CSIC,Estaciónbiolbgicadedoflana,Avda.Marialuisas/N,41013Sevilla,Spain

CurrentrangeandstatusoftheIberianlynx

FelispardinaTemminck,1824inSpain

AlejandroRodriguezMiguelDelibes

CSIC,EstaciónBiolbgicadeDoflana,Avda.MariaLuisas/n,41013Sevilla,Spain

Thecurrent distribution oftheIberian lynxFe/ispardinaisoutlinedanditspop­ ulation sizeinSpain estimated at about 95%ofitsworld range. There are48 separate breeding areas, generally small in size.In addition, 32non-breeding areashavebeenlocated,and50areaswherelynxpresenceisuncertain, someas­ sociated with,and others separated from, breedingareas. Amajor part ofthe rangesupports lowlynxdensities. Rangefragmentationisthemostnoteworthy feature ofthedistribution pattern, although dispersal mightlinksomeadjacent nuclei.Ninegeneticallyisolated populationsare recognized,although probably only two, accounting for about 70%of the total population,are viableinthe short term.TheSpanish population sizeisestimated atabout 1100individuals, withfewerthan 350reproductive females.The risksarising from thissituation areanalysedandconservation policiesproposed.

INTRODUCTION

The pardel or Iberian lynx Felispardina,Tern­ minck1824hasbeenconsideredby severalauth­ orities asasubspecies oftheEurasian lynx, Felis lynxL.(EllermanMorrison-Scott,1951;Corbet,

1978;CorbetHill,1986).However,coexistenceof bothformswithout hybridization hasbeenproved fromthePleistocenetorecenttimes(Kurtén, 1968; vandenBrink,1970,1971).Inaddition, studiesof the skull morphology and dentition reveal that theIberianlynxis theoldestrecentlynxspecies, phylogeneticallymoreseparated from F.lynxthan thisspeciesisfrom F.canadensis and F.rufus (Werdelin, 1981).The specificstatus ofFelispar­ dmahaslatelybeenupheld, including areviewof

(‘arnivora systematics (Wozencraft, 1989, sup­

ported byHonacki etal.,1982;Werdelin,1990).

The Iberian lynx has been recognized as the most threatened European carnivore (Mallinson,

1978).It isincluded in Appendix II of the Bern

(‘onventionontheConservationofEuropeanWild­ lifeand Habitats,and recently has been moved fromAppendix IItoAppendix IofCITES (Jack­

son, 1990).In the same way, it iscatalogued as

‘Endangered’ in the IUCN Red List of Threat­ enedAnimals(IUCN, 1988)and theSpanish Red Data Book(ICONA, 1986).However,noupdated field surveyexistsonitsstatusanddistribution. Iberian lynx distribution maps have been based oncompilationsofunsystematic reports (seeVal­ verde,1963,Garzón, 1978,and Delibes,1979,for Spain, and Palma, 1980,for Portugal). Allidenti­ fiedafewisolatedpopulationsinthesouthwestern mountains,andsometimesalsosomescatteredlocal­ itiesin the north and east of the Iberian pensin­ sula.Nevertheless,remarkable differencesoccurin themapsandpopulationestimates.

Theaimsofthispaper aretodescribeprecisely thecurrentdistribution andnumbersoftheIberian lynx,and todiscusstheconservation risksarising fromthepatternofdistributionandpopulationsize.

METHOD

BetweenNovember 1987and May 1988 4000 questionnairesrequestinginformationonthepres­ enceoflynxwere senttorangers,gameboards, taxidermists,andnaturalists’associations inpoten­ tiallyfavourable areasforthelynx(i.e.thesouth-

189

em two-thirds ofpeninsularSpain).Thirty-fiveper cent of the enquiries were answered. Of these,

245%(255) contained somepositiveinformation. Simultaneously, advertisements were insertedin hunting and nature journals requesting reports, and thescattered published and unpublished data onlynxdistributioncollected.

FromNovember1987toOctober1988fieldwork

was focused on: (1) checking reports from the postalenquiriesby interviewinglynx observersor hunters;(2)obtaining newreports fromthosewho couldnotbereachedbypost,suchasprivategame- keepers,shepherds andtrappers, and(3)makinga roughevaluation ofhabitat features,suchasvege­ tation cover, preyabundance and human activity (landuse)thathavepreviouslybeenrelatedtolynx presence(Delibes,1979;Palomares etal.,1991).

The field study area covered 74000 km2and about 2500peoplewere contacted (average34 interviews/l00 km2). Onalargescale,sampling intensity wassimilar over thewholeof thestudy area.

Reports were carefullyreviewed,onlybeing accepted as positive (1) if they were supported

by evidencesuchasfur,skeletalremains,tracks, faeces,and/or photographs,verifiedbyourselves; or (2) in the absence of definite proof, when severalaccuratelynx descriptionsfromseveral experienced observers wereinagreement, locating thespeciesinthesameareaandatthesametime.

Combining bothmethods—mail dataandfield checks—reduces thedisadvantages ofusingeach alone(Filion,1980),andprovidesahighefficiency/ effortratio.

Atotal of1583positive reports for the period

1978—1988were accepted(asimilarnumberwere rejected). A current distribution map was pre­ paredfromthesereports,assumingno significant changes in range boundaries during the present decade.(Thismaybeanunrealisticassumption so that ourresultsmaybeover-optimistic.)

Inthismapwewereabletodelimittwotypesof

areas:

(1) breedingareas,wheretherewasacontinuous spatio-temporal pattern in the distribution of reports,anddataonbreedingoccurred (recordsofkittensorpregnant females).

(2)areas of occasional lynx presence, where there were few reports, separated in time (>5years) and/orthroughspace(>10km betweencontiguouslocations),andtherewere norecordsofbreeding.

Furthermore,weregarded asareasofuncertain presencethosewherereports (usually <10)tended to aggregate within afewsquare kilometres. Be­ causealloftheseisolated recordswerelocatedfar from those where lynx presence was confirmed and noconsistent evidencewasfound, wecannot eitherconfirmorrejectlynxpresencethere.

Assumingsomedirectrelationship betweenlynx

density and frequency of reports, the proportion ofbreedingdata andhabitat quality (according to thestandard lynxhabitat previouslydescribed; Delibes,1979;Palomares etal.,1991),we distin­ guishedsubjectivelythreeclassesofbreedingareas: (1) high-density (many records, more than 10% relatedtobreeding,with good-qualityhabitat);(2) intermediate(breedingdataaccountingforlessthan

10%);(3)lowdensity(breeding rarelyrecorded).

Toestimate numbers weusedasareferencethe number oflynxlivingin50km2 oftheDoñana region.Herelynxdensitywasestimated bothfrom signcounts(Rauetal.,1985) andradio-tracking surveys(Beltrán,1988; unpublished data),as16 individuals/100 km2,excluding kittens (Palomares eta!., 1991).Inthelightofourexperienceinthe Doñana area,where therelationshipbetweenlynx reports and lynx abundance is known, we were abletoassigndiscretevalues ofrelativedensityto each of the remaining areas, taking into account the quantity andqualityofthereportsonlynx presenceobtained there. Asarule,breedingareas withhighlynxdensitywere estimated tosupport between 13 and 21 individuals/100 km2, inter­ mediateareasbetween8and 12,andthoseoflow density between4and7.Weassignedtotheareas ofoccasionalpresencedensitiesbetween2 and3 individuals/i00km2.

RESULTS

Figure 1shows thecurrent breeding range ofthe Iberian lynx. Although not an objective of our fieldstudy,data fromPortugal (Palma,1980)were includedtorepresentthe entireprobableworld breedingareaforF.pardina.

Thepardellynxoccursonallthefoursouthern mountain chains ofIberia, and alsoinaflatarea at sealevel,near themouth oftheGuadalquivir. A simpleanalysisshowsthatthelynxinhabits relativelyundisturbed Mediterranean habitatsfar from the main human settlements. Except in Doñana (at sea level),it occupies areas between

400and 1300maltitude, wherethere isMediter

Fig.1. Current breeding range (•)of theIberian lynx.Portuguese data from Palma (1980)and others unpublished. mountains >1000m.Principal Iberian riversarealsoshown.

raneanforestorshrubland.Highvegetationcover, an abundanceof rabbits (which prefer a mixture of coverandopengrassland)andlowhumanin­ fluenceseemtobe importantfeaturesofthelynx habitat. Inmountains, rockyplaces areusually selectedforbreeding,usuallycoincidingwithhunt­ ingreservesandnatural parks.

The lynxbreeding range occupies about 11000

km2,containing over48breeding areas ofgreatly varyingsize.Around the breedingareas, 32areas ofoccasionalpresencehavebeenestablished (Fig.2), whose total extent isestimated to beabout 3900 km2, although the limits are difficult to draw. Since thethreeisolatedlynxnucleireportedin Portugal hardly account for 700 km2 (Palma,

1980),weestimate that atleast95%ofthepardel

lynx range occurs inSpain, the area covered by ourstudy.

Taking intoaccount theabilityofthespeciesto

disperse(studiesinDoñanabyBeltrán, 1988,and one of the authors, unpublished data, recorded young dispersing more than 30 km), it seems realistictoconsider the48isolated breedingareas and32areasofoccasional presenceasninegeneti­ cally isolated populations, corresponding to the

Western SierraMorena(SMH),Doñana(DOIcT), Central Sierra Morena (SMC), Subbetic Moun­ tains (SUB),CentralSpain(CP,includingEastern SierraMorena,SMO,andMontesde Toledo­ Villuercas, MTO-VIL), Sierra de Gata (GAT), Sierra de Gredos (GRE), Alto Alberche (MAD) andSierradeSanPedro(SSP)(Fig.2).

In addition, it is possible that lynx occur in

50areas where their presence is uncertain, scat­ teredmainlyovereasternandnortheasternregions (Fig.3).

Theextentof thebreedingareas,estimatesof density,and populationsizeareshowninTable1. Twelve high-densityareas(>13individuals!100km2, mean size150km2 range 25—500 km2)account for 17%of the total breeding range. The maxi­ mumestimatedabsolutedensityreaches21 mdi­ viduals/100km2ina500-km2zoneinarea26.

Nineteenmedium-densityareas(8—13individuals!

100km2, meansize200km2 range10—840km2) accountfor36% ofthetotalbreedingarea.The remaining47%concerns 31low-densityareas(5—8 individuals/100 km2),withsimilarsizevariation to thatofthemediumareas,buttendingtobesmaller (= 161km2).

40km

Fig.2. Lynxdistribution inSpain.Continuouslinessurroundthenineestimated populations,includingbreeding(•)andocca­

sional presence(ii)areas. Threesubpopulations(VIL, MTO,SMO) arealsoindicated bybroken lines.Straight linesrepresent minimum barrier breadth (km). Symbols indicate thedegreeof barrier penetrability for lynx:*,high;•,low;0,null.GAT,

SierradeGata; GRE, SierradeGredos; MAD, AltoAlberche;SSP,SierradeSanPedro;CP.Central population (VIL,Villuercas; MTO, Montes deToledo; SMO, Eastern Sierra Morena); SMH, WesternSierra Morena; SMC, Central Sierra Morena; DO1J, Donana;SUB,SierrasSubbéticas.

As expected,asignificantlypositivecorrelation existsbetweenestimated population sizeand area size(r=0985, p00001), but notbetweenlynx relativedensityandareasize.

A decreasing density is noted from south to north onapeninsular scale,andfromeasttowest withineachmountain range.Thehighestrelative densitiesarefound inDoñana,eastern Sierra Morena, andeastern MontescleToledo.

The estimated total population size is about

1100individuals, halfoccurringintheSMOpopu­

lation. IntheDoñana areaweestimatetheretobe

49lynx (Table 1),whilePalomares et a!.(1991)

established apopulation sizeforthesamearea of

40—50individuals. Accepting theselimitsof vari­ ationforalltherange,thetotalSpanishlynxpopu­ lation wouldcomprise880—1150individuals.

Theage-ratioisnotknownforanyoftheIberian lynxpopulations. Theproportionofyounginnon- exploited populationsofFe/isrufusamounts to16

(Lembeck Gould, 1979)to 55%(Toweill,1980, in Anderson, 1987). Because a male may mate withseveral females,productivity islimited bythe number ofadult females(asinthebobcat, Knick,

1990).Assuming the average proportionof juve­ nilestobe35%and thesex-ratiocloseto1:1,the numberofbreedingfemalelynxinSpainprobably doesnotsurpass 350individuals.

DISCUSSION

The current distributionof lynx,as estimated in ourstudy, roughlyagreeswithmapsofferedsome timeagobyotherauthors (Valverde,1963;GarzOn,

1978; Delibes, 1979), as most of the occupied areas are in the mountains of the southwestern quarter of the Iberian peninsula. However, none ofthepreviousmapsattainstheaccuracyof that presentedhere,andneverbeforehasanestimation

Fig.3. Areas withuncertain lynxpresence(•).Sourcesfor PyreneesareJordan etal.(1988)and J.Ruiz-Olmo (pers.comm.) and for northwest region, Clevenger(1987)and Grande Hernando(1982).Numbers correspond to thelocation ofbreeding areas(seeTable1).

ofdifferentialdensitiesofeachlynxareabeengiven.

Because both lynx numbers and range have declinedsince1950 (Delibes,1979;Rodriguez Delibes, 1990), new localities indicated by our work cannot be attributed to recent recoloniza­ tion,buttobiasedorabsentinformation inearlier studies.

Small northern and northeastern populations

suggested by Valverde, and then questioned by GarzónandDelibes, remainunconfirmed,despite occasional reports from these regions (Grande,

1978;Grande Hernando,1982;Clevenger,1987; Jordan et al.,1988;J. Ruiz-Olmo, pers.comm.). Ourdata areinsufficienttoprovelynxpresencein these areas. Possible lynx presence has also re­ cently been reported in the French Pyrenees (Chazel, 1989), although the last clear evidence found there wasaskullofFelislynx dating from about 1950(Beaufort, 1965).

Estimatedpopulationsizespublishedinthe1970s werealso based on fragmentarydata, but were fairlycorrect, ifslightlyoptimistic. Thus, Urquijo (1975) and Garzón (1978) calculated 1200—1600 and 1000—1400adult lynx, respectively. IUCN

(1978), supported by Garzón’s data, accepted

1000—1500 individuals,includingthePortuguese population.On the other hand, theSpanish Red Data Booklaterreportedabout 400lynx(ICONA,

1986).

Although thesubjectdeservescloserstudy,three phases,characterized by differentmaincausal reasons,canprobably berecognizedinthehistory ofthedeclining range ofthe Iberian lynx.In the firstperiod, from antiquity until the1950s,hunt­ ing was probably the most important cause of lynxdisappearance,ashappened withFelislynxin Central Europe (Eiberle, 1972).In this phase the lynx disappeared fromgoodpotentialareasnow separated fromoccupiednucleibyavastextentof unsuitablehabitat. Natural recolonization is thus scarcely possible. In the second period, sincethe

1960s, habitat destruction and scarcity of rabbits (themainlynxprey;Delibes,1980)due tomyxo­ matosiswereprobably themoreimportantreasons for lynx decline (Garzón, 1978; Delibes, 1979). Inthisphase,coincidingwith rapideconomic development in Spain, many dams, highways and railwayswerebuilt, and most Mediterranean

Table1. Areasize(AS), estimatedaverageabsolutedensityTable1—contd.

(AD), andestimated populationsize(PS) inthe48 breeding

areasandtheninelynxpopulationsinSpainBreedingarea”ASADPS

(km2)(individuals!(nokittens) Breedingarea” AS AD PS 100km2)

(km2)(individuals!(nokittens)

100km2)CP(=SMO+911288808

MTO +VIL)

I Andévalo2356415

2 Aroche26644

populations

3Cumbres274841 Granadilla

4 Rosal174042Santa Cruz

5 Encinasola5196543 Gata

6 Aracena8264544Cilleros

7 Zufre113647

Occasional presence

Occasional presence4852412GAT population

SMH population1036455345Candeleda

8 Moguer131486GRE population

9 Donana1681602746 AltoAlberche

10CotodelRey521608

II Aznalcázar38401MAD population

12Torrecuadros1540147Cedillo

Occasional presence13540648 SanPedro

DO1population5399149

Occasional presence

SSPpopulation

Spanish Lynx population

“SeeFig.1.

scrublands wereconverted toagricultureand plan­ tationsofpines andeucalyptus,reducingthelynx rangeandcausingthefragmentation ofitspopula­ tions.Atpresent,lackofprey(inadditiontomyxo­ matosis, a new viral disease is destroying wild rabbitsin Spain;VillafuerteMoreno, 1991), humanizationoftheenvironment—mostlynxdie onroads,drowned inirrigation wellsandcaptured in traps for rabbits and foxes (Ferreras et al.,

1992,thisissue)—and theproblems derived from

fragmentationofpopulationsseemtobethemain threats to lynxsurvival, sincedirect hunting has beenreduced(Rodriguez Delibes,1990).

Fromthepointofviewofconservationbiology, smallsize,isolationanddispersion oftheoccupied areas are the more noticeable features ofcurrent lynxdistribution. Thispermitsustoaddressthe problem asatypical island biogeographic model, inwhichisland(fragment) sizes,estimated density anddistancebetween islandsareknown,butthe possibility of lynx crossing the different inter- islandbarriers isatpresentdifficulttoassess.

Considering only the barriers betweenthenine

isolatedpopulations ofFig.2,itappearsthat intensiveagricultureinthe TajoandGuadalquivir valleysseparates thethree northern (GAT, GRE, and MAD) and onesouthern (SUB)populations,

respectively,fromacentralpopulation(CP).Al­ tered,man-madelandscapeisolates DOI1from SMH,andthesamefactor,alongwithdistance, isolatesSSPfromCP.Suboptimalhabitatsmay explainthe separation betweenSMC and CP, whereasthisandlowlynxdensityinthebreeding areasisolate theSMCandSMHpopulations. Distanceand lowdensityseemto bethemain causesoftheGAT—GRE—MADbarriers.Further­ more,asmentionedabove,lowlynxdensity,prob­ ablyrelatedtoscarcityofprey,tendstodividethe centralpopulationintothreesubpopulations:VIL, MTOandSMO.Inthenearfuture,several new obstacles,includinghighways,largedams, and fenced-offhigh-speedrailways,mightrestrictlynx movementandcontributetotheprocessoffrag­ mentation.

Only two subpopulations of CP (SMO and

MTO,Fig.2)arefoundinareasofmorethan

2000km2,andonlythesetwohave200ormore

individuals.Theremainingpopulationsneverreach

100individuals,livinginareasoflessthan1000km2.

Becauseoftheirsmallsize,allpopulationsexcept

perhapsCPareprobablynotviableinthelong term,duetoincreasedvulnerabilitytostochastic

events(Shaffer,1981).Ifisolationreallyoccurs betweenall48 breedingnuclei,96%could not reachaneffectivepopulationsize(Ne) of50, the theoreticalthresholdbelowwhichlossofgenetic variabilityrapidlyincreases(Franklin,1980).

Habitatdestructionisoftenresponsibleformost recent‘normal’extinctions(WilcoxMurphy,

1985;Diamond,1989) andthisseemsto bethe

maincauseofthedeclineofthelynxpopulation. Ourresultsconfirmtheendangeredstatusofthisrare

carnivore,suggestingthatfragmentationandhuman disturbancearethe mainthreatsthatcouldleadto theextinctionofthe remainingpopulations.Infact, several reduced lynxpopulationshavebecome extinctinthe20thcentury,asaresultoffragmen­ tationbyruraldevelopment(seealsoFerrerasetat.,

1992,thisissue).

Becauseofthepracticalproblemsofestablish­ ingreserveslargeenoughtoguaranteethesurvival oftheincreasingnumberofspeciesthat become threatened,achangefromaspeciesapproachto anecosystemlland-useapproachissuggestedfor conservation strategies(e.g.Grumbine,1990).A conservationplanforthelynxmustbeincludedin abroadMediterranean ecosystemconservation programmewhichharmonizesthedevelopmentof ruralareaswiththemaintenanceofhabitatsand traditionallanduses.

Inaddition,short-termactions,inouropinion, shouldbefocusedonthreegoals:

(1)preventnewmarginalhabitatlossesorfrag­

mentationbysettingupprotectedareas;

(2)guaranteesurvivalofthe viablepopulations, specificallythe CP population, whichin­ cludesabout70%of thetotallynxnumbers;

(3)managepopulations,byincreasingnumbers inlow-densityareas,linkingadjacentnuclei bynatural (corridors)or artificialcontact (lynx release,insemination),and through reintroduction(southernSpain)andcaptive breeding.

Moreresearch intothefactorsthatgenerate barriersandtheassociatedlynxresponseisneeded to understandthefragmentationprocessand to predictlong-termpopulationtrends.

ACKNOWLEDGEMENTS

TheInstitutoNacionalparaIaConservacióndeIa Naturaleza(ICONA)andtheConsejoSuperiorde InvestigacionesCientIficas(CSIC)providedfinan­ cialsupport.WethankJ.N.GuzmánandJ.C. delOlmoforfieldassistance,and F. Gonzalez Bernáldez,P.Jackson,A.LazoandP.Starrsfor commentsondifferentdraftsofthemanuscript.

REFERENCES

Anderson, E.M.(1987).Acriticalreview andannotated bibliography ofliterature on bobcat. Colorado Div.Wild!. Spec.Rep.,62,1—61.

Beaufort, F.(1965).LynxdesPyrénées, Felis(L.)lynx lynx

(L.).Mammalia,29,598—601.

Beltrãn, J. F. (1988). Ecologia yconducta espacio-temporal

dellinceibérico(LynxpardinaT.)enelParque Nacional deDonana.PhDthesis,Universidad deSevilla.

Chazel, L.(1989). Notes sur Iasurvivance du lynxdans les

Pyrénéesfrancaises. Mammalia,53,461—4.

Clevenger. A. P. (1987). Observación de an lince ibérico

(Lynx pardina)enlaprovincia deLugo, norte de Espana.

Doñana,ActaVert.,14,140—2.

Corbet, G.B.(1978).TheMammalsofthePalaearcticRegion: ATaxonomicReview. BritishMuseumNatural History& CornellUniversityPress,London.

Corbet, G. B.Hill,J. E.(1986). A World List of Mam­ malianSpecies.BritishMuseumNatural History, London. Delibes,M.(1979).LelynxdansIaPéninsuleIbérique:répar­

tition et regression. BulL mens. Off National Chasse, No. spec.Sci.Tech.,Lelynx,41—57.

Delibes, M. (1980).Feeding ecologyof theSpanish lynx in

theCoto Donana.ActaTheriol.,25,309—24.

Diamond, J.(1989).Overview ofrecentextinctions.InCon­ servationfortheTwenty-firstCentury,ed.D.WesternM. C.Pearl.Oxford UniversityPress,NewYork, pp.37—41.

Eiberle, K.(1972).Lebensweiseund Bedeutung der Luchses in der Kulturlandschaft.Mammalia Depicta, No. 8. Paul Parey,Hamburg, Berlin.

Ellerman, J.R.Morrison-Scott,T.C.S.(1951).Checklist of Palaearctic and Indian Mammals 1758—1946.British MuseumNatural History, London.

Ferreras, P., Aldama, J. J., Beltrán, J. F. Delibes, M.

(1992). Ratesandcausesofmortalityin afragmented populationofIberian lynxFelispardina(Temminck). Biol. Conserv.,61,197—202.

Filion,F.L.(1980).Humansurveysinwildlifemanagement.In Wild4feManagementTechniquesManual,ed.S.D.Schemnitz. TheWildlifeSociety,Washington DC,pp.441—53.

Franklin, I. F. (1980). Evolutionarychange in small popu­ lations.InConservationBiology:AnEvolutionary—ecological Perspective, ed.M.E.SouléB.A.Wilcox.Sinauer, Sunderland,Massachusetts,pp.135 49.

Garzón,J.(1978).DieSituation des Luchses inSpanien. In Der Luchs in Europa—Verbreitung. Wiedereinburgerund, Rauber—Beute—Beziehung, ed.A.Festetics.Kilda-Verlag, Greven, pp.161—9.

Grande,R.(1978).Losiiltimoslincesdelassierrascentrales.

VidaSilvestre,28,229—35.

Grande, R.Hernando,A.(1982).Localizadas dos nuevas areasdelince.Quercus,3,20—1.

Grumbine,R.E.(1990).Viablepopulations,reservesize,and

federal lands management: a critique. Conserv. Biol., 4,

127—34.

Honacki, J. H., Kinman, K. E. Koeppl, J. W. (1982).

Mammal SpeciesoftheWorld.ATaxonomicandgeographic reference. AllenPressTheAssociationofSystematic Collections, Lawrence, Massachusetts.

ICONA (1986). Lista Roja de los Vertebrados de España.

ICONA, Madrid.

IUCN (1978).RedDataBook.IUCN, Gland.

IUCN (1988). Red List of Threatened Animals. IUCN, Gland.

Jackson, P.(1990).CITES 1989.Catnews, 12,23—4.

Jordan, 0.,Ruiz-Olmo, J.Orta, J.(1988).Estudiosobreel

linceyotros carnivoros forestales enelPirineo aragonés. Basesparasuprotección ygestión.DGA-Departamentode Agricultura, GanaderiayMontes, Zaragoza (unpublished

report).

Knick, S.T. (1990).Ecology of bobcats relative toexploit­

ation and a prey decline in southeastern Idaho. Wildl. Monogr.,108,1—42.

Kurtén, B.(1968).PleistoceneMammals ofEurope.Weiden­

feldNicolson, London.

Lembeck, M.Gould, G. I.(1979).Dynamics ofharvested and unharvested bobcat population in California. Proc. BobcatRes.Conf,Nain.Wildl.Fed.Sci.Tech.Ser.,6,53—4. Mallinson, J. (1978). The Shadowof Extinction. Macmillan,

London.

Palma, L. A. (1980). Sobre distribucao,ecologia yconser­ vacao do lince iberico em Portugal. Actas I Reunion Iberoamer.ZooL Vert., 569—86.Sevilla,Ministerio deUni­ versidadeseInvestigación.

Palomares, F., Rodriguez, A., Laffitte, R. Delibes, M. (1991). The status and distribution of the Iberian lynx, Felispardina(Temminck)inCoto Donana area,SWSpain. Biol.Conserv.,57,159—69.

Rau, J. R., Beltrán, J. F. Delibes, M. (1985). Can the increaseoffoxdensityexplainthedecreaseinlynxnumbers atDonana? Rev.Ecol.(Terre Vie),40,145—50.

Rodriguez, A.Delibes,M.(1990).ElLinceIbérico Lynx pardina en Espana. DistribuciOny Problemasde Conser­ vación.ICONA, Madrid.

Shaffer,M.L.(1981).Minimumpopulationsizes forspecies conservation.BioScience,31,131—4.

Urquijo, A. (1975). Protection et apparition du lynx en

Espagne.ConseilInternationaldeIaChasse,XIII Assemblée

GénéraleTriennale.Paris-Chambord.

Valverde, J. A. (1963). InformaciOnsobreel Lince Español.

ServicioNacionaldePescaFluvialyCaza, Madrid.

Vanden Brink, F. H.(1970).Distributionand speciation of someCarnivores. Mammal Rev.,1,67—78.

Van den Brink, F. H. (1971). Le lynx pardelle en France.

Bull.Soc.d’EtudeSci.Nat.Nimes,51,109—17.

Villafuerte, R. Moreno, 5. (1991). Rabbit haemorragic

disease(RHD)inDoñana National Park (DNP). Abstracts

Congr.mt.Un.GameBiol.,20th,Gödöllo, Hungary,August

1991,pp.107—8.

Werdelin, L. (1981). The evolution of lynxes. Ann. ZooL Fennici,18,37—71.

Werdelin,L.(1990).Taxonomic statusofthepardellynx.Cat news,13,18.

Wilcox,B.A.Murphy, D.D.(1985).Conservation strategy:

theeffectsoffragmentationonextinction. Amer.Nat.,125,

879—87.

Wozencraft, N. C. (1989). Classification of the recent

Carnivora.In CarnivoreBehavior,Ecology,and Evolution,

ed.J.L.Gittleman. Chapman Hall,London, pp.569—93.