Human-Related Factors Affecting Brown Bear Recovery in the Rhodopi Mountains of Northeastern Greece

Home , Drama (regional unit), Nestos (river), Sidironero

University of Montana ScholarWorks at University of Montana

Graduate Student Theses, Dissertations, & Professional Papers Graduate School

1997

Human-related factors affecting brown bear recovery in the Rhodopi Mountains of northeastern Greece

George Giannatos The University of Montana

Follow this and additional works at: https://scholarworks.umt.edu/etd Let us know how access to this document benefits ou.y

Recommended Citation Giannatos, George, "Human-related factors affecting brown bear recovery in the Rhodopi Mountains of northeastern Greece" (1997). Graduate Student Theses, Dissertations, & Professional Papers. 8402. https://scholarworks.umt.edu/etd/8402

This Thesis is brought to you for free and open access by the Graduate School at ScholarWorks at University of Montana. It has been accepted for inclusion in Graduate Student Theses, Dissertations, & Professional Papers by an authorized administrator of ScholarWorks at University of Montana. For more information, please contact [email protected]. I s

Maureen and Mike MANSFIELD LIBRARY

The University of IVIONTANA

Pennission is granted by the author to reproduce tliis material in its entirety, provided that this material is used for scholarly purposes and is properly cited in published works and reports.

** Please check "Yes” or "No" and provide signature

Yes, I grant permission No, I do not grant pennission

Author’s Signature

D ate Or , V______

Any copying for commercial purposes or financial gain may be undertaken only with the author's explicit consent.

HUMAN-RELATED FACTORS AFFECTING

BROWN BEAR RECOVERY IN THE RHODOPI

MOUNTAINS OF NORTHEASTERN GREECE

George Giannatos

B.S., Agricultural University of Athens, 1984

Presented in partial fulfillment of the requirements for the degree of

Master of Science

UNIVERSITY OF MONTANA

1997

Approved

Chairman, Board of Examiners

Dean, Graduate Schoo

Date UMI Number: EP39203

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion.

UMT OisMftation Publishing

UMI EP39203 Published by ProQuest LLC (2013). Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code

uesf

ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 Giannatos, George M.S., March 1997 Wildlife Biology

Human-Related Factors Affecting Brown Bear Recovery in the Rhodopi Mountains of Northeastern Greece

Director: Dr. Christopher W. Servheen

The bear population in the Rhodopi Mountains of northeastern Greece is newly established, and natural recolonization has occurred in only part of the potential bear habitat. At the beginning of the century there were widespread human settlements in remote mountainous portions of the Rhodopi area. These were mainly inhabited by Muslims practicing primitive agriculture and using large quantities of wood for fuel. Saracatsani pastoralists grazed sheep in areas above 1000 m during summer. Bears were eliminated in the area by direct human-caused mortality and by habitat fragmentation and loss. Fundamental changes after World War II and the Greek Civil War, created more favorable conditions for brown bear reoccupation approximately 30 years ago. The study area includes all local municipalities in which bear sightings have been reported in the last decade. I divided the study area into three analysis areas according to bear presence/absence and compared the areas’ human population, settlement density, cattle and sheep densities, proportion of forest cover, and proportion of agricultural land from 1910 to present. Bear habitat quality improved in recent years since human impacts were reduced. Forested areas are larger, livestock density is lower, and the human population smaller and more concentrated. Historic human settlements are now important bear feeding sites and soft and hard mast has increased with the regeneration of forests. Finally, I offer a number of management recommendations to improve brown bear recovery in the Rhodopi Mountains.

11 ACKNOWLEDGMENTS

I am grateful to my graduate committee members Drs. Christopher Servheen, Daniel Pletscher, and Carroll L. Marcum for their support and guidance throughout my research project. I especially appreciate the help of Chris Servheen, my committee chair, for his role of tutor, expert advisor, and friend. Chris encouraged me while I was in Missoula and Greece, and in many ways helped make this study possible. Dan Fletcher and Les Marcum provided valuable advice on project design and analyses, and offered helpful editorial comments on my thesis. Funding for this project was provided by the European Union, the Greek Ministry of Agriculture, and World Wildlife Fund-Greece through the project ARCTOS. I thank my family, especially my father, Constantinos, who encouraged me despite his poor health, and my brother, Panos, who helped with several thesis drafts. I appreciate Nancy’s editorial assistance on my thesis. Many other friends provided moral support and encouragement, for which I will always be grateful. Finally, I would like to thank Eleni for her moral support and confidence, and for helping me to write this thesis during a difficult time of family problems.

Ill TABLE OF CONTENTS

Page

ABSTRACT...... ü

ACKNOWLEDGMENTS...... iü

LIST OF TABLES...... vi

LIST OF FIGURES...... vii

INTRODUCTION...... 1

Rhodopi Mountains ...... 4 Objectives ...... 4

STUDY A R E A ...... 6

Physiography ...... 6 Climate ...... 8 Flora and Fauna ...... 8 History of the Human Factor ...... 9 Timber Harvest ...... 11 Livestock ...... 12 Hunting ...... 13

METHODS ...... 15

Habitat Disturbance ...... 15 Human Influence ...... 19 Field Surveys ...... 22

RESULTS ...... 24

Agricultural Land ...... 24 Forest Cover ...... 26

IV Page

Timber Harvest and Road D ensity ...... 29 Human Population Density ...... 32 Livestock Grazing ...... 38

DISCUSSION...... 40

Habitat Condition...... 40 Early 1900s...... 40 Agriculture and Livestock...... 40 Human Presence ...... 41 Mid-1900s...... 43 Current Conditions ...... 47 R o ad s ...... 47 Hunting ...... 47 Tourists ...... 48 Forest Cover ...... 48 European Bear Populations outside Greece ...... 49 Zones of Connectivity ...... 50 Management Recommendations ...... 51

LITERATURE C IT E D ...... 53

APPENDIX

A. Calculation of the man-made pasturelands ...... 59 LIST OF TABLES

Table Page

1. Study Sections ...... 18

2. Analysis variables ...... 20

3. Data on the sections of the study area ...... 25

4. Study area forest and road data ...... 26

5. Percentage of each section above 1000 m e le v atio n ...... 28

6. Observed and Reported bear signs...... 28

7. Game reserves in the study a re a ...... 30

8. Human population trents (1920 - 1991)...... 33

9. Data on nomadic pastoralism in 1955 ...... 39

10. Habitat utilization in the 1900 - 1920 period and current situ atio n ...... 43

VI LIST OF FIGURES

Figure Page

1. Location of the study area in mainland G reece ...... 7

2. Current bear distribution...... 17

3. Settlements in 1920, 1940, 1961, 1991 ...... 21

4. Road network 1969, 1993 ...... 31

5. Human population change in Area I ...... 34

6. Human settlement change in Area I ...... 34

7. Human population change in Area I I ...... 35

8. Human settlement change in Area I I ...... 35

9. Human population change in Area HI (Christian Sections)...... 36

10. Human settlement change in Area HI (Christian Sections) ...... 36

11. Human population change in Area III (Muslim Sections) ...... 37

12. Human settlement change in Area III (Muslim Sections) ...... 37

Vll INTRODUCTION

The brown bear{Ursus arctos) is the most widespread of the eight bear

species. Two centuries ago, brown bears lived throughout the Holarctic zone

(Servheen 1989). Habitat destruction and human-caused mortality reduced numbers of brown bears throughout their range except in areas with very low human density

and strict conservation measures (Servheen 1990). Bears eventually disappeared

from areas in Europe where land use was intense and human tolerance of bears low.

European brown bears currently survive in small, endangered, and isolated pockets in western and eastern Europe, except in Scandinavia and Russia, where populations are larger (Servheen 1990). Populations in the Cantabrian Mountains of

Spain, the French Pyrennes, and the Italian Abruzzo and Trento are listed as

endangered and protected by law (Servheen 1990, I N.S.N.B. 1992).

Eastern European bear populations are larger, and have legal classifications

ranging from game animal to protected species. Brown bears are legally hunted in

Romania and most of the ex-Yugoslavian republics (Frkovic 1987, Rosier 1989,

Servheen 1990). Several east European countries, including Bulgaria, recently gave brown bears legal protection, and other countries, such as Slovakia, are moving toward similar reclassification (Perzanowski 1995, Spassov N. et al. 1996). Bear populations in central and southern Europe, including Poland, Greece,

Albania, and Austria are endangered and legally protected (Servheen 1990, Red

Data Book-Greece 1992, ARCTOS 1996). These populations are connected and continuous with the population of Slovakia, Bulgaria, ex-Yugoslavia and Slovenia respectively (I.R.S.N.B. 1992, Kaczensky et al. 1994, Mertzanis et al. 1994,

Perzanowski 1995).

European brown bear distribution is strongly correlated with a high degree of forest cover, high topographical relief, and low human disturbance. Continued existence of such areas is critical for the survival of small bear populations (Elgmorg

1978, Buchalczyk 1980, Zunino 1981, Camarra 1983, Roth 1983, Clevenger et al.

1987).

The brown bear is an adaptable species, and able to tolerate some human- caused landscape changes. Bears have continued to survive in Europe by becoming shy and elusive because humans have eliminated most of the aggressive and bolder individuals (Egmorg 1978, Mattson 1990).

Brown bears in Europe are opportunistic feeders, and some individuals take advantage of human-produced foods like garbage, bee hives, and livestock. Such bears are attracted to areas of human habitation and are often in conflict with human interests. Feeding on human foods is a major cause of bear mortality (Mattson

1990). Their ability to use a diverse range of foods is exploited by people trying to manipulate some bear populations in eastern Europe. Humans provide food to concentrate the animals in specific areas and allow easy hunting of them. In several places bears are managed as game animals in feeding reserves. In these cases, high numbers of bears live in a small area because they are provided with supplemental food including com and slaughterhouse refuse. This practice of feeding bears often changes their behavior so that they no longer feed on natural foods in the area.

Bears become concentrated around feeding sites. Many eastern European brown bear populations are kept artificially large because of this feeding (Weber 1987).

Brown bear populations in Scandinavia, Greece, Albania, and the Cantabrian mountains of Spain survive without human feeding (ARCTOS 1996).

One of the largest populations of European brown bears exists in Greece.

Two subpopulations, separated by 220 km, inhabit a total of approximately 9,600 km^. The subpopulation in northwest Greece ranges over more than 8000 km^ of the

Pindus mountains. At 39° N, this population represents the most southern distribution of the brown bear in Europe. The other subpopulation, located in the mountains of the western Rhodopi, extends over 1600 km- in northeast Greece. Approximately 75% of this area is characterized by year-round bear presence. The remaining 25%, located in the Vrondou Mountains, is an irregular, seasonal bear habitat, and is probably connected with the Bulgarian bear population to the north (ARCTOS 1996).

There was no systematic study of the size or status of the brown bear population in Greece or bear management plan, until 1988 when the Ministry of

Agriculture began researching bear distribution and status (Mertzanis et al. 1994). Currently, the bear population in Greece is estimated at 100-150 animals, the largest brown bear population in the European Union outside of Scandinavia (Van der

Elst et al. 1989, ARCTOS 1996).

Rhodopi Mountains

The bear population of the Rhodopi Mountains is of special interest because; 1) it is a newly established, 2) potential habitat for additional recolonization is limited, and 3) ongoing construction of a hydroelectric dam threatens to reduce the existing habitat. Bears began recolonizing the Rhodopi area during the mid-1960s (Zervas

1927, Zervas 1961, Couturier 1954), and were first reported in Rhodopi by local people in the early 1970s. Although the immigration route of these bears is unknown, they likely came from Bulgaria since this was the closest bear population during that time (Couturier 1954).

Bears have not recolonized the entire mountainous forested habitat of the

Rhodopi mountains. Since human activity is the most important factor in habitat change in the Rhodopis, more intensive conservation efforts are necessary in areas where human pressure is great and/or increasing (I.R.S.N.B. 1992).

Objectives

I examined land-use changes in relation to brown bear occupancy of habitat in the Rhodopi Mountains from 1913 to present. To identify the influences of human- related factors on bear distribution, I compared existing land uses in areas that bears are regularly present with those in areas that bears never or infrequently visit.

My main objective was to relate the trends of human habitation and land use to bear presence during the years before the bear recolonization and during the last 30 years of bear presence. The results of this study will be used to improve conservation measures for brown bear recovery in the Rhodopi Mountains. STUDY AREA

Physiography

The study area is located in northeast Greece between 41° 12' and 41° 36' N and

24° and 25° 06' E. The area includes the prefectures of Drama and Xanthi, which belong to the province of Eastern Macedonia and Thrace. Forests are managed by the

Nevrocopi, Drama, Stavroupoli, and Xanthi Forest Services.

The study area is within the drainage of the Nestos River valley and the mountainous area north of the town Xanthi, totals 2510 km^. The 234 km Nestos River flows through Greek territory for 138 km in a general northwest to southeasterly direction. The terrain is generally rugged and is characterized by steep mountain slopes and perennial, fast-flowing, cold water streams and rivers. Elevation ranges from 100-2232 m above sea level.

Numerous secondary streams and creeks have cut deep ravines in the area. The most important of these streams are: Despatis, Vathirema, Polikarpos, Arcoudorema, and Diavolorema. Most of these stream valleys run north-south.

My study area includes all municipalities in which bears are known to have been regularly present in the last decade and adjacent areas where bear presence has been recorded at least once in the last decade (Fig. 1). VO

Fig. 1. Location of the study area in mainland Greece Climate

The Rhodopi area has a combination of continental and oceanic climate types.

According to Emberger-Sauvage’s bioclimatic level classification, the climate is humid with severe winters in areas higher than 1000m, and subhumid with severe winters in lower elevation areas (Emberger 1955, Mavromatis 1980). Mean annual precipitation ranges from 500-1100 mm. August is the driest month with 20mm average precipitation. Rainfall increases significantly with increased elevation.

Snowfall can occur any time of year above 1400 m, and will remain on the ground above 1200m from late December until May.

Mean annual temperature ranges between 10° and 11.5° C. July and August are the hottest months, with maximum temperatures reaching 37° C in lower elevations. Winter temperatures remain below freezing for extended periods with lows of -27° C during January, February, and December. The average lows of the coldest months range from -4° C in the valleys to -8° C at 1200m elevation

(Mavrommatis G. 1980).

Flora and Fauna

Forests cover nearly 55% of the study area. Vegetation consists of second- growth mixed hardwood and conifer stands. At lower altitudes, the dominant vegetation is oak (Quercus spp.) or oak-hombeam {Carpinus spp.) forests. In areas above 1000 m, forests consist mainly of beech (Fagus sylvatica), spruce (Picea abies),

Scots pine (Pinus sylvestris), black pine(Pinus nigra), and birch (Betula verrucosa). The study area includes the largest birch-spruce forest in Greece, and is the most southern area in Europe where these trees form large forest blocks. An 870 ha area near the Bulgarian border is the largest stand of virgin forest in the Balkan peninsula as well as in the European Union (Bauer 1983, WWF 1988).

Nearly all the larger native European mammals, including wolf {Canis lupus), roe deer (Capreolus capreolus), wild boar {Sus scrofa), chamois {Rupicapra rupicapra), otter (Lutra lutra), and wild cat (Felis sylvestris) inhabit the Rhodopi

Mountains. The area is the southernmost limit for the distribution of central European forest birds including the Capercaile (Tetrao urogallus) and Wild cock (Bonasia bonasia), and is the only area in Greece were these endangered bird species still exist

(WWF 1988, Red Data Book 1992, University of Thrace 1993).

History of the Human Factor

The area's current vegetative cover and socio-economic setting are the result of human activities that took place over thousands of years. Human occupation and land use have especially influenced habitat in the study area during the last century.

The term "greater area" in this section refers to the area of the two prefectures of

Drama and Xanthi, a total 5261 sq.km (Figure 2).

Two main periods of human occupation and land use can be identified:

A. Prehistory-14th Century (end of the Byzantine Era)

Most human settlements in the greater area were concentrated along large rivers in the plains since these were major transportation routes. Mountainous areas 10 were sparsely populated and largely undisturbed (Kyriakidis 1960, Theocharis 1971,

Bakirtzis 1988).

B. 14th Century (Ottoman Occupation)-Present

1) 14th Century-early 1920s (end of the Greek-Turkish War). The human population in the greater area dramatically increased, especially by the end of 16th century. Immigration of other ethnic groups into the region greatly increased the human population. The mountainous region was heavily settled by Muslims during this time. By 1914, the study area was heavily populated and hosted 89% of the greater area’s population (Vakalopoulos 1919). Pastoralists grazed livestock in mountain pastures without restriction until 1924, when the Greek-Bulgarian border closed (Hatzimihali 1957, Minaidis 1984). There is no information about the bear status in the study area.

2) 1923-1940 (World War II era). The human population of the study area decreased slightly during this 17-year period, when the greater study area became

Greek territory. Muslims were forced to leave Greek territory to make it available to

Greek refuges from Asia Minor (Lagopoulos 1982, University of Thrace 1993). An exception to this exchange was the Muslim people who remained in the eastern part of the study area. However, not all of the recently-vacated settlements were occupied by newcomers. Overall, the resettlement of the Rhodopi mountain area was unsuccessful.

Many Greek settlers abandoned their new villages (Pallis 1925, Lienau 1990). Bears did not inhabit the study area at this time. 11

3) 1940-1950 (World War II, German Occupation, Greek Civil War). The study area was almost completely abandoned during this period of terror and looting by Bulgarians and Communists. Most of the area’s inhabitants fled from mountain villages to the cities and lowland villages. There were no recorded bear observations during this period.

4) 1950-1996 (Regeneration-Industrialization). The study area had the most dramatic depopulation in its recent history, which affected mainly the Christian population of the area. Nomadic pastoralism disappeared before 1980 (Psyhogios and

Papapetrou 1995). Currently, the primary human activities are timber extraction and extensive cattle, sheep, and goat grazing. Bear sightings were recorded by the mid-

1960s and were present in much of the study area by 1991.

Timber Harvest

The study area has the most productive forests in Greece—although they represent only 4.5% of the total forested area in the country, they produce 14% of the industrial lumber (Min. of Finance 1993). Timber harvest is managed by the local

Forest Service and consists mainly of selective logging (i.e., single tree harvest management with 10-12 years rotation). Small-scale clearcutting in the lowland oak forests (with 20-30 years rotation) also exists ( Forest Service, pers. comm.). Selective logging has resulted in high road density (1.11 km/km^) (Statistical Society of Greece,

1988). The public has unlimited use of forest roads except during the winter, when roads are blocked by snow. 12

Livestock

In most of the study area the predominant forms of livestock are:

1) Large-scale, extensive cattle grazing in the forest meadows or in old abandoned agricultural land and around ruins of old settlements,

2) Small-scale goat grazing in lower areas, usually below 1000 m elevation, predominantly in oak shrublands interspersed by meadows, and, to a lesser degree, aroimd some of the ruins of the old settlements.

3) Sheep flocks in either agricultural land around settlements or in large mountain meadows.

Most grazing animals graze for nine months in the study area (WWF 1988).

Goats and sheep are usually accompanied by the shepherd and several guard dogs and are kept in huts or stables at night. Grazing usually takes place within the administrative limits of the municipality but not in any specifically designated area.

There is no delineation of grazing laind within the community, so the animals, especially cattle, graze wherever they wish. As a result, local concentrations of livestock occur in some areas, usually in areas with larger meadows or in partly forested areas, where the herds are more visible. Cattle and goats disperse in small groups in small forest meadows and use forest cover during the warm summer days.

The law prohibits livestock grazing in the recently logged areas for five to ten years and there is a permanent prohibition of grazing in the 870 ha area in and around the remaining virgin forest. Herds are dispersed during the year, searching for better 13 pasture. At night animals are usually put in a specific place or stable, open or closed.

In some cases small numbers of livestock may be isolated in small forest meadows, where they become vulnerable to predation by bears or wolves. However, there are very rare cases of attacks usually at night on livestock in the concentration area.

Livestock depredation by bears and wolves along with damage on beehives are serious because, if they are left unchecked they can be a reason for farmers to take the law in their own hands and kill the predator. Livestock depredation makes local people develop a negative attitude towards conservation and conservationists. Unfortunately, the State did not deal with this predation problem until 1988. An antipredator electric fence, which was initially used experimentally in 1988, worked effectively and many beekeepers applied for new installations. Legislation for 100% compensation for bear damages on livestock was delayed until 1992. There are still problems due to bureaucratic delays in payments to farmers. Compensation takes usually six months to reach the farmer.

Hunting

The Rhodopi area is one of the best hunting grounds for wild boar in Greece.

Hunters from all over the country come during the hunting season and usually stay two or more days. Wild boar hunts are usually drive hunts, which can be detrimental to the bears, taking into consideration the rather insufficient game protection in much of the area. Drive hunts are an intense hunting method using many people and dogs. Large areas are searched for game and all wildlife larger than hares are flashed out toward 14 the waiting hunters. During the last two years, 15 bears are known to have been killed in Greece during the wild boar hunting season (ARCTOS 1996). If the animals escape, stress from being chased is high. There are no systematic data to estimate hunting pressure and the influence of hunting on local bear populations. During a radio tracking operation of a translocated wild bear, in the Rhodopi mountains a general avoidance of the area where a driven hunting operation started was observed

(ARCTOS 1996). METHODS

Habitat Disturbance

Pressure on wildlife habitat can be measured by the type and frequency or

intensity of human disturbances in the area. I examined the following variables to

measure the degree of human-related disturbances:

• human population density and number of settlements

• proportion of intensively cultivated land in the area

• free-ranging livestock density

• proportion of the area covered by forest

I grouped the municipalities in the study area into the following three

categories according to bear presence/absence (Fig. 2) :

Area I. This is the core area of bear distribution, with bears regularly present in

more than 80% of each municipality during the last decade.

Area II. This is the overflow zone of bear distribution, with bears regularly present

in more than 40% of each municipality during the last decade.

Area III. This barrier zone is characterized by bears either absent or regularly present in less than 20% of each municipality during the last decade.

The study area encompasses 24 municipalities ranging in size from 14.3 km^

15 16 to 577.2 km^, with substantial variation in sizes between administrative units. To avoid bias, the study area was divided into areas of approximately equal size. The following restrictions were taken into consideration when making these delineations, or "study sections":

1. The section sizes were set within the European brown bear home range size in

Spain (Clevenger and Purroy 1991) and Scandinavia (Bjarvall et al. 1990).

2. Study sections followed administrative limits of municipalities. Livestock are usually grazed within these boundaries, and the statistics I used in my analyses applied to municipalities.

The study area was divided into 13 sections, with each section composed of one, two, or four municipalities (Table 1). Municipalities were grouped according to the similarity of land uses, socio-economic status, and the existence of geographic continuity. Limits of Dear distribution 10 the South

Section Limits

Municipality limits

I Ilia

a 'lib • / i (/) *’3 K m 2 4 6 o>cd I ri bb ÜU Table 1. Study sections

Analysis Total Area I = 1057.8 km^ Total Area II = areas 333 km^ Section la Ib Ic Id Ha Ilb Area (km^) 130 257 93.6 577.2 195.8 137.2 H 1 1 1 1 2 2 Municipalities Main Forestry Forestry Forestry Forestry Forestry Forestry economic Livestock Livestock Livestock Livestock Livestock Agriculture activities (in Services ranking order) Religion Christian Christian Christian Christian Christian Christian

Analysis Total Area III = 1118.2 km^ areas Section Ilia Illb IIIc Hid Hie Hlf Hig Area (km^) 156 91 150.2 181.2 174.1 190 175.7 no. of 4 1 1 4 2 2 2 Municipalities Main economic Forestry Livestock Forestry Livestock Livestock Livestock Agriculture activities (in Livestock Forestry Agriculture Agriculture Agriculture Agriculture Livestock ranking order) Services Forestry Forestry Religion Christian Christian Christian Christian Muslim Muslim Muslim 19

Human Influence

Data on human population, land use, livestock numbers, and size of forested and agricultural areas were obtained from local and central Forest Services, Agricultural bureaus, local municipalities, the National Statistical Society of Greece, and the

Geographic Society of the Greek Army. All ruins of old settlements were investigated and locations recorded. I verified current numbers of cattle, sheep, and goats by interviewing local livestock owners.

Human population data and settlement distribution data came from national population censuses conducted since 1913, when the study area became part of the

Greek State. Statistics on municipality-based land use were available at ten-year intervals from 1961 to 1991 (Statistical Society of Greece 1961-91).

To examine changes in human population, and number and distribution of settlements within a longer time period, graphical comparisons were made. The population counts of 1913, 1920, 1928, 1940, 1951, 1961, 1971, and 1981, and 1991 censuses were used in my comparisons as well as settlement data (according to 1920,

1940, 1961, and 1991 censuses). These years were selected because they represent times when fundamental changes in settlement numbers occurred (Fig. 3).

The difference between Christian and Muslim land-use patterns in Area III is discussed. The variables I used in my analysis are listed below. 20

Table 2. Variables used in the analysis

VARIABLES ANALYZED COMPARISON

YEARS

Forested area (%) 1961 and 1991 % change in forested area (1961 = 100) 1961 and 1991 Human population density 1961 and 1991 % change of human population size (1920=100) 1920, 1961, 1991 Agricultural land {%) 1961 and 1991 Settlement density (number / 100 km^) 1961 and 1991 Sheep density (number / area) 1961 and 1991 Cattle density (number of animals / land area) 1961 and 1991 % loggers (# loggers / section population) 1993 Road density (km roads / km^ ) 1991

Areas not included in agriculture or forested land are either 1) grazing land: usually fragmented forested habitat in higher elevations and fragmented oak shrubland in lower elevations, or 2) alternative land-use areas such as settlements, open water, and roads.

These areas constitute a very small part of the study area. 21

Fig. 3. Settlements in 1920, 1940, 1961, 1991 22

Field Surveys

I gathered data on bear presence during field work for an ARCTOS project. The goal of this project was to create an integrated management plan for improvement of bear recovery in Greece.

Field work included eight field surveys of 20 to 22 days each to gather data on bear presence/absence. The surveys were distributed throughout the year and covered all non-denning seasons. Sampling periods were separated by 20-25 day intervals.

Surveys entailed searching for bear signs along forest roads, old trails, and seasonal feeding areas such as abandoned orchards. Search transects were set up according to information on bear sightings, quality of bear habitat, information from local people, and Hunting Association reports. Roads and trails totaling 826 km covering 2,400 km^ were searched for bear sign during each survey.All potential bear habitat within the study area was sampled. I also filled out field sign observation cards and collected bear scats.

This sampling method has some limitations relating to the collection of animal presence data due to the nature of ground substrate and to weather conditions during the survey (Harris 1986). Therefore, presence/absence data was supplemented by reports from local people (farmers, loggers, foresters), mountaineers, hunters, and recreationists. These reports have been verified whenever possible by immediate visits to the area. Unverified reports were carefully selected and investigated, and only the most reliable ones were used in my analysis. 23

Because this field work covered only one year (May 1994 to April 1995), I used data from the 1988 bear survey (Mertzanis et al, 1990), and Forest Service data which had been gathered since 1987 when determining bear presence in an area and when setting limits of bear distribution. Surveys were not standardized in effort between the analysis areas. RESULTS

Agricultural Land

The percentage of agricultural land during 1961 and 1991 was significantly lower in Area I than in Areas II and III. Between 1961 and 1991 the percentage of agricultural land decreased in all four sections of Area I and in much of Area II.

Section lib is the only section in Areas I and II where the percentage of agricultural land is currently higher than 1% and agriculture is commercial and intensive. In five of the six sections of Areas I and II the agricultural sector has been eliminated from the local economy. Agricultural land in these sections range from 0% to 0.6% in 1991

(Table 3).

Within Area III, there was a marked difference between the percentages of agricultural land and the land practices of Christian and Muslim settlements between

1961 and 1991 (Table 3). Section Illb (Volakas community) is the only predominantly-Christian section where the percentage of agricultural land remained stable between 1961 and 1991, During the last 30 years, there have been increases in the percentages of agricultural land in all three sections with Muslim villages.

24 Table 3. Data on the sections of the study area

Study % Forest % Population Settlement Cattle Sheep density Section cover Agricultral density Density (# per density (head /km^^ land (Humans/ 100 km^) (head/km^) km^) Year 1961 1991 1961 1991 1961 1991 1961 1991 1961 1991 1961 1991 la 32 67 0.6 0 1.4 0.4 1.5 0.77 1.7 6 24 0 Ib 59 72.8 1.4 0.3 3.3 1.5 1.96 1.16 2.9 1.5 19 3.35 Ic 31 51 1.1 0.09 2.3 1.4 1.07 1.07 3 5.2 36.5 0 Id 65 69 0.3 0.2 0.9 0.2 0.7 0.17 1 2.36 7.25 4.87 lia 34 36.5 1.2 0.6 3.3 1.6 1.5 1.02 4.8 1.55 16.3 5 lib 33 36 3.8 2.1 18.6 10.1 5.8 3.6 8.7 1.15 75.6 1.96 Ilia 59 60 7.2 6.5 13.2 5.6 5.9 3.8 10.6 8.2 39 33

Illb 35 32 5 5 15.5 12 1.09 1.09 12.2 5.4 120 69.6 IIIc 36 37.3 4.2 2.3 20 8 9.09 6.66 10.6 5.3 55.2 23.3 Hid 41.5 32 6.5 4.5 22 7.2 10.5 7.7 11 6.1 47.8 33.3 Hie 60.6 60 5.8 6.9 24 21 5.3 4.6 4.3 1.2 64.7 48 Hlf 75 40 5 6.7 30 38.6 14.7 14.7 7.9 1.3 81.5 122 HIg 79.5 78 4.8 7.5 28.7 25 5.1 5.1 5.3 0.65 58 50 26

Forest Cover

Between 1961 and 1991, forest regeneration was significantly greater in Area I, with increases in all sections, than in Areas II and III (Table 4). More than 50% of all

Area I sections are currently forested (Table 3).

Table 4. Study area forest and road data.

Section Area Forested area Road density % loggers

(km^) ’61 - ‘91 (km /km^) -

(index for 1961 = (1991) # of loggers /

100) total

population

(1993) la 130 209 0.6 43.6 Ib 257 123 1.25 36 Ic 93.6 164 1.16 67 Id 577.2 106 1.06 27 lia 195.8 107 0.86 17 lib 137.6 109 1.97 5 Ilia 156 101.6 1.35 29 Illb 91 92 1.18 5.5 IIIc 150.2 103.6 1.74 2.2 Hid 181.2 77 1.18 6 Ille 174.1 99 1.87 3* Illf 190 53.4 0.32 1* Illg 175.7 98 0.98 2* ^Estimation of the local Forest Service 27

Forested areas increased slightly in both sections of Area II, but to a lesser extent than that of Area I. Sections of Area III show either no change or a decrease in forested area (Table 4).

Reduction of forest cover is thought to be an important cause of the decline of brown bears numbers, especially in European bear populations. Reduced cover is correlated with decreased bear habitat security and increased human intrusion (Krott

1962, Camarra 1983, Clevenger et al. 1987).

In addition to the size of forested areas, the nature and degree of human activities in these areas have impacted bear populations. Muslims cut branches of broadleaf trees for various uses including livestock feed £ind bedding material in stables. This practice of cutting branches and logging has transformed these areas into monocultures with virtually no mast production and little bear food. This lack of food and scarcity of secure hiding places makes these areas unsuitable for bears (as well as other wildlife species). These open, oak-dominated woodlands are typically overgrazed by sheep and goats.

Oak trees in northern Greece produce maximum mast when 50-60 years old, with mast production at four to five year intervals (Gatzogiannis, pers. comm.). The majority of oak trees in the study area are young and dense due to long-term clearcutting with a rotation of 20-30 years. Since harvest occurs before trees reach full maturity, these oaks rarely produce substantial mast. These harvest practices are considered by the Forest Service to be the most profitable because most wood is sold 28 for fuel.

Young oak forests can sometimes be used as refuge areas for bears or other wildlife, especially in areas where there is a seasonal food source nearby, like wild orchards or isolated mountain crops. Traditionally-used oak woodland and regenerated oak bush are degraded habitats and have similar poor mast production.

Most of Area III is below 1000 m (Table 5) where forests are mostly oak- dominated. Due to the sylvicultural practices just described, these sections consist mainly of degraded bear habitat. Also, areas below 1000 m have more human disturbance and provide fewer opportunities for secure bear habitat (Table 6).

Table 5. Percentage of each section above 1000m elevation.

Segment la Ib Ic Id II II III III III III III III III a b a b c d e f 8 % of the area above 40 40 50 40 20 30 5 40 1 3 30 3 10 1000 m

Table 6. Observed and reported bear sign Elevation Area I Area II Area III Study Area (a) (1+11+III) >1000 m 103 33 12 148 < 1000 m 24 6 0 30 Total 127 39 12 178 29

Timber Harvest and Road Density

Systematic timber extraction in the area started in the early 1970s. The first information about systematic forest exploitation was published in 1962 (Min. of

Agriculture). The volume of lumber production between 1973 and 1977 increased fivefold and, consequently, forest road construction increased (Min. of Coordination

1979). Selective timber extraction practiced in the study area required high road densities. Road density has increased dramatically since the minimal forest road network of 1969 (GSGA 1966) (Figure 4).

Of the 13 study area sections, only one has a road density less than 0.61 km/km- (1 mi /mfi), the optimum maximum management road density target for bear conservation in U.S. national forests (USFWS 1993).

Many bears were sighted and evidence of bear presence observed in parts of the study area with high open road densities (ARCTOS 1996). However, these were the most remote parts of the study area, far from paved roads and large human settlements.

There are few game reserves in the study area (Table 7). Although protected, these areas are too small to support even one bear and are, therefore, inadequate by themselves for bear conservation. 30

Table 7. Game Reserves in the study area

Area I Area II Area III Section Area (in Section Area (in Section Area (in km^) km^) km^) Id 53.1 lib 9 Ilia +inb ** 80 Id 32.5 nie +mf 32.7 Id 22 me + m f 40 ** Note: Nearly half of this reserve is outside the study area 31

1969

1993 Fig. 4. Road network 1969, 1993 32

Human Population Density

Area I had a nearly constant decrease in population size and number of settlements from 1913 until present (Figure 5). The rate of population decrease in Area

I was greater than Area I between 1920 and 1961. In contrast, the human population in five of the seven Area III sections increased during this time.

Between 1961 and 1991, the human population decreased in all but one study section (Illf) (Table 8). Area I had significantly lower human population density than

Areas II and III in both 1961 and 1991. Similarly, settlement density in Area I was lower than Area II in 1961 and 1991, and both were lower than Area III. Settlement density decreased between 1961-1991 in nine sections and remained unchanged in four (Table 3).

Human population and settlement densities in the Christian sections of Area III decreased (Figures 9 and 10), but increased or showed no change in Muslim sections

(Figures 11 and 12). 33

Table 8. Human population trends (1920 - 1991) Value index for 1920 = 100

Section 1920 1961 1991 la 100 6.5 1.85 Ib 100 17 7.6 le 100 13.5 7.4 Id 100 9.7 3.5 lia 100 13.8 6.7 nb 100 71.6 38.9 lUa 100 39.6 16.8 Illb 100 76 59 nie 100 118 47.4 nid 100 108 36 nie 100 127 112 Illf 100 130 168 Illg 100 114 100 34

8000 ■ la 7000 ■ Ib 6000 ■ le 5000 □ Id 4000 3000 2000 1000

1913 1920 1928 1940 1951 1961 1971 1981 1991 YEARS OF CENSUS

Figures. HUMAN POPULATION CHANGE IN AREA I

25 j ■ 1a ■ 1 b 20 ■■

1920 1940 1961 1991 YEARS OF CENSUS

Figure 6. HUMAN SETTLEMENT CHANGE IN AREA I 35

7000 T

4000 •

Gl 2

1000 ■

1913 1920 1928 1940 1951 1961 1971 1981 1991 YEARS OF CENSUS

Figure 7. HUMAN POPULATION CHANGE IN AREA I!

YEARS OF CENSUS

Figure 8. HUMAN SETTLEMENT CHANGE IN AREA II 36

7000 T ■ Ilia 6000 - ■ Illb ■ IIIc 5000 ■ □ llld 4000 ' 3000 2000 1000

1913 1920 1928 19511940 1961 1971 1981 1991 YEARS OF CENSUS

Figure 9. HUMAN POPULATION CHANGE IN AREA IH (CHRISTIAN SECTIONS)

■ Ilia 20 T B lllb I------■ IIIc

uj 10

1920WÀÀ] 1940 1961 1991 YEARS OF CENSUS

Figure 10. HUMAN SETTLEMENT CHANGE IN AREA HI (CHRISTIAN SECTIONS) 37

8000 T 7000 ■ 6000 - 5000 4000

2000

1920 1928 1940 1951 1961 1971 1981 1991 YEARS OF CENSUS

Figure 11. HUMAN POPULATION CHANGE IN AREA IH (MUSLIM SECTIONS)

30 T

YEARS OF CENSUS

Figure 12. HUMAN SETTLEMENT CHANGE IN AREA III (MUSLIM SECTIONS) 38

Livestock Grazing

Sheep density decreased dramatically between 1961 and 1991 in Area I, and was significantly lower than that of Areas II and IE in both 1961 and 1991 (Table

3). The largest number of free-ranging sheep flocks in the study area were in Area I during 1955 (Table 9). In three of the four Area I sections, cattle numbers increased

(Table 3) due to a shift from sheep to cattle grazing. Nomadic pastoralism disappeared from the study area by 1980.

There was a dramatic decrease in sheep density in both sections of Area II between 1961 and 1991. Numbers of cattle in Area II also decreased during this period (Table 3).

All sections of Area III showed decreases in cattle density between 1961 and

1991 and decreased sheep grazing in six of the seven sections (Table 3). Only section Illf showed an increase in sheep numbers. This is the only section with a constant increase in human population since 1920, and exhibits the most rapid deforestation in the entire study area. 39

Table 9. Data on nomadic pastoralism in 1955 (Hatzimihali 1957)

Sections # of Pastoralist families # of sheep la + Ib* 247 52,025 Ic 61 18,300 Id 405 78,815 lia 55 11,745 nb ** 21 3,300 IIIc ** 37 10,500 md ** 19 5,500 nig 25 6,000

* Most of the flocks were within the area of section Ib (Sidironero municipality).

** Includes wintering flocks also DISCUSSION

Habitat Condition

Early 1900s

It is possible to estimate living conditions and make inferences about bear habitat in this area during the early 1900s by looking at available statistics on human population size, number of settlements, number of nomadic sheep, and known human living conditions.

Agriculture and Livestock

Although the extent of agricultural and grazing land of the old villages is unknown, it must have been larger than the area around present-day settlements for the following reasons: Early agricultural methods were far less productive. Peasants had to grow crops to feed themselves and their families, to feed livestock during winter, and for commercial tobacco cultivation (Min. of Finance 1993, Giannopoulou 1994). Crops mainly grew in infertile land, so large areas were needed (Min. of Coordination

1979). The climate limited the annual cultivation period to eight months or less. This is not long enough to have more than one crop in a given area of land. To grow two or more crops large farming areas were needed.

Domestic animals were grazed extensively and most grazing areas of local

40 41 residents were below 1000 m. Above this elevation was snow cover in winter and pastoralists free-ranging flocks in summer. Domestic sheep and goats can cause dramatic habitat conversion, compete with wildlife for food, and increase the possibility of bear - human conflicts due to predation (Mattson 1990). Sheep and goats are more vulnerable to predation and to surplus killing than cattle or horses

(Mysterud 1976, Griffel and Basile 1981, Jorgensen et al. 1983, Knight and Judd

1983, Camarra 1986). Therefore, areas with high sheep densities are usually high conflict zones for bears.

Cultivated and grazed lands were highly unsuitable for bears, and can be considered an exclusion zone characterized by major habitat modification and

intolerance of large predators. This zone is the area of settlements, built-up areas,

intensively-cultivated land, and pasture land with high numbers of livestock. Little habitat existed in the first two decades of this century outside this area.

Human Presence

Human presence is also an important factor influencing quality of bear habitat.

At the beginning of the century, people in Rhodopi area competed with wildlife for limited resources and tolerated only animals which they considered beneficial or indifferent. This intense land use, the relatively open landscape, and centuries-long absence of conservation and management led to the disappearance of bears from the

Rhodopi area. This loss probably occured long before the 1900s.

Labor-intensive primitive cultivation is more destructive than seasonal modem 42 cultivation because large numbers of people work in thefields and surrounding areas for long periods.The possibility of encounters between people and wildlife, including bears increases dramatically.

Using available statistics, I estimated the non-impacted habitat of Area I and compared it with the current condition (Table 10).

The situation in the first two decades of this century was likely highly fragmented with overgrazed pasturelands. The few, small, scattered wooded areas existing at that time were disturbed by peasants all year in lower elevational areas and by nomadic pastoralists during the summer in higher areas. Woodlands with high human use could not provide secure habitat for brown bears. Even today, woodlands around Muslim villages are highly degraded and devoid of wildlife despite the use of electricity instead of wood. It seems that there were few, if any, wild parts of the area before

1920,which explains the absence of bears. 43

Table 10. Habitat utilization in the 1900-1920 period and the current situation.

T = Total Area I (km^) 1,057.8

Year period 1900-1920 1985-1996

P = # of permanent 61 6 settlements p = Human population of 18,505 (1913 694 (1991 Census) permanent settlements Census) N = # of nomadic 15-20 * - settlements

n = Human population of 5,000 * - nomadic settlements Total settlements = P + N 76-81 6

Total population = p + n 23,505 694

Number of nomadic sheep 149,140 -

G = Minimum open 894.8 - grassland requirements of nomadic flocks (in km^) (1) Land for other uses (see in 162.9 1,057.8 the text): A = T-G ( in km^) * Estimated (1) : Calculation method detailed in Appendix A

Mid-1900s

The Rhodopi area, especially Area I, was significantly transformed during the last 70 years. The number of settlements and people declined in Area I until World

War II. A dramatic decline followed, resulting in actual depopulation of the area. 44

Similarly, nomadic pastoralism was eliminated completely during the first two decades after World War II.

With this massive human population density decline and reduction of nomadic

flocks in the highlands, forests began to regenerate. According to locals, numbers of wild ungulates increased dramatically. According to their reports, bears reappeared in

Area I by the mid-1960s.

In general, Christian villages have tended to reduce agricultural land and to

abandon marginal land. Christian areas were depopulated due to aging of the

inhabitants and mechanization of agriculture. Section Illb (Volakas community) is the

only Christian section where the percentage of agricultural land remained stable between 1961 and 1991. Increases in Muslim agricultural areas can be attributed to high population density and to the primitive agricultural practices of the local people.

Mattson (1990) reports that pressure on bear habitat increases as cultivated

land increases. The reverse is also true: in North America, cropland abandonment

resulted in highly productive bear habitat.

Resettlement of the highlands during the 1923 human population shift was generally unsuccessful because most newcomers preferred the lowlands and nearby fertile agricultural land. This benefited bears because the forest regenerated in the abandoned areas.

As early as 1928, the distribution of human population changed. It shifted from a dispersed population, with many similar-sized settlements, to a centralized 45 structure with one large settlement and several smaller “satellite” settlements. This centralization was more intense in areas with fertile valley land or regional advantages such as occurrence on main transportation routes or local administrative, agricultural, or trading centers (Lieneau 1990). Settlement centralization also reduced human pressure on bear habitat.

Pressure from human populations increases when the human population density increases and the settlements are more dispersed (high settlement density).

Mattson (1990) reported that there is relatively little overlap between occupied bear habitat and high human density (>25 / sq.km).

The majority of bear signs were found in areas with larger numbers of ruined settlements (Area I). These areas had larger amounts of food supply in the form of soft and hard mast trees in abandoned old orchards (Table 7). Ruins by area are:

Area I: 55; Area II: 21; Area III: 20.

Old villages and surrounding abandoned farms increased diversity of bear foods because of abundant fruit bushes (Rosa sp., Rubus ideaeus, Cory lus avellana) in these areas. Rosa sp., an important food for bears, especially in spring, became numerous. In bear scat collected during 1994 and 1995, the overall frequency of occurrence of Rosa sp. was 25% while in spring the occurrence of Rosa sp. was 70%

(ARCTOS 1996). Consumption of wild-growing domestic fruits was recorded in other

European studies (Zunino and Herrero 1972, Berducou et al. 1983) and North

America (Servheen 1983, Garner and Vaughan 1987). 46

Area I, the core bear habitat, experienced the most dramatic changes during the last 70 years. This area shows the highest recovery of forest cover, the greatest human depopulation and abandonment of settlements, the greatest reduction of sheep flocks, the lowest sheep density, and the lowest proportion of agricultural land. The population and settlement density of Area I have been the lowest in the study area since 1961.

Area H, the overflow area, showed declines in human population density and settlement density, and intense sheep flock reduction in the last 30 years. However, reforestation has not been as extensive here as in Area I.

Area III, the barrier zone, showed marked differences between Christian and

Muslim sections in both land use and human population numbers and trends. The

Christian sections show decreasing trends in population size and settlements.

Numbers of sheep have generally decreased, but sheep densities are still much higher than Areas I and II. Although Area III agricultural land remained constant in one section and declined in three, it still occupies a larger proportion of land than in Areas

I or II.

Human-caused pressure on bear habitat is now the lowest in the core area, the highest in the barrier zone, and intermediate in the overflow area. In general, human pressure declined between 1961 to 1991. 47

Current Conditions

The Rhodopi area has different problems and conservation challenges today than in the beginning of the century or even a few decades ago.

Roads

Road density is so high that there are only small forested areas further than

350 m from a forest road (Mattioli 1994). A dense road network is a major cause of disturbance and mortality in areas where poaching is a problem. This makes road management a high priority.

The effect of roads on bears is variable. Studies in North America show that some bears avoid open roads regardless of use level on the road (Archibald et al.

1987, Mace et al. 1996). This may result in loss of availability of important roadside habitat. Females or family groups often avoid roads (Archibald et al. 1987, Mace et al. 1996). McLellan (1990) reported high bear densities in remote areas with high open road densities but where road use was minimal and roads were located far from paved roads and towns.

Hunting

Hunting pressure in Rhodopi appears to be high. Hunting affects bears both directly and indirectly. According to the Hunters Association of Drama and local people, there has been a rapid decline in game species numbers in the study area during the last decade. Bear populations benefit from high ungulate densities by scavenging ungulate carcasses or actively killing weak animals (Cole 1972, Kaal 48

1976, Fillonov 1980, Mattson et al. 1987, Reynolds and Gamer 1987, Clevenger and

Purroy 1988). This high quality food can be important in bear diets, especially in the spring when they emerge from hibernation (Clevenger and Purroy 1988).

Tourists

The Rhodopi area has recently attracted an increasing number of motorized tourists, an additional cause of disturbance to bears. Human recreational intrusion into bear habitat has been detrimental to relict bear populations elsewhere in Europe, where mass tourism increased bear dispersal outside the park because of habitat security loss (Zunino and Herrero 1972, Zunino 1981). Although the annual number of recreationists visiting the Rhodopi area is a fraction of the tourists else were, it is necessary to initiate measures soon, such as restricting access in some areas to prevent disturbance. At present, the virgin forest is the only portion of the Rhodopis with limited access.

Forest Cover

Fruit production in wild orchards throughout the study area is likely to decline as native forests intrude into these areas (A similar phenomenon was described by

Garner and Vaughan, 1987). Reduction of these important food sources may have serious consequences on the local bear population. Declining mast production could result in reproduction failures (Rogers 1978) and/or increased movements and altered distribution of bears with consequent increased depredation on animals and crops (Gamer and Vaughan 1987). 49

European Bear Populations outside Greece

In some areas of western Europe brown bears have survived for a long time as small relict populations in areas with large human populations (Zunino and Herrero

1972, Zunino 1981, Roth 1983, Clevenger et al. 1987, Camarra 1989). Some bear populations use high protein foods produced by agro-pastoral activities (Zunino and

Herrero 1972, Zunino 1981). In these cases, dependence on human food has been a source of conflict between bears and people. However, because these bears survived in very small numbers, damage to livestock was minimal and generally tolerated

(Zunnino 1981). These small brown bear populations have been on the decline at least since the beginning of this century (Camarra 1989). The bear population in the

Cantabrian Mountains has been declining since the 1800s (Clevenger et al. 1987).

Most relict bear populations in Italy, France, and Spain were saved from immediate extinction by legal protection (Zunino and Herrero 1972, Zunino 1981,

Roth 1983, Clevenger et al. 1987, Camarra 1989). Even in these cases there were usually small, secure habitat areas inaccessible to or rarely visited by humans (Zunino and Herrero 1972, Zunino 1981, Roth 1983, Clevenger et al. 1987).

Dependence on agro-pastoral activity for food will not provide long-term viability. Human and bear co-existence depends on accessibility of habitat and tolerance of the local people.

The Rhodopi area has two serious disadvantages for maintaining a healthy brown bear population compared to other European areas. First, it has more rolling, open 50 landscape than the Alps, Pyrennes, Cantabrian, and Apennine mountains. This reduces the number of many secure, inaccessible areas. Second, there were no hunting restrictions or other conservation measures until the 1970s, even though the bear population was highly endangered.

Bear populations could only make a comeback after the area was abandoned by humans and secure habitat was restored. Also, the existence of a source population in adjacent areas of Bulgaria was essential for bear recolonization of the Rhodopi

Mountains.

Zones of Connectivity

Bear status and habitat conditions in Bulgaria are important for the future of brown bears in the Rhodopi area since the newly-established bear population originated in Bulgaria, the closest natural bear population (ARCTOS 1996, Spassov et al. 1996).

The study area is continuous with occuped bear habitat in Bulgaria. Together they form a habitat unit isolated from the other Balkan bear populations (Servheen

1990, Spiridonov and Spassov 1990, Servheen et al. (in press)). However, the future of bear populations in Bulgaria is not encouraging. High human pressure adjacent to the Greek border, high road density, and few protected and secure habitat linkages seem to exist (Servheen in ARCTUROS internal report 1995, Servheen (in press)).

Human population density in Bulgaria is high for the concurrent existence of 51

bear populations (Mattson 1990). Settlement densities are higher than those of the best

bear habitat in Greece. Road densities in Bulgaria along the border are similar to

those in Greece.

The Bulgarian border patrol recorded bear tracks during the last five years

(Spassov et al. 1996). These reports show an average of 12-15 bear crossings per

year between Bulgaria and Greece. More data are needed to assess the status or possibility at linkages with southern Bulgaria bear populations.

Management Recommendations

The Rhodopi area has lower human pressure on brown bear habitat than other bear populations in southern Europe, and most sections in the barrier zone can be compared with the best areas in the Cantabrian Mountains. Improvement and enforcement of an appropriate conservation plan is imperative if bears are to survive

in the Rhodopi area. Conservation measures should aim to: 1) improve habitat condition in the study area so that all potential bear habitat can be occupied uniformly by bears, and 2) maintain migration corridors with the Bulgarian bear population.

These objectives may be achieved by reducing human pressure on habitat, creating more secure zones through road closures, and creating larger security.

Improving habitat in areas below 1000 m through better management and protection of low elevation oak woodlands will augment the potential food supply and careful management of existing wild orchards will maintain their productivity. 52

Increased conservation efforts in the barrier zone, especially in the Muslim sections, is also important.

Considering that even under ideal conservation and management conditions the existing habitat within the study area may not be large enough to ensure long-term survival of a brown bear population, it is vital to improve linkages to neighboring bear populations in Bulgaria.

Finally, it is important to research the situation in adjacent areas of Bulgaria and to combine Greek conservation efforts and actions with Bulgarians for long-term preservation of the brown bear population the two countries share. LITERATURE CITED

Archibald, W.R., R. Ellis, and A.N. Hamilton. 1987. Responses of grizzly bears to logging truck traffic in the Kimsquit River Valley, British Columbia. Int. Conf. Bear Res. and Manage. 7:251-257.

ARCTOS, January 1996. Final Rep. LIFE Proj. 140 pp.

Berducou, C., L. Faliou, and I. Barrat. 1983. The food habits of the brown bear in the national park of the western Pyrenees (France) as revealed by faeces analysis. Acta. Zool. Fenn. 174:153-156.

Bjarvaal A., F. Sandegren, and P. Wabakken. 1990. Large home ranges and possible early sexual maturity in Scandinavian bears. Int. Conf. Bear Res.and Manage. 8: 237-241.

Buchalczyk, T. 1980. The brown bear in Poland. Int. Conf. Bear Res. and Manage. 5: 229-232.

Boer, H.J. 1979. The need for the protection of virgin forest of Rhodopi. Proceedings Congress for protection of Flora and Fauna of Greece, Athens.

Camarra, J.J. 1983. Habitat utilisation of brown bears in the western Pyrenees. Acta Zool. Fenn. 174:157-158.

______. 1986. Changes in brown bear predation on livestock in the western French Pyrenees from 1968 to 1979. Int. Conf. Bear Res. and Manage. 6: 183-186.

. 1989. L’ Ours brun, Hatier (ed.).

Cambell, K.J. 1964. Honour, Family and Patronage. A Study of Institutions and Moral values in a Greek Mountain Community, Oxford Univ. Press.

Cole, G.F. 1972. Grizzly bear-elk relationships in Yellowstone National Park. J. Wildl. Manage. 7: 1-8.

53 Clevenger, A.P., F J . Purroy, and M.S. De Burguaga. 1987. Status of the brown bear in the Cantabrian Mountains, Spain. Int.Conf. Bear Res. and Manage. 7:1-8.

Clevenger, A.P. and F.J. Purroy. 1988. El oso en Leon. 100 pp.

______. 1991. Ecologia del oso pardo en Espana Monographia. 147 pp.

Couturier, M. 1954. L ours brun. Grenoble, France, 905 pp. (In French).

Elgmorg, K. 1978. Human impact on a brown bear population{Ursus arctos L.) Biol. Conserv. 13: pp. 81-103.

Emberger, L. 1955. Une classification biogeographique des climats. Rev. Tran. Fac. Sa. Montpellier, Bot.7, pp. 3-43.

Fermor, P.L. 1991. Roumeli. Travels in mainland Greece.

Filonov, C. 1980. Predator-prey problems in nature reserves of the European part of the RSFSR. J. Wildl. Manage. 44 (2): 389 -396.

Frkovic, A., R.L. Ruff, L. Cicniak, and D. Huber. 1987. Brown bear mortality during 1946-85 in Gorski Kotar, Yugoslavia. Int. Con. Bear Res.and Manage. 7:87-92.

Garner, N.P. and M.R. Vaughan. 1987. Black bears’ use of abandoned homesites in Shenandoah National Park. Int. Conf. Bear Res. and Manage. 7:151-157.

Geographic Society of the Greek Army. 1966. Maps.

Giannopoulou, M. 1994. Typologie correlation of planning and ekistic characteristics of the rural settlements of Xanthi and Rhodopi prefectures. Ph D dissertation, Univ. of Thraki.

Griffel, D.E. and J.V. Basile. 1981. Identifying sheep killed by bears. Research note INT- 313. U.S. Department of Agriculture. 3 pp.

Harris, R.B. 1986. Grizzly bear population monitoring: Current options and considerations. Montana For. Conserv. Exp, Sta., School of Forestry, Univ. Mont., Missoula. Misc. Publ. 45. 84 pp.

Hatzimihali, A. 1957. The Saracatsani, Vol. A, Athens.

54 I.R.s .N.B. (Institut Royal des Sciences Naturelles de Belgique). 1992. Actions de conservation visant a la protection de 1’ ours brun en Grece, en Italy et en Espagne. Rapport final. 120 pp.

Jorgensen, C.J. 1983. Bear-sheep interactions, Targhee National Forest. Int. Conf. Bear Res. and Manage. 5: 191-200.

Kaal, M. 1976. Ecology, protection and prospect of utilization of the brown bear in the Estonian S.S.R. Int. Conf. Bear Res. and Manage. 8: 79-84.

Kaczensky, P., F. Knauer, T. Huber, M. Jonosovic, and M. Adamic. 1994. The Ljubljana-Postojna Highway-a deadly barrier for brown bears in Slovenia? A paper presented in the 1994 Symposium “A Coexistence of Large Predators and Man. ”

Knight, R.R., and S.L. Judd. 1983. Grizzly bears that kill livestock. Int. Conf. Bear Res. and Manage. 5: 186-190.

Krott, P. 1962. Beitrage zur Kenntnis des Alpenbars, Ursus arctos Linne 1758. Saugetierkundlinche Mitteilungen. 35 pp.

Kyriakidis, S. 1960. About the Thakian History. Research Institute for Balkan Peninsula 36. Thessaloniki.

Lagopoulos, P. 1982. Industrialisation regional labour market and productive investments by remigrants in a peripheral region. Rep. No. 3. Univ. of Thessaloniki, 20 pp.

Lienau, C. 1990. Europark in Nordost-Griechenland? Hellenica Jahrbuch fur die Freunde Griechenlands.

Mace, R.C., J.S. Waller, T.L. Manley, L.S. Lyon, and H. Zuuring. 1996. Relationships among grizzly bears, roads, and habitat in the Swan Mountains, Mont. J. App. Ecol. 33(6): 1395-1404.

Mattioli, M. 1994. The Utilization of GIS for brown bear habitat quality assessment in the area of Rodopi-Greece. M.Sc. Lausanne. 59 pp.

Mattson, D.J. 1990. Human impacts on bear habitat use. Int. Conf. Bear Res. and Manage. 8:33-56.

55 Mattson, D.J., R.R. Knight, and B.M. Blanchard. 1987. The effects of developments and primary roads on grizzly bear habitat use in Yellowstone National Park. Int. Conf. Bear Res. and Manage. 8:57-64.

Mavrommatis, G. 1980. The bioclimatic zones of Greece. Relation of the climate and Vegetation. Bioclimatic zones of Greece. Relation Forest Research, Athens, Greece. 63 pp.

McLellan, B. 1990. Relationships between human industrial activity and grizzly bears. Int. Conf. Bear Res. and Manage. 8:57-64.

Mertzanis, G. 1992. Aspects biogeographiques et écologiques des populations helléniques d’ours brun (Ursus arctos L.) cas d’une souspopulation de Pinde: application a la conservation de 1 ’espece et de son habitat. These Université de Montpellier H. 200 pp.

Mertzanis, G., C. Grivas, G. Giannatos, H. Papaioannou, Y. loannidis, D. Bousbouras, C. Poirazidis, and M. Gaetlich. 1990. The status of the brown bear (Ursus arctos L.) in Greece. V. International Congr. of Zoogeogr. and Ecol., Greece and neighbouring countries, Iraklion, Crete. 7 pp.

Ministry of Agriculture, Forestry Div. 1962. Forest exploitation and creation of Wood Factory in W. Rhodopi.

Ministry of Coordination 1979. Special project for the development of the northern part of Drama Prefecture. Part B, Kavala. 281 pp.

Ministry of Finance. 1993. Special Development for the mountainous border zone with Bulgaria. Vol. A, B Athens (INTERREG 4).

Mysterud, I. 1976. Sau drept av bjorn, Ursus arctos, I Engerdal, syd-Norge I arene 1953-1963. Fauna (Oslo) 27(3): 121-138.

Pallis, A. 1925. Statistics of tribal immigrations in Macedonia and Thraki during 1912-1924. Athens.

Perzanowski, K. 1995. Carpathian bears research agreement. IBN(4): No. 2. 13 pp.

Psyhogios, D. and G. Papapetrou. 1995. The movements of pastoralists in Saracatsani: A Hellenic pastoralist population. Proc. of Congress in Serres 1- 3 Oct. 1983. Athens 1985. p. 27-46.

56 Red Data Book of threatened vertebrates of Greece. 1992. Athens. 353 pp.

Reynolds, H.V. and G.W. Garner. 1987. Patterns of grizzly bear predation on caribou in northern Alaska. Int. Conf. Bear Res. and Manage. 7: 59 - 68.

Rogers, L. 1978. Effects of food supply and kinship on social behavior, movements, and population growth of black bears in north-eastern Minnesota. Wildl. Monogr. 97. 72 pp.

Rosier, R. 1989. The status of the brown bear in central and eastern Europe, pp. 15- 25 in Proc. of the workshop on the situation and the protection of the brown bear in Europe. Oviedo, Spain. 12-20 May 1988. Council of Europe, Environ. Encounters Ser., No. 6.

Roth, H. 1983. Diel activity of a remnant population of European brown bear. Int. Conf. Bear Res. and Manage. 5: 223-239.

Servheen, C. 1983. Grizzly bear food habits, movements and habitat selection in the Mission Mountains, Montana. J. Wildl. Manage. 47(4): 1026-1035.

______.1990. The Status and Conservation of Bears of the World. Int. Conf. Bear Res. and Manage. Monogr. Series No. 2. 32 pp.

. (In press). Habitat fragmentation factors and linkage zone considerations for the maintenance of Global bear populations. Intl. Asso. Bear Res. and Manage. 10: 8 pp.

Spassov, N., G. Spiridonov, and K. Georgiev. 1996. Status of brown bear in Bulgaria and the Balkans. Paper presented in the European LIFE Project Meeting. Nympheo Greece. 1995.

Spiridonov, G., and N. Spassov. 1990. Status of the brown bear in Bulgaria. Aquilo Ser. Zool. (27).

Statistical Society of Greece. 1988 and Yearbooks 1933, 1950, 1961, 1971, 1981, 1991.

Theocharis, D. 1971. Prehistory of Eastern Macedonia and Thrace, A.G.S. 9, Athens.

University of Thrace. 1993. Local Development Project. Nestos Valley. Final Rep. Xanthi. 332 pp.

57 United States Dept, of the Interior, Fish and Wildlife Service. 1993. Grizzly Bear Recovery Plan. Missoula, Mont. 181 pp.

Vacalopoulos, K. 1919. The history of the Northern Hellenism. Thrace. 519.

Van der Elst, M., R. Lafontaine, M. Beudels, P. Devilliers, and R. Beudels. 1989. Actions d’urgence visant a I’elaboration d’une strategic de conservation de Tours brun dans le partie orientale et centrale de la Communauté (Grece et Italie). Raport a la Commition des Communautés Europeenes.

Weber, P. 1987. Observations of brown bear movements in the Hargita Mtns., Romania. Int. Conf. Bear Res. and Manage. 7: pp 19-21.

World Wildlife Fund-Greece. 1988. Ecodevelopment Project of Rhodopi and Nestos Valley. Vols. A and B. Athens.

Zervas, P. 1927. Game management. Min. of Agriculture. 275 pp.

______. 1961. The game animals in Greece. 2nd Ed. Min. of Agriculture. 333 pp.

Zunnino, F. 1981. Dilemma of the Abruzzo bears. Oryx XVI (2): pp. 153-156.

Zunnino, F., and S. Herrero. 1972. The status of the brown bear (Ursus arctos) in Abruzzo Natl. Park, Italy. Biol. Conserv. 4(4):263-272.

58 Appendix A. Calculation of the man-made pasturelands

The area of the minimum grassland requirementsfor the nomadic flocks was estimated as follows;

- One sheep needs a minimum 0.1 ha of open grassland per month (WWF 1988).

- Flocks were in the mountains for at least 6 months.

- Sheep grazed in Area I totaled 149,140 (Table 9).

The minimum open grassland area needed for this size of flocks is:

0.1 ha/sheep/month x 6 months x 149,140= 89,484 ha = 894.8 sq.km.