ECOGRAPHY 19 122-138. Copenhagen 1996

Population dynamics (1869-1994), demography, and home ranges ofthe lynx in Bialowieza Primeval Forest (Poland and Belarus)

Wiodzimierz J^drzejewski, Bogumila J^drzejewska, Henr\k Okarma, Krzysztof Schmidt, Aieksei N. Bunevich andLech Milkowski

J^drzejewski W J^drzejewska B Okarmd H . Schmidt K Bunevich, A N andMilkowski L 1996 Population dynamics (1869-1994) demograph>, and homeranges of the lynx in Bialowieza Primeval Forest (Poland and Belarus) - Ecography19 122-138

Population dynamics, demography and home ranges of the Eurasian lynx Lvn\ hn\were studied in Biatowieza Primeval Forest (BPF, 1250 km-), the best preservedmixed and deciduous forest in the lowlands of Europe, 4O'V(i of BPF area belongs toPoland and 60% to the Belarus Republic Results of radiotelemetry of lynx (1991 -1994) were combined with the Polish and Belarussian game departments' inventoriesof lynx numbers (1946-1994), archival huntmg statistics (1869-1989), observationsand snowtracking of lynx

In 1991-1994, 12 lynx were radiocoilared Their home ranges covered from 50 to246 km- (mean 147 km-), depending largely on the time the lynx was radiotrackedDuring a given period, le the autumn-winter seasons (1 October-30 April), thehome ranges were largest in adult males (90-148 km-), then in adult females (82-108km-), and smallest in subadult lynxes (39-55 km-) Home ranges overlappedextensively

In winters 1992 93 and 1993 94, 21 and 29 lynxes respectively, were recorded b>the mapping of radiotracked and snowtracked individuals in the Polish part of BPFOf them, 40"/<. were "transborder" individuals utilising both Polish and Belarussianparts of BPF Winter densities were c 3 adult lynx 100 km"- and 5 lynx 100 km -if kittens were included Adult males formed, on average, 29'/o and reproducingfemales 23"/i of all lynx Subadults and kittens constituted, respectively, \2"/u and 35"'"of the population Sex ratio was 1 1 During the first 3 months of kittens' life, onaverage 3 3 kittens mother were recorded, only I 6 young/mother survived tillindependence Mortality of kittens was at least 48'Mi, and the rate of mortality washighest during the early stage of kittens' life Mean annual reproduction rate of lynxpopulation was 0 59 In the protected population, annual mortality rate of subadultand adult lynx was on average 0 37 Poaching was the most important factorcontributing IVVa to the total annual mortality rate

During the last 125 years (1869-1994), three periods with relatively low harvest oflynx by man and thus with fairly natural functioning of lynx population, wererecorded before 1875 (density 2-3 lynx 100 km"-), in 1920-1959 (4-6 lynx 100km -) and after 1970 (2 -5 lynx 100 km^-) The levels of lynx densities were mostprobably determined by the varying abundance of roe deer Capreolus capreolus andred deer Cerius elaphus (lynx s main prey) in the ungulate community in BPF Twoperiods of near extermination of lynx occurred (1890-1914 and 1960-1970), bothcaused by deliberate persecution of lynx As soon as persecution was abandoned,lynx population recovered rapidly, mainly due to immigration from vast continuousforests in the east and north-east

Review of the long-term data on lynx dynamics in the Palaearctic revealed that inthe Far North-East (Yakutia), the 10-year cycles of lynx and the blue hare Lepustimidus. Its mam prey, were recorded Towards west the cycle period becomes shorter

Accepted 26 July 1995

Copyright C ECOGRAPHY 1996ISSN 0906-7590Pnnted in Ireland - all rights reserved

122 ECOORAPHY 19 2 (W6)

(5-6 years in the Komi region) In the SW regions of the Palaearctic. where lynxrelies on ungulates, lynx numbers are more stable but. periodically also more affectedby man

W J^rzejenski. H Okarma and K Schmidt Mammal Res Inst. Polish Academy ojSciences 17-230 BiatoHieza. Poland - B Jpdrzejenska Workshop for Ecology andProtection of the Natural Eniironment. 17-230 Biakmieza Poland - A V Bunetich.State National Park Beloiezhskaya Pushcha. Brest Oblast. Kamenei Raton 225063Kamenvuki. Belarus Republic - L Milkonski Krzvze 10. 17-230 Bialonieza. Poland

The 200-year population dynamics of the Canadian lynxLynx canadensis in North Amenca has become theclassic example of cycles dnven by the cyclic abundanceof prey, the snowshoe hare Lepus amencanus (Elton andNicholson 1942. Quinn and Parker 1987) Eurasian lynxLrnx lynx, although much less publicised than its Amer-ican congeneric, is still more interesting m this respectIn Its. originally vast, geographic range covering alltypes of Palaearctic forests from the northernmost taigato the steppe woodland, the lynx shows clear latitudinalgradient in its foraging In the boreal forests, lynx huntmainly for the blue hare Lepus iimidus. whereas in thenemoral boreal forests they rely on ungulates(J^drzejewski et al 1993) Moreover, the boreal Siberianpopulations of lynx follow in numbers the 10-year cyclesof blue hare (e g Tavrovskii et al 1971) Southern andwestern populations of lynx hunting for ungulates(mainly the roe deer Capreolus capreolus. seeJ^drzejewski et al 1993). do not show multiannualcycles Instead, in many regions, they are most seriouslyaffected by man hunting policy, predator controls,poaching and availability of forested habitats (e gNovikov 1967) Thus, lynx is a good subject for studyingthe predator population dynamics in the gradient ofprey resources changing from clearly cyclic (hare) torelatively stable (ungulates)

Spatial structure of Eurasian lynx populations ispoorly known Data on lynx home range size are scantyand show enormous variation from 10-20 km- (Heptnerand Sludskii 1972) to 1862 km- (Breitenmoser andHaller 1993). which seems to question the applicabilityof a simple territonality scheme to lynx ecology Incontrast to the well studied Canadian lynx (e g Brandand Keith 1979). demography of Eurasian lynx remainsunknown

In this paper, we analysed population dynamics oflynx in Biatowieza Primeval Forest located on thePolish-Belarussian border It is the best preserved tem-perate lowland forest of its size in Europe By combiningthe results of radiotelemetry of lynx (in 1991-1994).with the game departments" inventories of lynx numbers(1946-1994). hunting statistics (from 1869 until 1989when lynx became protected), observations, snowtrack-ing and all available reports on lynx, we aimed atshowmg 1) the spatial structure of lynx population and

size of home ranges. 2) sex and age structure of popula-tion litter size and mortality of lynx, and 3) dynamics ofharvest and lynx numbers in 1869-1994 We analysedhow man and natural factors had aflected the function-ing of lynx populations dunng the last 125 yearsFinally, we reviewed the available long-term data onlynx population dynamics from Palaearctic to show thegradient from cyclicity in the parts of l>nx range whereblue hare dominates its diet, to potential stability(though disturbed by man) in the parts, where lynx reheson ungulates

The study was part of a long-term research on preda-tor-prey relationships in Bialowieza Primeval Forest(eg Jgdrzejewski et al 1992. J^drzejewski andJ^drzejewska 1993. J^drzejewska et al 1994) Feedingecology of lynx and its choice and utilisation of ungulateprey was described in detail by J?drzejewski et al (1993)

Study area

Bialowieza Pnmeval Forest (BPF 52°30'-53°N23°30'-24°15'E) located on the Polish-Belarussian bor-der. IS a vast woodland connected continuously withother large forest tracts (Pruzhana Forest in NE. Shere-shevo Forest in SE. Knyszyn Forest in NW) (Fig 1) Itlies in the boreal nemoral zone, and is composed of richmultispecies tree stands In its historical borders. BPFcovers 1 250 km-, but in the recent decades its adminis-trative coverage increased to nearly 1 500 km- due toamendance (in both Polish and Belarussian parts) ofperipheral woods and woodlots. mostly of secondaryorigin

From the 15th to the end of 18th century. BPF wasprotected as the royal hunting forest of Polish kings Inthe 19th centur>' and until 1914 (under the Russian rule).It became a protected forest for monarchial hunts of theRussian tsars Industrial scale exploitation of timber inBPF was begun only in 1915 (dtiring World War I) byGerman occupants After WW I. Polish state forestryand the English company "The Century European Tim-ber Corporation" continued exploitation In 1921. themost valuable and nearly untouched oldgrowth (47 5km-) located in the heart of BPF was proclaimed asBiaJowieza National Park The remaining part of BPFunderwent economic exploitation of timber until 1941

ECOGRAPHY 19 2(1996) 123

Poland Belarus

forests

swamps

cities

towns

villages

state border

Poland Belarus

- state border

Browsk - forestNikorskoe administration

units

- „ - forestcompartments

C^ - towns^ and villages

0 2 4 6 8 10 km

Fig I Upper panel Bulovvieza Primeval Forest among the woodlands of eastern Poland and western Belarus Lower panelschetridlic map of BPF with division into forest administration tinits

124 EtOORAI'HV 19 2(19%)

In 1945 the new state border between Poland andthe Soviet Union (Belarussian SSR) divided the BPFinto two parts (Fig 1) with different management rulesIn 1981 the Soviets built a 2 5-m high wire fence alongthe border, and it became a real barrier to ungulatemovements Lynx and wolves can cross the border

The Polish part of BPF (580 km-) includesBialowieza National Park (BNP. 47 5 km") and theexploited forests administered b\ the State Forestry(530 km-) BNP is a strict reservation where no hunt-ing, timber exploitation and motor transportation isallowed The iich deciduous tree stands dominated b\oak Queicus lohui, lime Tilia coidata and hornbeamCaipinus betidiis cover 48' i of the park's area Mixedconiferous forest dominated b> pme Pmits sikestiis andspruce Puca ahies (with admixtures of oak) cover 17"''oand wet alder Almi.s gtutmosa and ash Ftaxmus excel-stoi forests - 1 &"/n The average age of tree stands is 130years In the 197O's the park became UNESCO s Man& Biosphere Reserve and World Heritage Site

The exploited forests on the Polish side of the borderare managed b\ means of small clear cuts and selectivecutting of large trees, and replantation Mixed conifer-ous forests dominated by pine and spruce (with admix-tures of oak) cover 54% of the area of exploited forestalder and ash forests 2O'/'(i, and oak-lime-hornbeamforests 13'y(i Mean age of tree stands is 72 yr

The Belarussian part of BPF (670 km-) has beenpartially protected since 1945 Cutting is small andselective (dead and dying trees only, no cleaicuts)Coniferous forests dominated by pme and mixed conif-erous forests composed of pine and spruce (with admix-tures of oak) cover 69% of area, oak-lime-houibeamforests 6"A,, and wet alder-ash forests 16'! ) The averageage of tree stands is 100 yr In 1991. the whole Belarus-sian part of BPF became the State National Park (withsome strict reserves and mostly with partial protectionregime) and the UNESCO's Man & Biosphere ReserveMore detailed information about vegetation of BPF isgiven by Fahnski (1986)

BPF harbours a rich community of ungulates consist-ing of European bison Bison hoiiasus moose Akesalecs, red deer Ccivus elaphus. roe deer Capreoluscapieohis, and wild boar Su.s scioja Large predatorsinclude lynx and wolf Cams lupus The native brownbear Uisus aictos was exterminated m the 2nd half ofthe 19th century Few reintroduced bears lived in BPFfrom 1938 till 1950 (Buchalczyk 1980. Bune\ich un-publ )

History of lynx management in BialowiezaPrimeval Forest

In BPF. lynx was neither protected nor particularlypersecuted until the 2nd half of the 19th century. Aselsewhere m the Polish-Lithuanian Commonwealth, it

was valued for its fur and hunted sustamably b\ thelocal game wardens Jarocki (1830). after having visitedthe Forest and talked to the wardens, reported 'Thelynx (Fehs lynx) although appears here quite often, is atransitive rathei than native inhabitant of the Forest"In the i860's the Russian administration of BPF beganthe 'modern' game management (Karco\ 1903) In1869 predator control (aimed at extermination ofbrown bear wolf and hnx) began and lasted until 1914In 1869-1876. from 14 to 23 drive hunts against wolveswere organized annually, during which all lynx seenwere also shot In 1877-1882 additionally, strychnineballs were laid out in the Forest to poison wohes andKnx (Karco\ 1903) Since 1889 bounties were paid forkilled predators First (1889-1890) bounties wererather high (20 rubels for a Knx) then (1891-1895)much lov\ei (3 rubels) which made the predator exter-mination markedh less effective In 1896. bountiesraised to 10 rubels for a hnx During that period thesalar> of forestn ddministiation officer was 203 lubelsper year and that of a game v\aiden 93 rubels (Karcov1903) Complete extermination of wolves and I>nx fromBPF was difficult because they continuousK immi-grated from surrounding forests, but the campaign was,nonetheless, effective enough Karco\ reported m 1903Game wardens no longer find ungulates mutilated b\

wolves and lynx'Detailed game management reports are not available

from 1903-1914 but it is knov\n that extermination ofwolves and lynx as a measure of protection and promo-tion of ungulates was conducted til! 1915. when Ger-man army invaded Biatowieza During 1915-1918.German administration of BPF focused at fast andlarge scale exploitation of timber Game was practicallynot protected and poaching (especially b) the end ofWW I and during the Pohsh-Soviet wai in 1919-1920)was disastrous for ungulates (J^drzejewska et al 1994)Beginning from 1915. tracks of lynx and wohes wereoften seen in BPF (Voit 1917) In 1919-1939. the wholeBPF was again under the Polish rule The Forestbecame a place of diplomatic hunts for the Polishpresident invited politicians and monarchs (Wiecko1984) Lynx were valued trophies of the hunts In1939-1941 BPF passed under the Soviet rule, and in1941-1944. It was established the 'Reichjagdgebiet' b}the Nazi occupants By the end of World War II. thenumbers of wohes and hnx were very high

After 1945, when the new Polish-Soviet border di-vided BPF game management rules were different meach part In the Belarussian part, from 1946 till 1955.lynx was a subject of sustainable hunting harvest In1956-1980. It was treated as pest but the deliberatepersecution occurred in 1956-1967 only, mainly bydrive hunts like those in the second half of the 19thcentury In 1981-1992. lynx was a subject to occasionalsport hunts In 1993. it was included into Red DataBook of the Republic of Belarus as a rare species

ECOGRAPHY 19 2 (1996) 125

(Chyrvonaya Kniga Respubhki Belarus 1993) In thePolish part of BPF. the lynx was fully protected in 1946-1953. hunted (under a licence system) in rather smallnumbers in 1954-1988. and again protected since 1989

Methods and material

Radiotracking

From February 1991 through March 1994. in the Polishpart of BPF. 12 lynxes were captured with spring-pow-ered footsnare traps at fresh kills or marking places (12captures, including 2 recaptures to change collars) andin a large double-door box trap (3 captures, including 1recapture) All traps had radio alarm systems thatsignalled the capture of an animal and shortened thetime lynx spent m a trap to 1-2 h A captured lynx wasimmobilized with a 0 4 - 1 0 ml of KetaminXylazinmixture (583 mg Bayer "'Rompun" dissolved in 4 mlParke-Davis "Ketavet 100 mg 1 ml") (Seal and Kreeger1987) We determined the sex. age (based on toothwear), and weight of a lynx, and fitted it with aradiocoUar (150-200 g) After handling, the lynx wasinjected 0 4 - 1 0 ml of Effortil to help recover thenormal heart action Data on collared lynx are given inAppendix I

Five to 7 days a week we tried to locate the radio-marked lynxes from the church tower (33 m high) inBialowieza village The audibility range of signal fromthe tower was 6-10 km After getting an approximateposition of a lynx we localised it precisely from theground (signal range 1-3 km) Searching for lynxes notpreviously located from the tower was conducted bydriving the system of forest roads (see map in Fig 1) In1991-1993. the radiotracking was done in the Polishpart of BPF (few localisations from the Belarussian partmade from the church tower are included) In the winterof 1993 94. once a month, a search of radiocoUaredlynxes was also conducted m the Belarussian part Theterrain of BPF is flat but the dense, not easily accessibleforest made our efforts of daily locating of lynxessuccessful on 19'/i)-50"/,, (on average 41'^)) of all days thelynxes lived in BPF and wore active collars In 1993-1994. in addition to daily locations, a continuous 5-dayradiotracking (with locations at 30-mm intervals) of onerandomly chosen lynx was conducted once a month

For estimating home ranges, all daily localisations ofthe autumn-winter seasons (1 October-30 April) andthe whole periods of radiotracking of each lynx wereincluded From the series of continuous tracking, onelocation per day was taken if lynx was more or lessstationary, and two extreme localisations if it movedHome range size was calculated by Minimum ConvexPolygon method using a program Tracker (A Angerb-jorn. Sweden) Detailed analysis of home range shapeand size variation will be presented elsewhere

Snowtracking

Snowtracking was a supplementary method of estimat-ing lynx numbers and densities Snowtracking of indi-vidual lynxes was done in the Polish part of BPF. first(1991/92) to locate non-collared lynx for livetrappingIn 1992/93 and 1993/94. all lynx tracks noticed duringany field work were recorded Some trails were followedfor few kilometers The total length of snowtracking ofnon-collared lynx was 20 km in 1991,92. 90 km in1992/93 and 230 km in 1993 94 There were c 75 dayswith snow cover in 1990/91 and 1993/94 and c 52 daysm 1991/92 and 1992/93

We visited the central part of BPF more often thanperipheral parts The numbers of lynxes walking to-gether and their marking behaviour were recorded Sizeof the foot print and the pace length were assessedvisually Snowtracking of radiocollared lynxes ofknown sex and age showed that adult males had dis-tinctly larger footprints and a long pace Moreover,most of the captured lynx were snowtracked beforecapturing In all such cases, the sex, age determined bysnowtracking proved to be correct when an individualwas trapped Tracks of radiocollared lynxes were distin-guished from those of non-collared individuals by ra-dio-localisations done concurrently with snowtrackingAdult females were determined as such when leadingyoung, the number of kittens was also counted duringsnowtracking Lonely adult females could be difficult todistinguish from adult males by their tracks However,adult females with no kittens were obviously very rare(no such females were recorded among captured indi-viduals) For density estimation, all records of radio-tracked and snowtracked lynxes were mapped for eachwinter, with an attempt to distinguish all individualspresent in the Polish part of BPF

Additionally, we used data on the snowtracking oflynx in Biaiowieza National Park collected in 1956/57-1961,'62 (courtesy of A Kawecki). and 1985/86-1991/92 (see details m Jfdrzejewski and J?drzejewska 1993)Lynx were surveyed on the grid of transects covering47 5 km- (entire BNP) in 1956-1962 and 11 2 km^ (SWpart of BNP) in 1985-1992 The mean index (N trackskm" ' d~') for the entire study area for each winter wascalculated and used for estimating the lynx densitiesaccording to the method by Pnklonsky (1965) Fordetailed description of rationale of Pnnklonsky'smethod see J?drzejewski and J^drzejewska (1993)

Long-term dynamics of lynx population - sourcesof data and methods of inventories

Information on predator control conducted in 1869-1902 was taken from Karcov (1903) Game departmentdocuments from 1920-1939 archived by the PolishState Forestry Administration were destroyed by Soviet

126 ECOGRAPHY 19 2(1996)

managers in 1939 Only few information were derivedfrom the articles on diplomatic hunts published in theforesters" monthly journal (Anonymous 1937d.b. 1938)We are not aware of any German game archives from1941-1944 Data from 1869-1938 are listed m Ap-pendix II

In 1946-1994, hunting quotas and lynx number esti-mates in the Belarussian part of BPF were archived bythe Game Management Department of the State Na-tional Park Belovezhskaya Pushcha (before 1991 GameHunting Preserve Belovezhskaya Pushcha) Each win-ter, the inventories of lynx were done by game wardensAfter new snow fall, the snowtracking on all accessibleforest compartment lines (1066 x 1066 m = 1 x 1 verst)was conducted twice (usually on two consecutive daysto get a reliable estimate) Tracks of lynx crossing thelines and their direction were noted Then, all trackswere mapped and. by drawing the reconstructed routesof lynx movements, the places of lynx daily rest siteswere determined (l e forest compartments, where atrack went into but not out of it) Some of the data onlynx numbers were published by Filonov (1989)

In the Polish part of BPF, data on lynx numbers andhunting quotas were collected and archived by theGame Department of the Bialowieza State Forest Ad-ministration, but only the archives from 1958-1994persisted Some earlier information (1946-1950) werefound in the typescript of memoirs (archived at theLibrary of Bialowieza National Park) of the late direc-tor of Game Department. Edmund Wagner In thePolish part of BPF, inventories of lynx (conducted bygame wardens) were done by snowtracking

Because of the political isolation in 1945-1990. therewas no exchange of information about predator num-bers and hunting policy between the respective GameDepartments in the Polish and Belarussian parts ofBPF Thus, the data from the two parts can be treatedas independent All data from 1946-1994 are listed inAppendix III

Totally, dunng the 125 years, the estimates of lynxnumbers in the whole BPF or part of it, are availablefor 51 years {4\"/n), and data on lynx hunting statisticsfor 83 years (66'>;i) Reconstruction of lynx populationsize was possible for nearly the whole series of 1869-1994, based on regression between number of lynxkilled during the years of persecution and estimatednumbers of lynx and between independent censuses oflynx in the Polish and Belarussian parts of BPF (seedetails in Appendices II and III)

Demographic parameters

Information on females with kittens (from snowtrack-ing and visual observations) were gathered by the Be-larussian game wardens and scientists during any fieldwork Available archives of their reports covered years

1947-1951 and 1982-1994 In the Polish part of BPF,data gathered in the same way by State Forest gamewardens, wardens of BNP, and zoologists covered years1957-1959 and 1986-1994 One earlier information(from the 1930"s) was derived from a documentaryphotograph taken of lynx litter by J J Karpihski(Anonymous 1937b) Totally, 61 observations (visualobservations or snowtracking) of 47 litters were docu-mented (if any litter was observed more than once, theearliest observation and the latest one were taken)

Data about non-harvest mortality of lynx was col-lected in the Polish part of BPF during radiotrackmg oflynxes and by autopsies of any dead lynx reported tous by forestry services (full data in Appendix IV)Annual mortality rates were calculated for the radio-tracked sample of lynx population in 1991'92-1993 94according to the method by Heisey and Fuller (1985)

Results

Spatial structure of lynx population in 1990/91-1993/94

During winter 1990 91. only two adult males wereradiotracked, Bazyli and Makary Their ranges over-lapped considerably (Fig 2)

In winter 1991 92, we captured 3 lynxes An adultmale Borys, trapped north of BNP. most likely visitedthis area during the mating season only Later he wasrecorded mainly in the western part of BPF (Fig 2) Ajuvenile male Bazyhszek. captured in early February,stayed with his mother and a sibling till March 21, sowe could estimate the home range of his mother Ajuvenile female Mala, trapped in December, was proba-bly a daughter of a female that had led 2 kittens andwas shot in the Belarussian part of BPF (near theborder) in October/November 1991 Makary was notradiotracked dunng that winter because of collar fail-ure, but tracks of a single male were recorded in hisformer home range We interpreted those as Makary'stracks, because in the following year, he was recapturedm the same area Bazyli was found dead in a snare inlate December 1991, and Mala disappeared from thestudy area by the end of January 1992 Additionally, amale and a female with two kittens were recorded bysnowtracking during 1991/92 (Fig 2)

In winter 1992'93, we radiotracked 7 lynxes, includ-ing 4 newly captured ones Adult male Borys andsubadult male Bazyhszek had overlapping territoriesand they were occasionally found close to each otherIn December 1992, Borys was poached After a femaleNatasza had been radiocollared in February 1993,Bazyhszek was recorded to accompany her severaltimes Their home ranges overlapped considerablyMakary stayed farther east than in 1990/91 (Fig 2) Heoften accompanied females Sonia and D, the latter one

ECOGRAPHY 19 2(1996) 127

Winter

\

[

0 2 4

1990/

\^ 7 ^

6 8 10

' Makan/rf * K /

cfBazylicf MakBCV

24 Feb-30 Apr 9110 March - 30 Apr 91

cfMakaiYrfBazylicf Bor/s$ rvlala9 C + 2 |uv

(CfBazyliszek

9March-XApf 911 Oct -8Feb-11 Dec9 Dec

* NN)

23 Dec 9130 Apr 9291 -21 Jan 92

91 - 28 March 92

a9BA + 2 juv

31 Jan21 Jan

92- 3 Feb 92

rfBorys<3 Bazyliszekcf Nikifor( f Makary9 D.t 9Zo^Juv9 Sonia9 Natasza ^ 1 }uv

1 Oct - 7 Dec 921 Oct 92 - X Apr 9326 Jan -23 Feb 9327 Feb - X Apr 9329Dec 9 2 - X A p f 9312 Jan - X Apr 933 Feb - 30 Apr 93

rfLrfH

16 Dec 92-31 March 9329 Dec 92-3 March 932 Jan 93

rf Dvfupalcy 9 Feb - X March 939 E<9 J +9 K ^

. liuv 14 Dec 92 -3 Feb 932)uv 19 Feb 93

. 2|uv 3 - X March 93

rf Bazyliszek 1 Oct 93-21 Nov 94rfMakary 1 Oct 93 - X Apr 94rf Igor 22 March - X Apr 949 Natasza+ljuv 1 Oct 93 - 28 March 949 Zoja 1 Oct 93 - X Apr 949 Sonia 1 Oct 93 - 30 Apr 949 Tamara 17 Nov 9 3 - X Apr 949 Olga.|.1|uv 16 Feb - X Apr 94

128

rfS 12 Nov 93-8 March 94rfM 16 Nkjv 93-18 March 94rfN 16 Nov 93-21 March 94cf T 4-28 March 94d Dwupalcy 7 - 8 March 949 P+2 |uv 1 Nov 93 - 28 March 94O 2 7 Nov 93-4 March 949 R+2 |uv 16 Nov 93 -10 March 949 W+2 )uv 24 Nov 93 - 9 March 949 U+3 |UV 3 Feb - 8 Apr 94

ECOORAPHY 19 2 (1996)

Table 1 Home ranges of lynx in Biatowieza Primeval Forest estimated by the Minimum Convex Polygon with lOOVnlocalisations Life-time home range total area utilised by a lynx dunng the whole penod of radiotracking (dates given mAppendix I) Winter home ranges size of area utilised by a given lynx during 1 October-30 April (calculated only for individualsthe radiotracking of which dunng any cold season spread >3 months and included 25 or more localisations) Numbers oflocalisations in parentheses

Lynx sex, age atcapture, andname

Life-time homerange (km*)

(N localisations)

Winter home range size (km-)

1990/91 1991'92 1992 93 1993'94

M, ad, BazyliM, ad, MakaryF, ad. Mother of

Bazyliszek'M, subad.

Bazyliszek-M, ad, BorysF, subad, SoniaF, subad, ZojaF, ad, NataszaF, subad, TamaraM, subad, Igor

222 2 (152)245 9 (200)

99 2 (38)

185 6(134)239 9 (58)118 7 (255)49 7 (54)

158 2(209)73 1 (94)76 3 (28)

148 2 (36)90 3 (25)

_

_

99 2 (38)

91 2 (68)

55

38

82

1

5

1

(32)

(62)

(67)

108 447 198 149 0

(131)(26)(55)(41)

' This female was not radiocoilared Her home range was calculated based on radiolocations of her offspring (Bazyliszek) whileIt had stayed m a family group

- Included are only localisations taken after he had become independent

leading a radiocoilared kitten Zoja Adult male Nikiforwas radiotracked for one month (25 Jan-23 Feb) anddisappeared Ten months later he was killed as rabiedon a farm yard in Knyszyn Forest, 60 km N from theplace of capture Probably, he came to BialowiezaForest only for the mating season, when he wasrecorded together with the female E leading a kitten

By snowtracking, we mapped 4 males and 3 femaleswith kittens Male H was recorded once, as he camefrom the Belanisian part (Fig 2) He was distinguishedfrom Makary by obviously larger tracks and pacelength Another male lacked two toes on his foot print(Dwupalcy = two-toed in Polish) His tracks were oftenrecorded together with the tracks of Natasza andBazyliszek (Fig 2)

In winter 1993/94, 8 lynxes were radtotracked, in-cluding 3 newly trappwd ones Makary was found in thesame area as in the former winter He visited femalesSonia, Zoja, and a newly captured female Olga with akitten Home ranges of subadult Zoja and adult Olgaoverlapped extensively Three lynxes were radiotrackedin the western part of the Forest (Fig 2) The range ofmale Bazyliszek overlapped that of female Natasza witha kitten In the end of December 1993, Bazyliszek waspoached Afterwards, Natasza moved north-east InMarch, she was observed without a kitten and wasleavmg blood on her trail No prey of her own werefound, but she scavenged on prey remams of female R

with 2 kittens In late March 1994, we lost contact withNatasza A subadult male Igor was trapped in thecentral part of the Forest in late March 1994 Later hemoved to the NW part and stayed there The homerange of a subadult female Tamara was located mostcentrally

Additionally, snowtracking revealed the presence of 5males, 4 females with kittens, and one lonely individualwith distinctly small tracks (lynx Z in Fig 2) MaleDwupalcy was recorded on 2 days only, he probablycame from the Belarussian part for the mating season

In winters 1992/93 and 1993/94, 43% and 38% (onaverage 40%) of all lynxes recorded in the Polish part,respectively, were found either on the state border or inthe Belarussian part of the Forest (Fig 2)

Home range size

For 10 lynxes. Minimum Convex Polygon home rangesestimated for the entire periods of radiotracking cov-ered, on average, 147 km' (SD 73) (Table 1) Theaverage home range of males was 194 km- (SD 70,n = 5) and that of females 100 km^ (SD 42, n = 5)However, the size of home range increased with thenumber of days the lynx wore active collars (Y =80 8 - 0 23X, n = 10, R^ = 0 473, p = 0 028) The abovegiven estimates of home range size corresponded to an

Fig 2 Spatial distnbution of lynx in the Polish part of BPF Solid lines (and all names of lynx except of Dwupalcy) denoteradiotracked lynxes (thick lines - males, thin lines - females) Penods of radiotracking (and snowtracking if tracks of a givenlynx were recorded pnor to capture) given under each map Minimum Convex Polygon home ranges are plotted for allradiotracked lynxes irrespectively of the numbers of localisations Broken lines (lynx Dwupalcy and lynxes denoted by letters)denote non-collared individuals recorded by snowtracking (thick hnes - males, thin lines - females) Penods of snowtrackinggiven under the maps Polygons of 'ranges' embraced all tracks that were unambiguously prescnbed to a given lynx

ECOORAPHY 19 2 (1996) 129

Table 2 Winter densities of lynx in BPF. estimated by various methods (see Fig 2 for distribution of recorded lynx) MCP -minimum convex polygon Methods (1) and (2) yielded underestimated measures of density, whereas method (3) is most reliableIn method (3). the numbers of lynx in early winter and late winter (the latter m parentheses) are given

Method

(1) N adult lynx per wholePolish part of BPF

N lynx (lncl kittens) perwhole Polish part of BPF

(2) N adult lynx corrected b>transborder lynx per wholePolish part of BPF

N lynx (lncl kittens)corrected by transborderlynx per whole Polish partof BPF

(3) N adult lynx per MCP areautilised by them

N lynx (lncl kittens) perMCP area utilised by them

Winter

1992 931993 94

1992 931993 94

1992 931993 94

1992 931993 94

1991 921992 931993 94

1991 921992 931993 94

N lynxrecorded

14!8

2129

11145

16523 5

6 (6)14(12)18 (16)

11 (10)2! (19)29 (27)

Area (km-)

580580

580580

580580

580580

2163430 7562 0

216 3430 7562 0

Density(N 100 km--)

2431

3650

1 925

284 0

2 8 (2 8)3 2 (2 8)3 2 (2 8)

5 1 (4 6)4 9 (4 4)5 2 (4 8)

average of 360 days m males and 215 days in femalesThe trend of increasing size of home range with longertime of radiotrackmg occurred both in males and fe-males, but the regressions were not significant due tothe small sample sizes (5 male and 5 female ranges)

During the autumn-winter seasons (l e equal periodsof radiotrackmg). the ranges were smaller, and theirsizes were largely determined by lynx body mass (Y =-118 1 + 12 OX. n = l l , R- = 0 58. p = 0 006) Adultsof both sexes held large ranges, on average 110 km- inmales (SD 33 n = 3 ranges) and 97 km* in females (SD33 n = 3) Subadults had much smaller ranges, in fe-males on average 45 km- (SD 6. n = 3). and m a male55 km^ (one case only) (Table 2) Average autumn-win-ter home range of all lynx was 82 km- (SD 33)

Winter densities of lynx in 1991-1994

During the four years of livetrapping and radiotrackmgof lynx, only a part of BPF area was censused and anunknown proportion of lynx population was recordedThus, we estimated the density by 3 methods 1) totalnumber of recorded lynx divided by the whole area ofthe Pohsh part of BPF, 2) the number of recorded lynxcorrected by the percent of 'transborder' lynxes (l e by2O'!4. lower) divided by the whole area of the Polishpart of BPF, and 3) total number of recorded lynxdivided by the area covered by their home ranges (thatarea was estimated as Minimum Convex Polygon) In

all three methods, density was calculated first for adultand subadult lynx and then for all lynxes (kittensincluded) (Table 2)

The densities of adult lynx were from 1 9 to 3 2 lynx100 km -, and those of all lynx (kittens included) from2 8 to 5 2 inds 100 km"- (Table 2) In winter 1993/94,when the study was most intense and covered thelargest area, the 3 methods yielded similar densityestimates The most reliable estimate (method 3) gavesimilar results in all years 2 8 - 3 2 adult lynx 100 km"-and c 5 lynx 100 km"- including kittens These areestimates for beginning of each cold season If all casesof mortality and disappearance are considered, thedensities (representative of late winter) were 2 8 adultlynx 100 km^^ and 4 4 -4 8 lynx 100 k m " ' includingkittens (Table 2)

By extrapolating the density estimates on the wholeBPF, we can assess that during 1992-1994, 1 250 km^harboured 24-35 adult lynx and 35-62 lynx includingkittens In the same years, the inventories by gamedepartments reported 29-44 lynx, if Polish and Be-larussian estimates are summed, or 27-41 lynx, if acorrection for double-counted transborder lynx is made(l e numbers of lynx in the Polish part of BPF loweredby 20%) Therefore, the game departments' estimatesare within the range of densities calculated from radio-tracking and snowtracking Both game departments'inventories and radio- and snowtracking showed aWyi<-30"/a increase in adult lynx numbers in 1992-1994

130 ECOORAPHY 19 2(19%)

Table 3 Sex and age structure of lynx population m BPF in 1990 91-1993 94 R radiocoilared lynx, S - lynx recorded bysnowtracking (see Fig 2 for distribution of all lynx) Mean bod> masses of lynx of each age sex group (in kg) are given inparentheses (data averaged from live-captured lynxes)

Winter

1990 91I99I 92

1992 93

1993 94

Mean percent(1992 93-1993'94)

RRS

RS

Total

RS

Total

Ad, M(212)

T

22

34

7 (33';',,)

25

7 (24",,)

29"'!.

Recorded

Ad F(157)

2

14

5 (24"',,)

24

6 (21"o)

23" „

numbers of

Subad(15 0)

2

2 (10"o)

4

4(14",,)

12';,,

Ju\(9 0)

23

16

7 (33';,,)

1111 (38"'o)

35",,

Undet

11 (3"',,)

1';,,

Sex and age structure, reproductive parameters,and mortality rate of lynx population

Data on sex and age of lynxes radiotracked and snovv-tracked in winters 1990/91-1993 94 is given in Table 3In 1992 93-1993/94, adult males formed on average29% and reproducing females 23'/'ii of all lynx Subadultindividuals constituted 12% and kittens 35% of the lynxpopulation (Table 3) Sex ratio was 1 1 Mean numberof young was 1 4 kittens per reproducing female (1 2kittens female including subadult non-reproducing fe-males) in 1992/93, and 1 8 kittens per reproducingfemale (1 2 kitten/female) in 1993 94 Totally, in 1991-1994, 13 females were leading 21 kittens, 1 e 1 6 kittensreproducing female One orphan kitten. Mala, wasrecorded

Long-term data on lynx litter size showed that duringthe first 3 months of kittens' life, there were on average3 3 young mother and the rate of mortality was higherdunng the early period of kittens' life than during theirolder age (Fig 3) Mortality of young before indepen-dence amounted to 48'M), but it may be somewhatunderestimated, because snowtracking and visual obser-vations could not reveal females that lost entire littersespecially during the first weeks of kittens' life

Mortality of subadult and adult lynx was estimatedm the Polish part of BPF Totally, in 1978-1994. 11cases of mortality were recorded (Appendix IV), includ-ing 6 poached lynxes (54'KJ), 1 shot after it had come tothe farmyard, 1 rabied, 1 dead from gun-wounds, and 2from unknown causes Natural factors were responsiblefor 9% only (or 27"/. if unknown causes of deaths wereincluded) of the recorded deaths On average annualmortality rate of radiotracked lynx was 0 372, in otherwords 37yo of subadult and adult lynx died each year(Table 4), 71% of that mortahty was due to poachingNatural mortality was low, on average 5% per year

Total mortality after 3 yr was 0 763 1 e only c 24"/(, ofinitial number of lynx survived 3 years (Table 4)Annual mortahty rates were compared to the annualrate of increase due to reproduction (calculated asnumber of young produced per year by one individualin a population) On average, the annual increase ratewas 0 593 (Table 4) Thus, during the three years of ourstudy when no hunting harvest took place, the increaserate overcompensated for the mortality

Dynamics of lynx harvest and population numbersin 1869-1994

In 1946-1994 (49 years), three independent counts ofannual numbers of lynxes in BPF were done in theBelarussian part (data for 44 years) in the exploitedforests of the Polish part (42 years), and in BiatowiezaNational Park (11 yr) (Appendix III, Fig 4) The resultsof these censuses are correlated, thus verifying thereliability of lynx number estimates (numbers of lynx inthe Belarussian and Polish parts r = 0 59 n = 38 years,p < 0 0005. numbers in BNP and Belarussian partr = 0 75, n = 11 p = 0 008, numbers in BNP and ex-ploited forests of the Polish part r = 0 55, n = 10, ns)

Huntmg for and persecution of lynx in the Belarus-sian part of BPF affected its numbers in the wholeBialowieza Forest During the period of the most in-tense persecution, lynx in the Polish part declined ear-lier and recovered later than in the Belarussian part,which adjoins other continuous forests inhabited bylynx BNP, a strict reserve, where huntmg was neverpermitted, is too small (47 5 km-) to serve as a refugefor lynx Numbers of lynx in BNP underwent dramaticfluctuations hnked with those in the Belarussian part

ECOGRAPHY 19 2 (1996) 131

4-0(0E£ 3-mc£ 2-z

1 -

n-

(3 3)•

(2 1)

. . . . . . . .

*(1 8)

" 1 -

(1 6)•V

1-100 101-200 201-300 > 300Days of kitten's life

Fig 3 Changes m lynx litter size from the kittens" birth tillindependence, 61 records (snowtracking and visual observa-tions) of 47 litters m the Polish and Belarussian parts of BPFDay 1 = May 15th Bars are mean values for each 100-daypenod (numerical values in parentheses) Vertical lines +SDDots represent individual sample points

of BPF and the exploited forests of the Polish part(Fig 4)

During the last 125 yr, three periods of high numbersof lynx were recorded before 1875, in 1920-1959 andafter 1970 (Fig 5) These penods were charactensed byrelatively little harvest of lynx population by manDensity levels vaned among these penods The recon-structed numbers mdicate the density of 2-3 lynx 100km~^ before 1875 Game inventories showed 4-6 lynx100 km-- in 1920-1959, and 2-5 lynx 100 km^^ after1970

Two penods of near extermination of lynx occurred(1890-1914 and 1960-1970), both caused by dehberate{persecution of lynx in order to protect ungulates andother game Dunng the most severe hunting, up to 1lynx 100 km^^ were killed annually (Fig 5) As soon aspersecution was abandoned, lynx population recoveredrapidly Beside reproduction of the local survivors, theintense recolonisation of BPF by lynx from the east andnorth-east must have taken place, because during 5 yr

(1967-1971) lynx numbers increased from 2 to 60 (30times) Given the average annual reproduction rate(0 593), the local survivors could have increased from 2to 13 individuals, only

Discussion

Home ranges and demography of lynx

In Bialowieza Forest, the mean home range surveyedfor longer time (c 1 yr on average) covered 147 km-,and basically depended on the time when lynx woreactive collars The home ranges estimated during theequal length periods (autumn-winter seasons, on aver-age 82 km- ) correlated mainly with the body masses oflynxes Goszczyiiski (1986) reviewed the available infor-mation on wmter home ranges of Eurasian lynx cen-sused by snowtracking and gave the mean figure of71 km- The home ranges of radiotracked lynx in thereintroduced population in the Swiss Alps (Berne re-gion) were from 39 to 450 km- (Breitenmoser andHeller 1993) Variation in the home range size of 7 lynxwere in 71% determined by two factors, time the lynxwere radiotracked and body mass (calculated by usfrom Table 1 in Breitenmoser and Haller 1993) In theVosges Mts (France), radiocollared lynx held ranges of98 to 516 km^ (Herrendschmidt et al 1986) In seemsthat space use pattern of lynx is not based on strictterntonality It is a more labile structure of 'drifting'home ranges Thus, to conduct reliable comparisons ofhome range size from various studies one must stan-dardise the time of radiotrackmg

Reproductive parameters of Eurasian lynx were sofar established only on the basis of carcass autopsiesKvam (1991) showed that Scandinavian lynxes had onaverage 2 5 embryos (SD = 0 53) and that 63% of fe-males were sexually active In BPF, c 75% of fe-males > 1 yr old were reproducing

In Bialowieza, man-caused mortality, especiallypoaching, in the protected population of lynx was veryserious Lynx get into snares set for ungulates Dra-matic increases of poaching of ungulates for meat were

Table 4 Annual mortality rates (calculated from radiotrackmg data according to the method by Heisey and Fuller 1985) andreproduction rates (estimated from the proportion of kittens in the population) of lynx m BPF in 1991-1994 Minimum estimateof mortality included only evident cases of death, maximum estimate included also animals that disappeared from the study area

Penod

February 1991-March 1992Apnl 1992-March 1993Apnl 1993-March 1994

Mean annual rate(SD)Total mortality after 3 yr

min

0 3880 2540 299

0 314(0 068)0 680

Rate of mortality

max

0 6260 2540 413

0 431(0 187)0 836

averaged

0 5070 2540 356

0 372(0 127)0 763

Rate ofnatural

mortality

000 149

0 050(0 086)0 068

Rate ofmortality

due topoaching

0 3880 2540 149

0 264(0 120)0611

Rate ofincrease

0 6670 5000 611

0 593(0 085)

132 ECOGRAPHY 19 2(1996)

Belarussian part of BPF

50

40

30-

sport hunting spon hunting protection

I1946 48 50 52 54 '56

Polish part of BPF

70 72 74 76 78 80 82 84 86 88 90 92 94

sport hunting protection

1946 4« 50 52 54 56 58

Year— number ot lynx estimated by game inventory— reconstructed numbers of lynx

numbers of lynx in Biatowieza National Park• lynx harvested

Eig 4 Dyndmics of lynx numbers (censused by snowtracking). and hunting harvest of lynx in the BeUrussian and Polish partsof BPE in 1946-1994 (data in Appendix III)

recorded in times of political chaos and economic re-gression that affected local people Recently, an expo-nential increase of numbers of poached ungulates foundby game wardens has been recorded after 1989 in thePolish part of BPF (Okarma et al 1995) Similarly, inthe Spanish population of fully protected Lynx pardmaannual mortahfy rate was 0 37. and 50% of that mortal-

ity was due to poaching and another 17% due fo trafficaccidents (Ferreras et al 1992)

Prospects of lynx survival in Biatowieza are goodbecause the Forest joins other large forest tracts in theNorth (Knyszyn Forest) and especially m the East(vanous Belarussian forests), which are populated bylynx and where lynx is protected

Lynx harvested

•,••. • • "!• 1

1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

O 5-

03 . I

3 3-

Lynx population density

I860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

Year

Fig 5 Long-term dynamics of iynx density and hunting harvest in BPF (1869-1994) Density in 1869-1946 (thin line)reconstructed based on regression between numbers shot and population size in 1958-1967 Density m 1946-1994 (thick line)- snowtracking censuses, numbers of lynx recorded in the Belarussian part combined with those m the Polish part corrected by20% ('transborder' lynx, see text) Broken hne - smoothed population dynamics Data in Appendices II and III

ECOORAPHY 19 2 (1996) 133

Karelian ASSR Pechora llych Reserve Yakutia (Russra)(Russia) w ^ 1

193E 1M0 1945 1950 1930 193E 1940 194S 1960 1965 1900 1965

NovgofcxJ IN W Russia) Central Forest Reserve (Russia)

I '"«-

Eurasian lynx Lynx lynx Blue hare Lepus timidus

Fig 6 Population dynamics of lynx (and in some regions of the blue hare Lepus timidus) in the Palaearctic region fromnorth-east Asia (Yakutia) to south-central Europe (Romania) Sources of data Yakutia- Tavrovskii et al (1971). Pechora-Ilychreserve and Komi ASSR - Teplov and Teplova (1947. cited after Heptner and Sludskii 1972) and Naumov (1947)Central-Forest Reserve - Zheltukhm (1986). Karelia. Estonia. NW Russia. Lithuania, Latvia Novikov (1967), BerezinskiReserve Geltman and Dolbik (1983), Romania - Kratochvil (1968)

Dynamics of lynx population in BialowiezaPrimeval Forest

During the three penods of fairly undisturbed function-ing of lynx population in the last 125 yr the levels oflynx density differed markedly We think that the num-bers of lynx were determined by availability of prey Inthe first half of the 19th century and until 1890. ungu-late community in BPF consisted of 4 species only (inorder of decreasing abundance) wild boar, roe deer.European bison, and moose The native red deer be-came extinct by the end of the 18th century (Brincken1828, Jarocki 1830) Numbers of wolves were veryhigh (up to 8 lnds 100 km~-) (J?drzejewska etal unpubl ) and brown bear (also preying on ungulates)still occurred in small numbers (Karcov 1903) Thus,the roe deer must have been the staple prey toboth lynx and wolves, although the latter hunted alsofor wild boar, moose and European bison (Karcov1903)

After 1920. ungulate community changed m favourof the lynx preferred prey Roe deer, red deer (the latterone reintroduced in 1890). and wild boar inhabitedBPF. whereas European bison and moose were tempo-rarily exterminated (Gavrin and Donaurov 1954.Wi?cko 1984. Pucek 1992) Wolves were so heavilypersecuted after 1946 (Gavrin and Donaurov 1954).that already by 1950 their numbers were much lower

than those of lynx That period was characterised bythe highest recorded numbers of lynx

In 1970-1994. densities of lynx were somewhat lowerthan during its best years The structure of ungulatecommunity has gradually changed Red deer becamethe dominant species, followed by roe deer and wildboar European bison was restituted and moose recov-ered, but the two latter species are still the least numer-ous ungulates in BPF (J?drzejewska et al 1994.Okarma et al 1995) Although up to 20 wolves are stillharvested annually in the Belarussian part, their num-bers are quite stable at 2-4 lnds 100 km -. l ecomparable to those of lynx (J^drzejewska et al un-publ )

In Bialowieza. the impact of man heavy enough todisturb the lynx-ungulate relationships occurred ratherlate in comparison to other European regions and itwas restncted to c 40 yr in the 19th century and some20 yr in the 20th century

Patterns of population dynamics of lynx in thePalaearctic region

Eurasian lynx shifts from the blue hare Lepu,s timtdus asbasic prey in the north and north-east to ungulatesat more southern and south-western latitudes of thePalaearctic region (J^drzejewski et al 1993) Long-term

134 ECOGRAPHY 19 2 (1996)

data from fur returns, hunting statistics and inventoriesshow, that there is a clear geographic gradient m thecyclicity of blue hare and, consequently, of theEurasian lynx populations (Fig 6) The 10-yr cyclesoccur m north-east Siberia (Yakutia) Towards west.the cycle length becomes shorter (5-6 yr m Komiterritories) In the northern part of Europe and atsouthern latitudes in the whole Palaearctic. fluctuationsof blue hare numbers are irregular and do not dnvefluctuations in lynx numbers (e g Central-Forest Re-serve), because lynx utihses ungulates as alternative and- still more south - as primary prey However, inregions, where Jynx depend primarily on ungulates (e gLatvia Estonia. Lithuania. Romania), its long-termdynamics can be quite strongly affected by humanpolicy of game management

Aeknowledgements - We sincerely thank U and C Breiten-tnoser (Bern, Switzerland) for their methodological instruc-tions and help with obtaining the equipment We are gratefulto A Kawecki for his unpublished data of lynx snowtrackingm Bialowieza National Park in 1957-1962. and to Polish andBelarussian game wardens who took part in censuses of lynxR Kowalczyk. S Sniezko. E Bujko, students of FarnboroughCollege of Technology (England). Earthwatch ResearchCorps, and students of Fachhochschule Hildesheim Holz-minden (Gottmgen. Germany) helped in radiotracking oflynxes K Zub prepared the figures. L Szymura and E Bajkohelped m analysing the snowtracking data L Nowacki drewthe silhouette of lynx We thank the priests of the Catholicchurch in Biatowieza. R Paszkiewicz. S Lukaszewicz. and JBuzun, who allowed our access to the church tower at anytimes of day and night This joint work was made possible bythe agreements of cooperation by the Bialowieza NationalPark, the Mammal Research Institute (Poland) and the StateNational Park Belovezhskaya Pushcha (Belarus) We aregrateful to C Okotow, Z Pucek and V Semakov for encour-agement and providing facilities during our exchange visitsComments by Z Pucek and O Liberg on the earlier versionare gratefully acknowledged The study was financed by grantsKBN 4 4416 91 02. KBN 6 P205 034 05. and partly byEarthwatch

ReferencesAnonymous 19'!7a Puszczajako teren reprezentacyjny [Prime-

val Forest as a representative terrain] - Echa lesne 16(14)366-374. in Polish

- 1937b Zarys historyczny Dyrekcji Lasow Panstwowych wBiatowiezy [Historical outline of the State Forest Direc-torate in Biatowieza] - Echa lesne 17(14) 379 406. inPolish

- 1938 Kronika towiecka [Hunting chronicle] - Echa lesne15(10) 204. m Polish

Brand, C J and Keith. L B 1979 Lynx demography duringa snowshoe hare decline in Alberta - J Wildl Manage43 827-849

Breitenmoser. U and Haller, H 1993 Patterns of predationby reintroduced European lynx in the Swiss Alps - JWildl Manage 57 135-144

Bnncken, J 1828 Memoire descnptif sur la foret lmperiale deBiatowieza en Lithuanie ~ N Glucksberg, Varsovie

Buchalczyk, T 1980 The brown bear in Poland - In Mar-tinka. C J and McArthur, K L (eds). Bears - theirbiology and management Bear Biol Assoc Conf Ser 3229-232

Chyrvonaya Kniga Respubliki Belarus 1993 Beiaruskaya En-cyklapedyya lm P Brouki - Minsk, in Belarussian

Elton. C and Nicholson, M 1942 The ten years cycle mnumbers of the lynx in Canada - J Amm Ecol 11215-244

Fahnski. J B 1986 Vegetation dynamics in temperate low-land primeval forests Ecological studies in Biatowiezaforest - Dr W Junk Publ, Dordrecht

Ferreras P , Aldama. J J , Beltran, J F and Delibes, M1992 Rates and causes of mortality in a fragmented popu-lation of Iberian lynx Fehs pardina Temminck. 1824 -Biol Conserv 61 197-202

Filonov. K P 1989 Ungulates and large predators in wildlifereserves - Nauka Moskva, in Russian

Gavrm. V F and Donaurov, S S 1954 Volk v BelovezhskoiPushche [Wolf in Belovezha Primeval Forest] - Zool Zh33 904-924, in Russian

Geltman. V S and Dolbik M S (eds) 1983 BerezinskiiBiosfernyi Zapovednik Belorusskoi SSR [Berezinsku Bio-sphere Reserve of the Belarussian SSR] Uradzhai.Minsk in Russian

Goszczynski. i 1986 Locomotor activity of terrestrial preda-tors and Its consequences Acta Theriol 31 79-95

Heisey. D M and Fuller, T K 1985 Evaluation of survivaland cause-specific mortality rates using telemetry data - JWildl Manage 49 668-674

Heptner V G and Sludskii A A 1972 Rys Fehs (Lvnx)hnx Linnaeus. 1758 - In Heptner. V G and Naumov,N F (eds) Mlekopitayushchie Sovetskogo Soyuza [Mam-mals of the Soviet Union] Izd Vysshaya Shkola. MoskvaVol 2 pp 385-457. in Russian

Herrendschmidt V . Leger F and Terrier G 1986 Moded occupation spatiale des lynx {Fehs lynx L ) lntroduits.dans le Massif Vosgien - premiers resultats - GibierFaune Sauvage 3 67-81

Jarocki. F P 1830 O Pu!)Czy Bialowiezkiey l o celnieyszych wniey zwierz?tach [About Biatowieza Primeval Forest and itsmore prominent animals] - Warszawa. in Polish

J?drzejewska B . Okarma. H . Jfdrzejewski, W andMifkowski, L 1994 Effects of exploitation and protectionon forest structure, ungulate density and wolf predation inBiatowieza Primeval Forest. Poland - } Appl Ecol 31664-676

J?drzejewski. W and Jfdrzejewska. B 1993 Predation onrodents in Bialowieza primeval forest. Poland - Ecogra-phy 16 47-64

- . J?drzejewska. B Okarma. H and Ruprecht, A L 1992Wolf predation and snow cover as mortality factors in theungulate community of the Biatowieza National ParkPoland - Oecologia 90 27 36

- . Schmidt, K Mitkowski. L . J?drzejewska. B andOkarma. H 1993 Foraging by lynx and its role in ungu-late mortality the local (Bialowieza Forest) and thePalaearctic viewpoints - Acta Theriol 38 385-403

Karcov G 1903 Belovezhskaya Pushcha Ee lstoricheskuocherk, sovremennoe okhotniche khozaistvo i vysochaisheokhoty V Pushche [Belovezha Primeval Forest Historicaldescription, contemporary game management and monar-chial hunts in the Forest] - A Marks S Petersburg inRussian

Kossak, S 1988 Winter penetration and movements of a malelynx Acta Theriol 33 354-360

Kratochvil, J 1968 The lynx population in Rumania - Actasc nat Brno 2(5-6) 65-70

Kvam, T 1991 Reproduction in the European lynx. Lynxlynx - Z Saugetierkunde 56 146-158

Naumov. S P 1947 Ekologiya zaica-belyaka [Ecology of theblue hare] - Izd Moskovskogo Obshch lspyt Prirody,Moskva, in Russian

Novikov. G A 1967 An experience of quantitative estimationof the present condition of lynx populations in the westernpart of the USSR - Byull Mosk Obshch lspyt Prir(Otd Biol) 72(3) 5 18. in Russian with English summary

ECOGRAPHV 192(1996) 135

Okarma. H . J^rzejewska. B , J?drzejewski. W . Krasinski. ZA and MiJkowski. L 1995 The roles of predation. snowcover, acorn crop, and man-related factors on ungulatemortality m Biatowieza Primeval Forest, Poland - ActaThenol 40 197-217

Pnklonsky, S G 1965 Coefficients to treat the data of wintertranssect method of census taking of game animals by theirtraces - ByuII Mosk Obshch Isp Pnr (Otd Biol) 70(6)5-12. in Russian with English summary

Pucek. Z 1992 Conservation strategy for European bison -Proc intern Symp Ongules/Ungulates 91 (Touluse-France, Sept 2 6, 1991) 589-594

Qumn N W S and Parker, G 1987 Lynx - In Novak, M .Baker. J A Obbard. M E and Malloch. B (eds). Wildfurbearer management and conservation in North Amer-ica Mm Nat Resources, Ontario, pp 683-694

Seal, U S and Kreeger, T S 1987 Chemical immobilizationof furbearers - In Novak. M Baker, J A . Obbard, ME and Malloch, B (eds). Wild furbearer management and

conservation in North America Mm Nat Resources, On-tano, pp 191-215

Tavrovskii, V A , Egorov, O V , Knvosheev V G , Popov,M V and Labutm, Y V 1971 Mlekopitayushchie Yakutii[Mammals of Yakutia] - Izd Nauka, Moskva, in Russian

Teplov, V P and Teplova, E N 1947 MlekopitayushchiePechoro-Ilychskogo Zapovednika [Mammals of the Pe-chora-llych Reserve] - Trudy Pechoro-Ilychskogo Za-povednika. 5, in Russian (cited after Heptner and Sludskii1972)

Voit, E 1917 Die Erischliessung des Urwaldes - Bialowies mDeutscher Verwaltung Heft I Paul Parey, Berlin, pp 9-54

Wi?cko, E 1984 Puszcza Biafowieska [Bialowieza PrimevalForest] - PWN, Polish Scientific Publishers, Warszawa, inPolish

Zheltukhin. A S 1986 Biocenotic relationships of the Eu-ropean lynx (Lvttx hnx) in the southem taiga of the UpperVolga - Zool Zh 65 259-271. m Russian with Englishsummary

Appendix I

Lynxes captured and radiotracked in Bialowieza Primeval Forest in 1991-1994 If not stated otherwise, theradiotracking was continued after 30 Apnl 1994

Name age, sex

Bazyli, ad, MMaicary, ad M

Mala, JUV, FBazyliszek juv M

subad. MBorys. ad MNikifor. ad. M

Sonia. subad, FZoja, JUV, F

subad, FNatasza, ad, FOlga, ad, FIgor, subad, MTamara, subad, F

Bodymass(kg)

2019

23'9

11

14-2421

15109

165151615

Dates of radiotracking

24 Feb-23 Dec '919 March-3 July '9127 Feb '93-30 Apr '9411 Dec '91-21 Jan '927 Feb '92-21 March '92

28 March '92-21 Nov '938 Feb-7 Dec '9225 Jan-23 Feb '93

30 Jan '93-30 Apr '945 Jan '93-Apr '93Apr '93-30 Apr '943 Feb '93-28 March '9416 Feb-30 Apr '9422 March-30 Apr '9429 March-30 Apr '94

Cause of untimely end ofradiotracking and othernotes

poachedtransmitter failure

disappearedstayed with mother andanother kittenpoachedpoachedkilled as rabied in KnyszynForest (31 Dec '93), 60km N from capture place

stayed with mother

disappeared

^captured on 15 December 1993 for collar change, body mass 20 kg,-captured for collar change in June 1992 (weighted at the time of capture)

136 ECOGRAPHY 19 2(1996)

Appendix II

Numbers of lynx killed in BPF (1250 km-) in 1869-1938, numbers of lynx estimated by game wardens (1928,1935), and reconstructed population numbers (the latter in italics) Sources of data in Methods and materialReconstruction of numbers was done for years of intense persecution of lynx and was based on regression betweenthe number of lynx killed and estimated population size in 1958-1961, 1966-1967 (l e years with reliable data forboth Belarussian and Pohsh part of BPF. data in Appendix III) Y = - 0 066 + 4 783X, R - = 0 87, p = 0007,n = 6 "* - reliable reconstruction of lynx numbers not possible or available data uncertain

Year

18691870187118721873187418751876187718781879188018811882188918901891

Yearlynumber ofkilled lynx

-)5057541511100262

Reconstructednumber of

lynx

10240

243324195

24555

102910

Year

18921893189418951896189718981899190019011902192819341935193619371938

Yearlynumber ofkilled lynx

00030->012100

348

1116

Reconstructednumber of

lynx

000

140

1005

1050

300

95''9

9

76

Appendix III

Number and harvest of lynx m the Polish part (exploited forests 530 km*, and Bialowieza National Park 47 5and the Belarussian part of BPF (670 km') in 1946-1994 Numbers of lynx censused numbers are in Roman type,reconstructed numbers in italics (reconstruction of numbers in 1951-1957 in the Polish part and in 1983, 1985,1990 m the Belarussian part based on regression of lynx numbers in the Belarussian and Polish parts (Y = 3 36 +0 414X, R- = 0 352, n = 38, p < 0 0005) Question marks denote numbers of lynx censused but strongly departingfrom the long-term trend (and thus thought to be unreliable), for such years numbers estimated from the aboveregression equation are given in parentheses Sources of data and methods of censuses described in Methods andmaterial

Year

19461947194819491950195119521953195419551956195719581959196019611962

Numbersin the

exploitedforests

3230303032202520331823212012632

Lynx in the Polish part of BPF

N killed Numbersm BNP

000000000001 060 1 12 1000 010 02

Lynx in the Belarussian

Numbers

29233628533135314229

W (34)32344230105

part of BPF

N killed

12656

123123

1278

10949

ECOORAPHY 19 2 (19%) 137

Appendix III (cont.)

Year Lynx m the Polish part of BPF Lynx in the Belarussian part of BPF

Numbersin the

exploitedforests

N killed Numbersm BNP

Numbers N killed

19631964196519661967196819691970197119721973197419751976197719781979198019811982198319841985198619871988198919901991199219931994

5401039101422

15167192014291224131418911161212II131216

1 61 6I 4I 1

0108

472

92030493842

62'' (27)3326172041332440253529413726182417211728

Appendix IV

Cases of non-harvest mortality of lynx from the Polish part of BPF r - river, v - village, names of forest admin-

istration units (in parentheses) as in Fig 1 Names of lynxes refer to radiocollared individuals (see Appendix I)

Date Place Description of lynx Cause of death

Winter 1978

1986

April 1988

April May 1990

Winter 1990/91

Winter 1990 91

December 1991December 1992

December 1993

December 1993

April 1994

E of Biafowieza v(Biafowieza)Sacharewo v(Lesna)W of Krzywiec v(Ladzka Puszcza)Przewloka(Bialowieza)(Browsk)

Nowosady v(Narewka)Hwozna r (BNP)W of Nowosady v(Narewka)Skryplewo v(Narewka)

Knyszyn Forest

N of Pogorzelce v(Zwierzyniec)

Adult male

Adult male

Adult male BazyliAdult male Borys

2 5-yr old maleBazyliszek

Adult male Nikifor

Adult male

Poached with twodogs (Kossak 1988)Shot in a farmyard

Poached, probablypoisonedFound decayed, causeof death unknownFound dead fromgun-woundsPoached

Poached with snarePoached with snare

Poached, its collar(destroyed) found on4 January 1994Killed in a farmyard.rabiedFound dead, decayed.cause unknown

138 ECOGRAPHY 19 2 (1996)