ImpactofMarketHuntingonMammalSpeciesin

EquatorialGuinea

JOHNB. FA,*JAVIERJUSTE,tJAIMEPEREZDELVAL,$ANDJAVIERCASTROVIEJOt

jerseyWildlifePreservationTrust,LesAugrèsManor,Trinity,JerseyJE35BF,ChannelIslands,email fEstaciónBiológicadeDoflana,C.S.I.C.,PabeliondelPeru,AparladoPostal1056,41080Sevilla,Spain

*CooperaciónEspafiola,Malabo,GuineaEcuatorial

Absfract TheImpactofcommercial bunting onforestmammalswasstudied intworegionsonBiokoand Rio MuniinEquatorialGuinea,westAfrica.Harvestswereassessedfrom carcasscountsinthemainmarkets intheareas.A totalof10,812carcassesof13specieswererecordedinBioko,and6160carcassesof3Ospecies wererecordedinRioMuni.Biomassofharvested mammalswas111,879.63kginBiokoand64447.87kgin RioMuni.Forthe12preyspeciesselectedforstudyinBioko,harveststotaled 7.15animals/km2or6293kg’ km2.Harvestsforthe17preyspeciesinRioMuniwere3.22anlmal.s/km2or2406kg/km2.Weusedamodel developedbyRobinsonand Redford(1991)toestimatepotential harvestsbasedonanimal productionrates. Totalproductionwas147.90animals/km2and 139.12animals/km2inBiokoand RioMun4respectively.Po­ tential harvest figuresvaried considerablybyspecies.Comparisonofactual and potentialharvestsshowed thatfiveprimatespecies(CercopithecuserythrotisCercopithecusnictitansCercopithecuspogonias,Cercopith­ ecus preussi, and Mandriflusleucophaeus) and one ungulate (Cephalophusogilbyl) in Bloko were being buntedunsustainably.Onlytwo ofthe 17species(Cercopithecusnictitansand Cephalophusdorsalis)inRio Muniwerebeing buntedunsustainably.Percentdeviationofactual from potential harvestsaveraged 498 timesgreaterthansustainableharvestinBlokoand 1.03timesgreaterinRioMuni.Forthetwositestogether figuresrangedfrom dose to28timesgreaterthanpotential to0.08%ofthepotential harvest.Althoughbunt­ ingmethodsand thecommercializationpotential ofspeciesmayaffecttheirpresenceinmarkets, thesefig­ uresshowthatBiokoanimalsareheavilyexploite4someofthemunsustainably.Thisposessevererisksfor theconservationoftheisland’sunique faunathat mustbeaddressed immediately.

ImpactodeIacazacomercialsobrelasespeciestiemamiferosenGuineaEcuatorial

Resunien:ElimpactodelacazacomercialdemamiferosdeIaselvafue estudiadoendosregiones enBioko y RioMuni,enGuineaEcuatorial,AfricaOccidentalLascosechasfueronestimadasapartirdelconteodean­ imalesmuertosenlosprincipalesmercadosdelarea.Untotalde10,812animalesmuertosde13especiesfu­ erondocumentadosenBiokoy6160animalesmuertosde30especiesfuerondocumentadosenRioMuni.La biomasadelosmamiferosrecolectadosfuede111,879 kgenBiokoy66,447.87kgenRioMuni.Paralas12 especlesdepresasseleccionadasparasuestudloenBioko,larecolleccióntotalizó7.15animales/km2o62.93 kg/km2. LarecolecciOnpara his17espedesdepresaenRioMunifuede3.22animales/km2o2406kg/km2. Utilizamosun modelo desarrolladopor Robinson yRedford(199!)para estimarlascosechaspotenciales basadas enlastasasdeproducdtinanimalLaproducción totalfuede 147.90animales/km2y 13912an­ imales/km2 enBiokoyRioMunirespectivamente.LascifrassabreIacosechapotenclalvarariaronconsi­

dne

ranbtleentrelasdistintasespedes. Lascomparacionesdelasrecoleccionesreaiesylaspotenciales mos­

traronquecincoespeciesdeprimates(Cercopithecuserythrotis,Cercopithecusnictitans,Cercopithecuspogo­ nias,CercopithecuspreussiyMandriflusleucophaeus)yunungutado(Cephalophusogilbyi)enBiokoestaban siendocazadasenformano-sostenible.Solamente2dolas17especies(CercopithecusnictitansyCephalophus dorsalis)enRioMunlestabanslendocazadasenformano-sostenible.LadesviaciOndelporcentajedorecole­ cdonrealconrespectoalpotencial,fueenpromedio498vecesmayoraldolarecolecciOnsostenida enBioko y1.03vecesmayorenRioMuni.Paralosdossitiosfuntos,lascifrasoscilaronentre28vecesmayorquelaco­

sechapotenclalaun0.08%tieíacosechapotencial.Sibienlosmétodostiecazay elpotendaltiecomercializa­ dOn puede afectar sitpresencia enlosmercados,lasdfrasactuatesmuesfran quelosanimalesenBiokoes­ tanseveramenteexplotados,algunotieloscualesenfonnano-sostenible.Estoplanteaseverosriesgosparahi conservaclOntieíasingularfaunadehiida,portoqueesteprobtemadebesertratadoenformainmediata.

Introduction Methods

FaetaL

Wildanimalsareanimportantsourceofproteininmany tropicalforest countries inAfrica(Ajayi1971, 1983; AdeolaDecker1987).Forexploitedspecies,itisim­ portantthattherateofharvestdoesnotexceed thatof productionbecause over-exploitationleadstodepletion. Harvestshould beareplaceableformofmortalityand shouldsubstituteforsomeofthenatural,annualmortal­ ityratherthanincreasingtotalmortalityofapopulation (Caughley1977).

Robinsonand Redford(1991) developed a simple modeltoprovideestimatesofpotential (sustainable) harvest ratesfordifferentneotropicalforestmammals. Theyfocusedonspeciestraditionallyimportanttosub­ sistence huntersandcalculated maximum production foraspecies(innumbersofanimals/kin2)asthenumber producedyearlyunderoptimalconditions.Theyused measuresofpopulationdensityandtheintrinsic rateof natural increasetoestimatepotential harvest rates for differentspecies.Thisprovidesafigurefortheoptimum sustainableharvestwhen productionisatamaximum andharvestinghasminimaleffectsonthenaturalpopu­ lation.Theoptimumsustainable population isthenum­ berofanimalsofaspeciesthatresultsinmaximumpro­ ductivitybutnotexceedingthecarryingcapacityofthe habitat.Assessingtheimpactofhuntingonwildlifepop­ ulationsisthuspossiblewhenfiguresgenerated bythe modelarecompared toactualharvestdata.Thesecom­ parisonsareusefulinsituationsinwhich detailedlife- historyparametersforaccurateestimatesoftheeffectof huntingonpopulation structurearenotavailable.

Incountrieswhere commercial hunting ofgamefor

human consumption is important, datafrommarkets canprovideshort-andlong-terminsightintotheimpact ofhunting onbushmenspecies.InAfrica, countsof numbers ofmammals,birds,andreptilesenteringmar­ ketscanprovide understandingofseasonalandlongitu­ dinaldynamicsofwildlifeuseandexploitation(Colynet al.1987;Kalivesse1991)andeventhebiologyofspecies (Gevaerts1992). Weexamined theimpactofharvests forsomemammalspecies byrelatingcalculatedpoten­ tialharvests withactualtakelevels.Weexamined only thosespeciesthatcomprised aminimumof1.5%ofthe totalweight ofgametakenandforwhich dataonden­ sityandlife-historyparameterswereavailable.

StudyArea

EquatorialGuineaconsists ofaterritory ontheAfrican mainland, RioMuni,(26017 km5andfiveislands.Rio MuniisborderedbyCameroon tothenorth, Gabonto theeastandsouth, andtheGulfofGuineatothewest. ThetwomostimportantislandsareBioko (formerly Fernando P00, 2017 2)andAnnobon (formerly Pa­ galü,17km2).Intacttropicalrainforestisfoundafewki­ lometersinlandfromthecoastandstillcoversmostof the country,59% of RioMimiand28%ofBioko.Al­ thoughtheamountofprimaryrainforestinBiokoislow, becausemostforestwascuttoplantcacao,mostofthe islandiscoveredinwell-conserved, tallsecondaryforest (Fal992a).

HarvestData

HuntinginEquatorialGuineaIspracticedopenlyandin­ tensivelybyprofessionalhunters, andmeat isbrought intomarketsthroughouttheyear.Unlikeother West Af­ ricancountries, therearenoclosedhuntingseasons.

FromOctober1990throughOctober1991,harvestin­ formation was collected fromtwo marketsites(Mun­ doasiandCentral)inBats,RioMuni,andfromtheprin­ cipalmarket(MercadoCentral)inMalabo,BiokoIsland. Thelattermarketisdivided intoseparatesections(Luba andRiaba).

Vertebratecarcasseswerecountedbytheauthorsand bytrained,localobservers hmiliar withallentrypoints ofbushineattothemarketsandthespeciesconcerned. Reliability was checked regularly.Species were re­ cordedbytheircommon names toavoidconfusionin nomenclature.Samplingwasconductedon424market days, 212ateachlocality.Gamewasbroughtin dailyby intennediariesbetween hunters andmarket-standpro­ prietors.Wevisitedthe marketsdailybetween 0630 hoursand1200hoursbecauseallmeatarrivestobesold between0700and1100everymorning. Only freshcar­ casseswerecounted,although somesmokedmeatis broughtin. Numbersof carcassesrecordedrepresent minimumextractionbecausesomegameisconsumedin villagesorsoldbeforeitreaches themarket(Colelletal.

1995).Ageandsexinformationwasnotrecorded.No­

menclaturefollowsHaltenorthandDiller(1987).

Analysis

Preyspecies biomasswascalculated bymultiplying the numberofanimalsbytheindividualspeciesweights.Be­ cause assome ofthe carcasses forsaleareyoung ani­ mals,thismethodtendstooverestimatethetotalweight ofmeatsold.

Huntingareasthatsupplythemarketsitesweredesig­ nated “reservoirareas.”Marketspecies restricted tonv­ erineorswamp forests(collared mangabey,Cercocebus torquatus;DeBrazza’smonkey, Cercopitbecusnegiec­

nlterateofincreaseistheexponential oftheintrinsic rateofIncrease (e,...)andistheincreaseinthepopula­ tionsizefromtimettotimet±1.Variablessuchasden­ sityandintrinsicrateofincreaseinmammalshavebeen showntobepredictablyrelatedtotheirbodymassand trophicleveloccupied (Peters1983;RobinsonRed­ ford1986). Maximumfiniterateofincreasewascalcu­ latedusingageatfirstbirth(a),ageatlastreproduction (w), andbirthrateoffemaleoffspring(b)from Cole’s (1954)equation.

rm(w+1)

ins; water chevrotain, Hymoschus aquaticus; grey-

checkedrnangabey, Lophocebusaibigena; otter, Lutra maculicolils; talapoin, Miopitbecustalapoin; Bates’ dwarf antelope,Neotragusbatesi; sitatunga, Tragela­ phus spekel) were not included in these analyses be­ causeofdifficultiesinmeasuringthesizeofthese areas. Thesizeofternfirmeareaswasdetermined,conserva­ tively,frominterviews withhunters,in bothlocalities.

InBioko, hunterssending meat to the Lubamarket

section use the Malabo/Luba districts, especially the western slopesofPicoBasiléandthenorthernslopesof theGranCalderadeLuba.Theisland’seastern districts ofBaneyandRiabaandareasstretchingintothesouth­ eastern highlandsandcoastservetheRiabamarketsec­ tion.InRioMuni,meatprimarilyfromtheLitoraLdistrict (S2,000km2)enters both markets. BecauseBiokopri­ matesarerestricted tocertain partsoftheisland,distri­ bution data gathered by Butynski and Koster (1995) wereusedtocalculatereservoirareasforthesespecies.

Productionforeachspecies wasdeterminedusingin­ formation on populationdensity at carrying capacity, themaximum rateofpopulationincrease, andtheden­ sity that producesthe maximum sustained yield.We useddataonobserved densities ofspecies, asopposed topredicteddensities (seeRobinson Redford1991). Avengedensities (numberofanimals/kin2)were taken from anextensive surveyofthe relevant literature.No densityinformation wasobtained forDendrobyraxdot- sails,Crossarchus spp., Man/s gigantea,andThriono­ mysswinderianus,sothesespecies werenotincluded. Optimum harvest wasconsideredthenumber ofani­ maLsofaspeciesthatcanberemoved(perkilometer)by humanseveryyearwithout alteringthesizeofthestand­ ingpopulationandwasdeterminedusingRobinsonand Redford’s (1991) harvest model. Tocalculate produc­ tion(P,theaddition tothepopulation through births andimmigntions),these authors assumed that realistic maximumfigureswouldoccurat60%carryingcapacity to accommodate variation related to density depen­

denceandbirthrates.Hence,

“max(0.6 DXImax)— 0.6 D,

where Disthe populationdensity and imaxthe maxi­

mumfiniterateofincrease ofthespecies. Maximumfi

1em+beTm—bC

Forthecalculation ofr,,femaleageatfirstrepro­ ductionand totalnumber ofinfantsbornperyearwere takenfrom avariety of sources. Figuresforprimates camefromreferencescontainedinRoss(1988, 1992); thoseforduikers werederivedfromPayne(1992). For rodents,pangolins,andcarnivores,information wasex­ tractedfromEstes (1991) andNowakandParadiso (1992).Annualfemalebirthrate(numberoffemaleoff­ springbornperyear)wascalculatedfrommeaninter- birthintervalandlittersize,andweassumedthatallspe­ cieshaveasexratioatbirthof1:1. Dataonageatlast reproductionisnotusuallyavailableformostspecies, butlongevityismoreoftenrecorded.Thcrcfore,maxi­ mumrecordedlongevity,L,wasusedasasubstitutefor w when we calculated rm,as employed by Ross (1992).DatausedaregiveninTable1.

Potentialharvestwascalculatedfrom productionfig­ ures.Becausethelongevityofaspeciesisagoodindex of theextent to which harvestingtakesanimals that would have died anyway (species were divided into threecategories:long-lived,>10years;short-lived,5-10 years;andveryshort-lived<5years),RobinsonRed­ ford(1991)assumedthatharvestcouldtake60% ofthe production in veryshort-livedspecies, 40%inshort- lived,and20%inlong-lived.

Exploitationlevelsofeachpreyspecieswereassessed bycomparingproductionfiguresforeachspecieswith thevolumeobservedinmarkets.ThisyieLdedafigure thatwasconvertedto apercentagetodenotethedevia­ tionofactualfrompotentialharvests.

Results

ThirteenspeciesofmammalsinBiokoIslandand29(in­ cludingtheeightriverine species)inRioMimiwere recorded(Justeetal.1995). Fivespecies (38.5%),in­ cludingendemic subspecies such asPreuss’sguenon (Cercopitbecus preussi insularis),russet-eared guenon (Cercopitbecus eiytbrotiserytbrotis),drill(Mandriltus leucopbaeuspoensls),andredcolobus(Procolabusisa­ diuspennanti),wereuniquetoBioko,whereas24(82.8%) werefoundonlyinthecontinental sites.

Table1. Bodymass,dietclassification,andreproductivecharacteristicsofmammalspeciesusedasbushmenIaEquatorialGuinea

Rodentia

*CA,carnivore,FG,fruglvore-granlvore;FH,frugfvore-berflvore;FO,fruglvore-omnivore;HB,herbivore-browser;MY myrmecophage.

Wecounted17,571 carcasses (10,812 in Biokoand

6760inRioMuni),ofwhich artiodactyls(9species)ac­

countedfor 36.7%,followed by primates (16 species,

26.2%)androdents(3species,21.2%)(fable2).

Annualharvestwas12,974animals(111,879.63 kg)in Biokoand8112(66,447.87kg)inRioMuni(p0.001). Harvestsconsisted of6.49animals/km2or55.94kg/km2 inBiokoand 3.89animals/km2or33.22kg/km2inRio Muni (cia-square test, p0.001). The blue duiker (Cephalopirns monticota)and Ernin’srat(Cricetomys

therodentsandartiodactyls;thelowestfigureswerere­

cordedforprimates.

Figure1shows the percentage deviationof actual takesfrompotential harvestforallspeciesinBiokoand Rio Munimarkets.Percentagedeviationofactualfrom

Table2. Numberofspedes(S)andIndividuals(N)Inmammalian groupsavailable Inmarketsites InBloke(LubaandRiaba)andRio MunI(CentralandMundoasi).

etnini)inBioko,andC.monticolaandthebrush-tailed porcupine (Atberurus africanus) in Rio Muni, ac­

GroupSN

Luba

SN

Riaba

SN%

Totals

countedformorethanhalfofallcarcassesbrought into

markets(Justeetal.1995).Cephalophusmonticola was themostnumerous speciesinbothmarketsites,repre­ senting approximately30%ineach. (Jricetoniys emini wasthesecond mostcommon inBiokobutrankedonly

Artioclactyla Primates Pholidota Hyracoidea Rodentia Totals

22134 34.70

71652 26.87

1 119 1.94

17 0.11

22237 36.38

136149

21779 38.15

7 1044 22.39

173 1.57

14 0.09

21763 37.81

13 4663

23913 36.19

72696 24.94

1192 1,78

1It 0.10

24000 36.99

13 10,812

eighthinRioMuni.Threespecies,C.monticola,C.em-

Central Mundoasi Totals

casseson saleon the island.Cephalophusmonticola, A.africanus,andthegreaterwhite-nosed monkey(Cer­ copitbecusnictitans)representedmorethan 59%ofall carcassesinRioMuni.

Production and potential harvest estimates aregiven

Artiodactyla 81294 41.29 81594 43.95 8

Primates10 790 25.21 II 738 20.35 11

Carnivora443 1.37 478 2.15 5

Edentata256 1.79 256 1.54 2

Hyracoidea11 0.03 11 0.03 1

Rodentia3 950 30.31 31160 30.86 3

2888 42.73

1526 22.58

121 1.79

112 1.66

2 0.03

2110 31.22

InTable3.Thehighproductionfigureswere typicalof

ConservationBiology

Volume9,No.5,Octoberfl)95

TOTALS28313429 362730 6759

Table3. ComparIsonsofobservedhrvcst,cakvlatcdproduction,andsustalanbfrharvestforbushineatspecIesinEquatorialGuinea.

Observed / Observed / Animals
Harvested / Biomass
Harvested / Production
Siteand Species / Density(nolkm2) / Harvest(no/yr) / (perkm2) / (gms/1m2) / (no,m2)
BIOKO
PrImates
Cercopitbecuserytbrotis / 24.70 / 781 / 0.16 / 683.68 / 2.64
Cercopfthecusnictltans / 22.70 / 254 / 0.49 / 4263.00 / 0.12
Cercopithecuspogonias / 14.40 / 52 / 0.56 / 753.76 / 0.08
Cercopitbecuspreussi / 9.90 / 196 / 0.65 / 6272.50 / 1.51
Cobbussatanas / 20.40 / 514 / 0.29 / 3480.00 / 2.72
Mandrillusleucopbaeus / 6.70 / 551 / 0.29 / 4328.25 / 0.79
Procobobuspennanti / 156.30 / 348 / 0.66 / 5273.07 / 14.09
Rodentla
Atberurusafricanus / 55.00 / 1581 / 0.79 / 3160.00 / 27.12
Cricelomysemini / 134.00 / 2419 / 1.21 / 2420.00 / 81.49
Ardodactyla
Cepbabopbusogilbyi / 13.00 / 3181 / 1.59 / 30210.00 / 2.02
Cephabophusmonticola / 22.60 / 732 / 0.37 / 1850.00 / 8.68
PholMota
Manistricuspis / 10.90 / 192 / 0.10 / 240.00 / 6.63
TOTAlS / 490.60 / 10,801.00 / 7.15 / 62,934.26 / 147.90
RIOMUM
Cercopitbecuscephus / 17.30 / 430 / 0.22 / 1193.25 / 1.18
Cercopitbecus mona / 13.00 / 67 / 0.03 / 91.56 / 0.89
Cercopitbecus nictit.ans / 22.70 / 523 / 0.26 / 2275.05 / 1.55
Cobbussatanas / 20.40 / 152 / 0.08 / 912,00 / 2.72
Gorillagorilla / 0.70 / 3 / 0.00 / 176.32 / 0.03
Mandrillussphinx / 6.70 / 262 / 0.13 / 2153.64 / 0.79
Pholidota
Manistrlscuspis / 10.90 / 93 / 0.05 / 116.25 / 6.63
Ardodacb
Cephabophusdorsalls / 3.80 / 491 / 0.25 / 4910.00 / 0.50
Cepbabopbusleucogaste / 4.10 / 6 / 0.00 / 40.50 / 1.31
Cephalopbussylvicultor / 0.90 / 6 / 0.00 / 187.50 / 0.29
Cephubopbusmonticola / 22.60 / 2107 / 1.05 / 5267.50 / 8.57
Potamochoerusporcus / 3.10 / 93 / 0.05 / 2790.00 / 1.89
Rodentla
Atberurusafricanus / 55.00 / 1698 / 0.85 / 3396.00 / 27.12
Cricetomysemlni / 134.00 / 390 / 0.20 / 390.00 / 81.49
Carnivora
Genetta servallna / 0.80 / 32 / 0.02 / 16.00 / 0.45
Nandiniabinotata / 3.60 / 72 / 0.04 / 72.00 / 2.01
Civecitictiscivetta / 2.80 / 15 / 0.01 / 75.00 / 1.68
TOTAI.S / ,322.40 / 6440.00 / 3.22 / 24,062.57 / 139.12

potential harvests averaged 498.99% in Bioko and

103.27%inRio Muni.Huntinglevelsdifferedbetween taxonomicgroups;Prinutesandungulates appearedto bethemostexploited.InBloko,estimatesofactualtake forallexcept fourprimates and oneungulate(30.7%of allrecorded species) were below potential harvests. Takesrangedfrom28timesgreaterthansustainablehar­ vestforthecrownedguenon(Cercopitbecuspogontas) to0.96timeslessthanthesustainable harvestforEmin’s

Rio Mum presentsadifferentpicturethanBiokobe­ causeonlytwospecies (11.77%)were beingexploited in an unsustainablemanner:Cercoplt1.ecus nictitans andthebayduiker(Cephalophus dorsalts).Mostspe­

des werehuntedbelow sustainableharvests.Takes rangedfrom 13limesgreaterthanpotentialharvestfor C dorsalis to0.67 timesbelow potentialharvestfor Cepbalopbusleucogaster.

Discussion

Thereare fewdetailedstudies atpresentontheimpact ofhumanexploitationongameanimalsinwestAfrican countries.Assessmentsofoptimalsustainable harvests usingtheRobinsonand Redford(1991)modelcan offer well-founded baselineinformationforregulatingpopula­ tion-uselevels.Assuggested byRobinson andRedford

1112MarkeilluntingofAfricanBushmeatspeeiesPaelat.

Cpo CprCogMieCnICsaCerPpeAelCmoMrCern

Cni Cdo CceMapGgoCmnQseCmo Aaf Csa Nbi Ppo Mfr Cci Cam Cay CI.

Species

Figure1. Deviation of actualharvestfromcalculatedpotentialharvestsforbusbmeatspeciesinBioko(a)andRio Muff(b).Speciesabbreviations:AafAtherurusafricanus;Cce,Cercopithecuscephus,Cci,Civecttictiscivetta;Cdo, Cepha[ophusdorsaiis,Cer,Cercopithecuserythrotis;Cern,Cricetomysemini,Cle,Cephalophus[eucogaster;Cmn,Cer­ copithecusmona;Cmo,Cephalophusmonticola,-Cm Cercopithecusnictitans;Cog,Cephalophusogilbyi,Cpo, Cercop­ ithecuspogonias;Cpr,Cercopithecuspreussi;Csa,Colobussatanas,Csy,Cephalophussylvicultor, Ddo,Dendrohyrax dorsalisGgo,Gorillagorilla;Gse,Genettaservalina;Mie,Mandrillusleucophacus;Msp,Mandrillussphin,çMtr,Manis thcuspis,Nbi,Nandiniabinotata;Ppe,Procolobuspennanti;Ppo,Potamochoerusporcus.

ConservationBinlogy

Vc.lume9,Nc’,5,Ocer19

(1991),eventhoughthedataaresubjecttovariationasa resultofinaccuraciesinestimation, theassumptionsin themodelareinlinewithproductivityawlharvestinfor­ mationfromotherecosystems.Acautionarynoteisnec­ essary,however.Potential harvests, the maximum bio­ logicallypossible productionfor agivenspecies under ideal conditionscalculated herein, incorporatefigures thatrelatetoreproductionparametersassumedtobeac­ curate andthatuseavengepopulationdensities.

Currentdatapointtoacleardifferenceinexploitation

ofspecies betweenBioko andRioMuni.Varianceinto­ talnumberofpreyspeciesrecorded betweenislandand continentalmarkets relatesdirectly tothe richer verte­ bratefaunainRioMuni(Fa1992a).Butit isclearthatall marketsrelyheavilyonthesaleoftwospecies, anante­ lope (Cepbalophusmonticola)and arodent (Criceto­ mysemini)inBioko andthesameantelopeandanother rodent(Atberurusafricanus)inRioMimi.

Unless reproductiveparametersreportedherein are

verydifferentfromactualfigures,thecurrent rateof hunting iscertainly farbeyondthe abffityofsomespe­ ciestosustainpopulations. Ourfiguresfortheduiker species in Bioko,are below those observed by Payne (1992)fromastudyinnearbyKorup, Cameroon. Payne (1992) calculated a higher take of 11.5-13.2 times greater thanthesustainable harvestforC ogilbyi (1.96 timesgreaterInourstudy)and1.3-2.2timesgreaterfor Cnwnticola(0.11timesgreaterInourstudy).Thesitu­ ation isdifferent for primates because levelsforsome Biokospeciesarearound twotimesabovepotential har­ vest,buttakeforCpogonlasis28timesabovesustain­ ableharvest.Thisspecies, together withC.nictitans,C. perussi,andMandrilluleucophaeus,isunsustainablyex­ ploited; thelasttwospecies areamongthemostendan­ geredprimatesinAfrica(Oates1986;Leeetal.1988).

Exploitation levelsgeneratedfrom themodelassume generallyunburnedconditions, butinrealityactualden­ sitiesarewellbelowwhatthemodelassumes,andthere- foretheeffectofhuntingonthese speciesmustbeseen ascvenmore worrisome. Becauseanydisruptionofthe population(skewed sex ratios,isolatedsmallgroups of

EquatorialGuinea(Castroviejoetal.1986;JusteCan- tern1991;Fa I992a). Nevertheless,proteindeficiency exists,particularlyawayfromthemain urbancenters (CooperationFrançalse1984),becausehalfof allprotein intheurbanareasisderivedfrombushineatwhereas the sameamount invillagescomesfromprotein-rich oleagi­ nousgrains(peanutsandsquash).Thisseemingcontra­ dictionistheresultnotofa lowper-capitaavailabilityof meatinruralareasbutofanemphasisonsellingpreyin towns because ofthegreater purchasingpower there. Thisisclear ifone comparesthe volume ofextracted meatbetweenlocalities.ThemeatvolumeattheMalabo markets inBiokowas 70%higher than the volume in Bata,despite onlyaslightdifference inpopulationsize betweenMalaboand Barn(52,000 and 55,000, respec­ tively;datafromFNUAP).Malaboistheseatofgovern­ ment andforeign aidorganizations;therefore, ithasa more urbanized populationwithgreater buyingpower. Thus,the rateofmeatextractionrespondstothe eco­ nomicpotentialofaproportionoftheurbanpopulation rather thantothenutritionaldemands ofthepopulation atlarge.

Attributes of species that influence hunters’ prey choiceare difficulttodetermine, butspeciesfoundin marketsbothontheislandandincontinentalareasmay representonlythoseanimalsthatfailwithinaparticular weightrange,dictatedby huntingchoiceoreconomic constraints(such astransport costs) rather than—or as wellas—market preferences.JusteandCantero (1991) arguefortheexistenceofanoptimalpreyspectrumde­ termined bymeatsale prices between1000 and1500

FCFAperkg($US3-5).Thus,smallerpreysuchassquIr­

rels(8speciesinBloko andiiinRioMuni) andbats(Fa

1992a), aswellasthelargeanimals(>200kg)areex­ cluded, andmoreemphasisisputonmedium-sizedani Both size extremesrequire specialized hunting techniques.Thehirgerspecies (forexample, theforest buffalo[Syncerus caffernanus]andthe forestelephant [Loxodonta africana c-yclotisJrequireanexaggerated investmentinextractionandtransporttomarket,Hence, species appearingin markets are generally easily

animals,orincreased mortalityofpre-reproductIveani-tracted through theuseoftrapsandsnares(rats,porcu­

mals)mayhavealreadyaffectedproductioninthereser-pines, duikers) or have acceptableeconomicreturns. voirareas, over-hunting ofsome species maybe even Peop1es preferenceforcertain meatsmayalsoaffectse than we have shown. Furthermore, be- lectionofspeciestakentomarkets(SabaterPiGroves

causeactual harvestfiguresare minimal(more animals thanappearinmarketsaretaken),current ratesofhunt­ ingfaroutstrip the abilityof some populationsto re­ place the animalskilled.Datafrom Colelletal.(1995) demonstratethat whereas 80%ofantelopescaught in villagesinsouthernBioko weresenttomarket, just10% of thesmallergame(Criceromys,Manis)weresoldat market.Thishasalsobeen observed inZaire(Colynet a!.1987).

Our study hasfurther confirmed the importanceof bushmcat asasubstantial source ofanimal protein in

1972).Acombinationofthesefactorsaffectsthevarying proportionsofspeciessoldinmarkets.

Hunting technology andmethod (snares versusguns) arerelatedtothesizeoftheanimal.Therefore, aplausi­ bleargument toexplain over-exploitationofthearbo­ realguenons andforestbaboons isthathunters actively these animals. Snaretrapping does not require mucheffortandcanbeundertakensimultaneouslywith gun hunting. Snaresarechecked atone-week intervals; thus, huntershave more time to stalk arboreal game (Cold etal.199).

1114MarketHuntingofAfrfcanBushmeatSpeciesPactaL

InRioMuni,over-huntingseemslesssevere,butas earlyas1968Sabater P1andGroves(1972) notedthat heavypredationofguenonsby humanshadmadethese smallmonkeysrareinsomeareas. Afurther indication oftheimpactofhuntingonBiokoguenonsIsthatdensi­ tiesinundisturbedareasareshowntobethree tofour timeshigherthaninareaswherehunting ispresent(Bu­ tynskiKoster1995),despItethefactthatguns andam­ munition are not readilyavailable(Juste Cantero

1991).

Ifharvestcontrolsarenotimposed,increases inhu­ manpopulation,changesinaboriginalsubsistencemeans, andsubsequentdeforestationwillimpoverish native fauna,astheyhaveinothertropicalcountries(forexam­ ple,Panama;Bennett1968).Ingeneral,fiunal exploita­ donforsubsistenceinWestAfricaisconsideredhighby severalauthors(Mittermeier1989),butitisthegreater emphasisonhuntingforprofitthatisdisquieting.This hasbecomemore conunonpiacebecause commercial huntingoffersasignificantmonetaryincentivetorural people.Wilkieetal.(1992) hasshown thathuntingin Congoconstitutesthemajorrevenueforupto51%of theemployeesofamajor loggingcompany.Without theseearningsmanyfamilieswouldbeunabletobuyba­ siccookingutensils,clothing,medicine,andeducational materialsforschool-agechildren.

Aneffectivewaytoprotectandexploitgamesustain- ablyinEquatorialGuineamaybethemaintenanceofal­ readydecreedprotectedareas,especiallyontheisland, wherecontrolofhuntingisvitalforconservingtheen­ demicfauna.Established,protectiveboundaries(Castro­ viejoetal.1986;Fa1992a,199Th)mustbemadeeffec­ tive, althoughexploitationin someareascouldcontinue onarotationalbasisinwhichMiowareasserveasrepro­ ductiverefuges.Mostareas,however,atleastinBioko, arebeinghuntedcontinuouslyandintensively,andfew sitesmayremainassourcepopulationsthatcouldre­ populatetheheavilyhuntedareas.Dataonactualdensi­ tiesandharvestratesfromspecific sitesareurgently needed.Moreimportant,althoughItwillbemorediffi­ cult,huntingandmonitoringregulationsneedtobees­ tablished.

Acknowledgments

WethankJoséAngueMangueinBiokoandRafaelDel­ gado,AntonioOndó,PascualSacogo,andCornelloCis­ nerosinRioMuniforcollectingthedailymarketinfor­ mation.WearegratefultotheMinisterof Cultureofthe Republicof Equatorial Guinea,Excmo,LeandroMbo­ nile,andtheDelegadoContinentallilmo.WethankJosé Obamaforhelpandsupport.Thisstudywasfundedby theO.C.G.E.,atpresentthe Institute deCooperación panciDesarroilo(LC.D.)oftheSpanishMinisteriotie AsuntosExteriores.WealsothankCarolineRossforpro-

vidinguswith computerprogramsandJohnRobinson andJohnPayneforreviewingearlierdraftsofthemanu­ script.

LiteratureCited

Adeola,M.0.,andE.Decker.1987.UtilisationtieIafaunesauvageen milieururalanNigeria.NatureetFauna3:15-21.

Ajayi,S.5.1971.WildlifeasasourceofproteininNigeria.Nigerian

FIeld36:115—127.

Ajayi,S.5.1983.Domestication ofmammalsinAfrica.NigerianField

47:145-156.

Bennett,C.F.1968,Human influencesonthezoogeography ofPan. ama.Ibero-Americana,no.51.UniversityofCaliforniaPress,Berke­ ley.

Butynski,T., andS.Koster.1995.Distributionandconservationstatus ofprimatesinBioko.BiodiversityandConservation3:393-409.

Castroviejo,J.,J.Juste,andR.Castelo.1986.Proyectodeinvestigación

yconservaciónIanaturalezaenGuinea Ecuatortal.Ministériotie Asuntos Exteriores, Espada,Secretaria tieEstadopan Ia Cooper­ aclónInternacionalypantheroainérica—OficinadeCooperaciOn conGuineaEcuatorlal,Madrid,Spain.

Caughley,6.1977.Analysisofvertebratepopulations. Wiley,Chiches­

ter,England.

Cole,L.C.1954.Thepopulationconsequences oflifehistoryphenozn

ena.Quarterly ReviewofBiology29:103-137.

Coleil,M.,C.Mate,andj.E.Fa.1995.HuntIngbyMokaBubisinBloko: dynamicsofexploitationatthevillagelevel.BiodiversityandCon­ servation3:939-950.

Colyn,M.,A.Dudu,M.Mankoto,andM.A.Mbaelele. 1987.Exploita­ tiondopetitetmoyengibierdesforersombrophilesduZaire.Na- toreerFaune3:22-39.

Cooperation Française.1984.AnalysedeIssituation agro-pastoralede

Iaregioncontinencale(RioMimi):propositionsd’actions.Repub­ liquede GuineeEquatoriale, Mlnistèrede1’Agriculrureetde Ia PEche,CeiluiedeConceptiondeProjetsdeDéveloppementRural, Malabo,EquatorialGuinea.

Estes, R.1991.ThebehaviorguidetoAfricannnmznals.Universityof

CaliforniaPress,Berkeley andLosAngeles.

Fa,J.E. l992a.ConservacidndelosccosistemasforestalesenGuinea

Ecuatorial.WorldConservationUnion,Gland,Switzerland.

Fa,J.E.1992b.ConservationinEquatorialGuinea.Oryx26:87-94. Gevaetts, I-I. 1992. BirthseasonsofCercopitbecus,Cercocebusand

ColobusinZaire.FoilaPrlmatologia59:105-113.

Haltenorth,‘1.,andH.Diller.1987.LargermammalsofAfrica.Collins,

London.

Juste,J., andJ.Camera1991.Isoformsnacionalsobrernedioambiente ydesarrolo:GuineaEcuatorlal.ConferencladelasNacionesUnidas sabre MedicAmbienteyDesarrollo(UNCED),Malabo,Equatoriai Guinca.

juste,J..J.F.Fa,J.PerezdelVal,andJ.Castroviejo.1995.Marketdy­

namicsofbushmeatspecies inEquatorialGuinea.JournalofAp­

pliedEcology 32:454-467.

ICalivesse,A.1991.L’approvisionnementdesmarchescitBanguienvi­

andedechasse.NatureetFaune17:14-20.

Lee,P.C.,F.Bennett,andJ.Thonback. 1988.Threatenedprimatesof Africa.WorldConservation UnionspeciesMonitoringCentre,Cam­ bridge,England.

Nowak, R.M.,and J. L. Paradise.1992. Walker’smammals ofthe world,vols.III.TheJohnsHopkinsUniversityPress,Baltimore andLondon.

Ones,J.F.1986.I1JCN/5SCprimatespecialistgroupactionplanforAt ricanprimateconservation:1986-90.World Conservation Union

(IUCN)/WorldWideFundforNature,Gland,Switzerland.

ConservationBiology

Vob,nie9,No.5,October 5995

Faeta!MarketHuntingofAfrtcanBushtneatSpecies1115

Payne, 3.1992.Afieldstudyoftechniquesforestimating densitiesof duikersinKorupNationalPark,Cameroon. M.S.thesis.University ofFlorida,Gainesville.

Peters,It H.1983. Theecological implicationsofbodysize.Cam­

bridgeUniversityPress, Cambridge,England.

Robinson,). G.,andK.H.Redford.1986.Bodysize,dietandpopula­

tiondensityof neotropicalforestmammals.American Naturalist

128:665-680.

Robinson,).G.,andK.H.Redlbrd.1991.Sustainableharvestofneotro­ picalforestanimals.Pages415-429inJ.C.Robinsonand K.H. Redford,editors.Neotropicaiwildlifeuseandconservation.CM-

cagoUnIversityPress,Chicago.

Ross,C.1988.Theintrinsic rateofnaturalincrease andreproductive

effortin primates. JournalofZoology(London)214:199-219.

Ross,C.1992.Theintrinsicrateofnaturalincrease andreproductive

effortin primates.Oecologia214:199-219.

SabaterPt,J.,andC.Groves.1972.Theimportanceofhigherprimates in thedietofthefangofRioMuni.Man7:239-243.

Wiikie,0.5.,J.G.Sidle,andG.C.Boundzanga.1992.Mechanizedlog. ging,markethunting, andabankloaninCongo.ConservationBio­ logy6:570-580.