros, 195.2 km Guillet, Dedreux in Simopoulos 1984). In the 18th Total length: 673.0 km century, there is evidence of the beginning of a With a totally walked distance of 2,350 km, dramaticpopulation decline (Mertzios in Papavassiliou interannual and interseasonal comparison of walked 1963) due essentially to massive bear exterminationfor distances shows satisfactory sampling pressure and the skin, coupled with habitat alteration. More recent distribution in time (year, season, and month) oral informationconfirms the species' extinction in the (F1 = 0.93; F2 = 0.82; a = 0.05). Data were 1940s from the southernmostand easternmostbranches cumulatedand partitionedby time of season. of the Pindus range (Mt. Parnassos, 2,457 m, and Mt. Habitat preferences and habitat-use patterns were Olympos, 2,918 m). One may assume that the main estimated according to the "Habitatuse-availability" orographicunits of Greece (Pindus range and Rhodope model (Marcumand Loftsgaarden1980). Habitattypes mountains), because of their inaccessibility and were determined and classified according to spatial remoteness,have been the refuges and dispersal centers distributionand availability of vegetation communities of the species in Greece throughouthistoric times. in the study area. Bonferroni's intervals were calculated to estimate significance of bear habitat Present Status preferencesby time of year (annualcycle) and seasons. During the 1986 investigation campaign, 160 BROWNBEAR IN GREECE * Mertzanis 191

questionnaireswere sent, to all villageslocated within - sporadicoccurrence of bears in the far eastern the presumedbear range. One hundredand two were sectors(area of Mt. Askion)was confirmedas well answered, confirmed, and mapped. That gives a as locally extra-limitalbear presence in the far distributionpattern of the brownbear for the western westernones (areaof Pogoni)(in 1990). population nucleus, covering the 3 - fromspring 1987 to autumn1989 severalcases of physiographic/phytogeographicunits previously extra-limitalsporadic occurrence of bears in the described(Fig. 1.) Accordingto this information,we extremesouthern sectors (area of Karpenission), noticethe followingpoints: were reportedand confirmedin the field (location - a concentrationof bear occurrencearound the of damageon beehives). In this zone, apparently majormountain units of the northernPindus range. isolated from the rest of the core bear range, - a geographiccontinuity with adjacent (neighboring) probably1 male, 1 subadult,and 1 femalewith a populations(Albanian, Yugoslavian) characterized cub were recognized. This is the firstevidence of by migrationphenomena along the borderline. bearoccurrence in this areafor the last 40 years. - a severefragmentation tendency of the distribution - thereare still someunexplored sectors of potential area nearbythe town of Kastoria(see arrowon bear occurrencein the extremenorth that need Fig. 1). This could be explainedby the ongoing confirmation(area of Mt. Voras). extension of agriculturallands upon forests (in - the risk of internalfragmentation at the same "discussion"topic). latitudestill persists. - a sporadicbear presence in the extremeeastern and Concerningthe eastern distribution nucleus (Rhodope southernparts of the range, characterizedin the mountains): samesectors by a retreatof the speciesrange. - sporadic occurrenceof bears in the extreme A total of 1,126 reportswere registeredduring the westernsectors (area of Serres- Lailias)as well as 1988 campaign. Of these, 1,013 (90%)were accepted probabilitiesof bearoccurrence in adjacentsectors as referringto the periodbetween 1983 and 1988, and (Mt. Belles) makes the delineationof the bear 113 (10%) were rejected. Of the 1,013 accepted rangethere more fluctuating. reports, 506 (50%) contain direct and second-hand - migrationphenomena across the Greek-Bulgarian informationreferring only to the year 1988. In total borderhave also been confirmed. 479 persons were interviewedand distributedas follows: PopulationSize shepherds 26% Estimationof minimumpopulation size is basedon forestersand forestwardens 24% recordsof observationsof females with cubs of the hunters 15% year, withinthe sameyear (Servheen1989). timbers 10% A minimumsize of the total bear populationin farmers 8 % Greece,in 1988, was roughlyestimated between 120 beekeepers 4% and 140 individuals. Systematicdata on illegal hikers 3% killing-poachingcovering a 5-yearperiod (from 1985 to other 10% 1990) shows an averageannual loss of 14 animals Five categories of signs of bear activity and (about12% of the totalpopulation) (Fig. 3). Besides, occurrencewere takeninto account,as follows: accordingto same data on a longer period (1962 to - directsightings of femaleswith cubs 1978), "harvest"levels reflectingpopulation levels - tracks,scats, and othersigns of activity (Elgmork1988) appearto be decreasing(Fig. 4). - damage to livestock Since the official protectionstatus of the species - cases of illegal killing remainshardly effective, there are seriousreasons to - confirmedcases of illegalkilling believethat human-caused mortality is still one of the Their distributionpattern (Fig. 2), extendingover main factors of bear populationsnegative trends in 8,000 km2, appears generally the same as reported in Greece. We may also noticethat high levels of human 1986 for the westernpopulation nucleus. They also causedmortality are particularly concentrated during the providenew dataon speciesdistribution for the eastern wild boarhunting period (Fig. 5). nucleus(Rhodope mountains), extending over a surface of about2,500 km2. Food Habits More specifically: Concerning the western The global(annual) diet is dominatedby food items distributionnucleus (Pindus range): of plant origin (87%) followed by animal material 192 Int. Conf.Bear Res. and Manage.9(1) 1994

/ Bulgaria-- yr ,--ej2"n ;-c

9'?- - Rhodope, 9 R T' yrkia * " -- 9 1T

Aegean Sea

IonianSea a zK '

Potentialbear sectors co a

Bear range

* Area of Study

0 50 100 150km ,I

: , j

Fig. 2. Brown bear range in Greece according to 1988-89 data.

(13%) mostly composed by insects (ants: 10%). Grass constituted the major part in summer (30% of the (herbaceousplants), fleshy and dry fruits composed the summer diet, n = 73) showing a maximum in autumn major proportion of the vegetable part in the bear diet. (61% of the autumn diet, n = 209). We noticed a In total, 54 species were identified, meeting nutritional differential (off season) use of certain vegetable food needs in proteins and glucids respectively (Mealey resources such as acorns (70% of occurrence in the 1980) (Fig. 6). We noticed that the general feeding winter sample), beech nuts (45 %), and apples and pears cycle appeared to follow plant phenology. We also (15%). When these were consumed in winter, they noticed that there seems to be a close relationship seemed to have a compensating role to the feeding between rhythm of consumption (massive) of several stress characterizingthis period of the year. Animal food items and their seasonal availability(short in time) material consisted primarily of ants (maximum and therefore optimal nutritional value, within their consumption in summer, 43% of occurrence in the annual phenology stages (cycle) (Fig. 7). Herbaceous summer sample--10% of the total diet). Percentage of plants constituted the principal spring food (60% of other mammalpreys is apparentlylow (2% of the total occurrence in the spring sample) whereas fleshy fruits diet) (taking into account high animal protein BROWNBEAR IN GREECE* Mertzanis 193

Number 10- of bears 8 killed

4- 7

.,P.1 I11 ,,I I,,, oI,62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 1985 1986 1987 1988 1989 Year Year Fig. 4. Brown bear human-caused mortality and tendency of levels to data from 1962 to Fig. 3. Confirmed cases of annual bear losses (human-caused population (according hunting 1978). mortality) from 1985 to 1989. digestibility, Herrero 1978, Mealey 1980). - influence of climatic factors (high temperatures) Micrographicstudy of hair samples showed dominance inducing "stational"modes in habitat use patterns of domestic ungulates (especially cattle, with the highest (Parde 1984). number of attacks concentrated in autumn) (Fig. 8.) The second maximum of activity occurs in fall, Remains of sheep, goats, domestic dog, and meles suggesting a close relationwith the pre-denningfeeding meles were also identified. Remains of tortoise were period. Bear activity did not cease entirely during identified in 1.5% of the total sample. winter (10% of total indices). We noted occasional Brown bear feeding spectrumshowed a significantly movements of bears as well as feeding activity signs. higher diet diversity (Student-Newman-Keuls test; We suggest 2 main causes of that activity: P < 0.05) in autumn, the period of fruit ripening and - Hunting that could be an important disturbance therefore of maximum food availability (Fig. 9); its factor explaining unusual winter mobility. main characteristics being continuity and - Intervals with mild weather conditions (usually complementaritywithin time. February) inducing a "re-activation" of some animals. Annual Activity and Habitat Use The brown bears' greatest activity levels occur Analysis of HabitatPreferences within a period of 6 months between May and October In the analysis of habitat selection, we investigated (almost 70% of total activity signs) with a first the relation between the utilization and availability of maximum in June (10% of total activity signs) and a habitattypes using "vegetationcommunities" variable. second one (more important) in September-October Different vegetation communities were grouped into (34% of total activity signs) (Fig. 10) These 11 types: Oak (Q. frainetto, Q. cerris, Q. pubescens) significantly differ from spring and winter activity forest; other deciduous forest types (Carpinus sp. levels (F = 6.09; P < 0.0005). Ostrya sp.); black pine (Pinus nigra) forest; fir (Abies Annual activity patternsdemonstrate 2 peaks (Fig. borisii-regis) forest; beech (Fagus sylvatica) forest; 10). One occurs in the beginning of summer(June) and balkanpine (Pinus heldreichii) forest; mixed (deciduous appearsto be relatedto the post-denningfeeding period and coniferous) forest communities; mixed forests and as well as to the breeding period (Clevenger et al. ecotones; ecotones; orchards and cultivated land; 1990). The mid-summer flexion (July) of annual subalpineand alpine pastures. activity could be attributedto 3 main causes: Habitatavailability was determinedby mapping the - significant human presence within bear habitat former vegetationcommunities (vegetation type grid = (logging, grazing, trekking), 1 km2)and producing840 coordinatesin the study area. - restrictedfood availability, Bear locations were also mapped. 194 Int. Conf. Bear Res. and Manage. 9(1) 1994

. 1 2

III A. F B. tI- t

SU seasons I SP

w [ ] I I I I I I 30 20 10 0 0 10 20 30 _ numberof bears killed I numberof bearskilled - huntingperiod

Fig. 5. Seasonal frequency of bear killing. (A) From 1962 to 1978. (B) From 1985 to 1989.

1 2

0 25 20 15 10 0 5 10 15 20 25 30 5 0 I I % /// / //// Rubussp. Rubussp. //////////Y//// % Cornusmos Herboce s //////;/j////l//////////I ///////// Herbacees Fo u rmis ////////////////////// Cornusmos ///,///// Fourmis Molus,Pyrus Molus, Pyrus Fogus sp. Fogus sp. Quercussp. Prunussp. Prunussp. Quercussp. Corylussp. Rosoc6es J. pousses Mammiferes Rosocees Corylussp. A. ~ Mammiferes Vitis sp. Vitis sp. J. pousses B. Reptiles Reptiles

Fig. 6. Absolute and relative occurrence of the main food items in the bear's diet, Pindus, Greece, from 1985 to 1988. BROWNBEAR IN GREECE* Mertzanis 195

100 - Reptiles

90 - Ants

80 - I Mammals

Accorns 70- B Ifl Cornussp. 60 - J- Coryllussp.

50- - Herbaceous

40 - Rubus sp. - 30- Young plants a Beech nuts 20 - a Prunus sp.

10 - - Malus sp.

0 H H P P P E E E A A A H

Fig. 7. Annual trophic spectrum of the brown bear in Pindus range, Greece, from 1985 to 1988.

With these locations and those from bear use we utilized more or less than expected. tested the "null hypothesis" that bears do not use the In fact, in 86% of the cases where this vegetation habitattypes selectively but ratherin proportionto their type occurs in our study area, understory vegetation availability or in a random manner. To determine with fruit trees and berries is well developed. Besides, whether there was a significant difference between the forest cover rate in 52% of sites of bear locations is expected and observed utilization, a chi-squaretest was above 75 %. used dependingon whetherthe assumptionof normality Global sub-use of ecotonic formations is probably was met. If a statisticallysignificant difference resulted, related to the significant avoidance of subalpine and Bonferronisimultaneous confidence intervalswere used alpine meadows and subsequentlyof the upper limit of to determine which categories of habitat types were the forested zones (interface between forest and open lands). 50?/ ~o- It is worth noting that use of black pine forests, 450//0- proportionalto their availability, which is considerable 40?//o- o official sources field observations within the study area, may be as well related to the 35?l/0- shrub 30O * reports occurring understory vegetation composed of which as we have seen a 250/Yo- essentially berries, play role in the bear's diet. 20?,vo- very important 150? 10?07o- CONCLUSION-DISCUSSION 50/ a of stabilizationof the brown bear 00//n ,n ,1 Despite tendency J F M A M J A S O N D range in Greece, with some isolated cases of extra- Month limital occurrence, imminent risks of further internal fragmentationshould not be underestimated.Therefore, Fig. 8. Annual frequency of bear attacks on domestic cattle, there is urgent need for a systematic monitoring of the Pindus, Greece. potential and effective causes of aggravation of the 196 Int. Conf. Bear Res. and Manage. 9(1) 1994

. 14 Total numberof food items/month 12

6 n= 95 4 -e- n = 343

Month

Fig. 9. Annual variation of food items diversity in bear's diet. (A) Winter, spring, summer average. (B) Fall average. phenomenon at "sensitive"areas. LITERATURECITED First systematic data on bear populationstatus show ANDERSON, D., J. LAAKE, B. CRAIN, AND K. BURNHAM. a close relation between human-caused mortality 1979. Guidelinesfor line transectsampling of biological (poaching) and negative populationtrends. A thorough populations.J. Wildl. Manage.43:70-78. quantificationof the phenomenon should be enhanced CLEVENGER,A., F. PURROY, AND M. PELTON. 1990. by systematic population censuses and combined with Movementand activity patterns of a Europeanbrown bear hunter's information campaigns. Effectiveness of the in the CantabrianMounts., Spain. Int. Conf. Bear Res. existing wardening network should also be reviewed. Manage.8:205-211. Results on food habits and requirements,as well as COUTURIER,M. 1954. L'Ours Brun. Arthaud, eds. 905pp. on rhythmsof annual activity and modes of habitatuse, CURRY-LINDAHL,K. 1972. The Brown bear (Ursus arctos provide informationfor the identificationof criticalbear L.) in Europe:decline, presentdistribution, biology and food resources, habitattypes, and timing of utilization. ecology. Pages 74-80 in S. Herrero,ed. Bears, their They should be seriously taken into account when biology and management. IUCN new Ser. Publ. 23. scheduling human activities involving especially 371pp. logging, hunting, recreation, projects of natural DEBAZAC,E., AND G. MAVROMMATIS. 1971. Les grandes resources exploitation, etc. divisionsecologiques de la vegetationforestiere en Grece continentale.Bull. Soc. Bot. Fr. 118:429-452. FALIU, L., C.L. BERDUCOU, AND J. BARRAT. 1980. Le regime carnivorede l'ours brun desPyrenees. Etude preliminaire.Ciconia 4:21-32. HAINARD, R. 1961. Mammiferes sauvages d'Europe-Insectivores-Cheiropteres-Carnivores(II). Delachauxet Niestle, eds. 322pp. HERRERO,S. 1978. A comparisonof some featuresof the evolution, ecology and behavior of black and grizzly/brownbears. Carnivore1:7-17. M J J MARCUM, C., AND D. LOFTSGAARDEN. 1980. A Month nonmappingtechnique for studyinghabitat preferences. J. Wildl. Manage.44:963-968. Fig. 10. Brown bear annual activity patterns, Pindus, Greece. MATSAKIS,J., G. KATSADORAKIS,AND F. STURER. 1981. BROWNBEAR IN GREECE * Mertzanis 197

Donneespreliminaires sur l'etat des populationsd'ours PAUSANIAS."Arkadika", "Attika", "Lakonika". Loeb brundans la regionde Zagoria. XV Congr.Int. Fauna Classical Library, eds. 1954, 1955, 1965, 1969. Cinegeticay Silvestre(Trujillo, 1981). 3pp. 1,2,3,4:573pp.(In Greekwith English translation.) MEALEY,S.P. 1980. The naturalfood habitatsof grizzly QUEZEL,P. 1967. A propos de quelques hetraies du bears in YellowstoneN.P., 1973-74. Int. Conf. Bear. Macedoine. Pages200-210 in Bull. Soc. Bot. Fr. 114. Res. Manage.4:281-292. RUSSELL,R. 1971. Summerand automnfood habits of MERTZANIS,G. 1989. Quelques elements sur l'etat des island and mainlandpopulations of polar bears - a populationset l'ecologiede l'oursbrun (Ursus arctos L.) comparativestudy. M.S. Thesis, Univ. of Alberta, dans le NW. de la Grece. Biologia Gallo-Hellenica Edmonton.87pp. (15):191-198. SIMOPOULOS,K. 1984. Xeni periigites stin Ellada; de 333 ___ . 1992. Aspects Biogeographiqueset Ecologiques des A.D. a 1850. Foreigntravelers in Greecefrom 333 A.D. populationshelleniques d'ours brun (Ursusarctos L.) to 1850. Papyroweds. 1,2:1,400pp.(In Greek). Cas d'une souspopulationdu Pinde: Applicationa la SUMNER,J., ANDJ.J. CRAIGHEAD.1973. Grizzly bear conservationde l'espece et de son habitat. These, habitatsurvey in the ScapegoatWilderness, Montana. Universitede MontpellierII, France. 220pp. MontanaCoop. Wil. Res. Unit. Missoula,eds. 49pp. PAPAVASSILIOU,A. 1963. I arctos eis ton ellinikon horon TISCH,E. 1961. Seasonalfood habitsof the blackbear in (the brown bear in Greece)-KyniyetikaNea (June the WhitefishRange of NorthwesternMontana. M.S. 1963):828-830. (In Greek). Thesis, MontanaState Univ., Missoula. 108pp.