focus and role of research in panda conservation - reid

7/21/2019 Focus and Role of Research in Panda Conservation - Reid

http://slidepdf.com/reader/full/focus-and-role-of-research-in-panda-conservation-reid 1/11

THE FOCUS AND ROLE OF BIOLOGICALRESEARCHINGIANTPANDACONSERVATION

DONALDG. REID,Ecology Group, Departmentof Zoology, Universityof BritishColumbia,Vancouver, BC, CanadaV6T 2A9

Abstract. The Ministryof Forestryof the Peoples Republicof Chinaand WorldWide Fundfor Naturehave cooperatedsince 1980 in an effortto conserve the giant panda (Ailuropodamelanoleuca)in the wild in China. This conservationprojecthas 4 majorcomponents: biological

research,populationsurvey, managementplanning,andtraining. Thispaperfirst evaluatesthe focus and resultsof thebiological researchusinga frameworkbased on populationviability analysis andlife-historytheory. Demographicparametersandthe causes of their variationare stillpoorly understood. A number of habitat-relatedcological processes arerelativelywell understood. Second the paperassesses the dominantrole of biological researchin the project. The principalthreatsto panda populationviability are anthropogenic:habitatloss and poaching.However, this conservationprojecthasnotsufficientlyaddressed hesocio-economicconditionsandbehaviorsthatcauseandinfluencethethreatsto pandapersistence. Incorporating ocial scientistsalong withbiological scientistsin a teamof investigatorsat the inceptionof a conservation

project should make the projectmore successful.

Int. Conf.BearRes.andManage.9(1):23-33

The Chinese Ministry of Forestry (MoF) and theWorld Wide Fund for Nature(WWF)have cooperatedsince 1980 in efforts to conserve the giant panda n the

Peoples Republic of China (PRC). Wildlife

Conservation International(WCI) cooperated in this

project from 1980 to 1986. The project goal is to

conserve a viable wild populationof giant pandas in

perpetuity, thereby avoiding extinction of this

endangeredspecies.The project has included more than 12 years of

cooperativeresearch,a population urvey, andan initial

managementplan producedby the 2 principalagencies(MoF and WWF 1989) and modified by the MoF for

ratificationby

the ChineseStateCouncil in 1992(MoF1992). However, panda populations and habitat

continue to shrink,andthe conservationprojecthas notreversedthis trend(Schaller1993). It is appropriateo

analyze the project, to determinewhether effort waswell spent, what componentswere lacking, and howbest to proceed. Currentlythe persistence of wild

populationsof all ursids in southeastAsia and SouthAmericais in seriousdoubt(Servheen1990), andthese

species are receiving or will receive conservationattention.

Between 1985 and 1988 I worked for WCI and

WWF as a field research scientist on the pandaconservationproject in Sichuan, China. In this papermy intentionis to evaluatesome of the successes andfailures of the biological research and managementplanningaspectsof the project. My specific objectivesare to: (1) briefly describethe componentsof the giantpanda conservationproject; (2) evaluate the focus ofone component, biological research, in the light of

populationviabilityanalysisandlife-history heory;and

(3) commenton the role of biological researchrelativeto other project components in attainingthe projectgoal, as judged by actions recommended in the

managementplans.The researchreviewed was supportedby the World

Wide Fund for Nature International,World WildlifeFund Canada,WildlifeConservationInternational,andthe ChineseMinistryof Forestry. G.B. Schaller,A.H.

Taylor, K.G. Johnson, and an anonymous reviewer

providedvaluablecriticalcommentson an earlierdraft.

COMPONENTSOF THECONSERVATIONROJECT

BeforeWWFandWCIbecameinvolved, the Chinese

governmenthadputconsiderableeffort into giantpanda

conservation. Twelve naturalreserves,with the goal ofconserving panda habitat, were established between1963 and 1979 (Schalleret al. 1985). A preliminarypopulation survey in 1974 and 1975 produced an

estimate of 1,000-1,100 animals in the wild, and

demonstratedhatpandadistributionwas fragmentedon2 scales as a result of expanding human activities inmountainvalleys. Insurmountable aps of agriculturallandandmajorriversseparatedpandas iving in each of6 mountainranges (Fig. 1). Each range had its owndistinctive panda habitats-montane forests with

understory of various bamboo species growing in

altitudinalbands. At a smaller scale within a range,loss of low elevation forests had forced remainingpandas to higher elevations with little opportunityformovement between habitat fragments. ThePRC/WWF/WCIgiant pandaproject, begun in 1980,has 4 components: biological research, populationsurvey, managementplanning, and training.

Biological ResearchThe biological researchhas employedfrom 10 to 50

people continuouslyfrom 1980 to the present, and hasbeen by far the largestcomponent. The field research



24 Int. Conf.BearRes.andManage.9(1) 1994

Fig. 1. The 6 mountainranges in southcentralChinawith giant pandahabitat,and locations of 12 pandareserves established

by 1980.

consistedof 2 studies of wild pandas, 1 in each of the

2 largest mountain ranges with pandas: WolongReserve in the Qionglaishan,and TangjiaheReserve in

the Minshan(Fig. 1). In additiona captive facilityandresearch laboratorywas built in Wolong Reserve to

breedpandasandstudy reproductive ehavior. Overall,the goal was to provide data on ecological processesand natural behaviors most affecting population

dynamics,and relate these data to managementactions.

Initially he researchconsistedof quantifyingpatternsof behavior, and long-term monitoring of known

individualsto quantifypopulationparameters Schalleret al. 1985). Studies began in Wolong, but ran

concurrentlyin Tangjiaheafter 1984 (Schaller et al.

1989). From 1983, research focussed mainly on 3

topics: (1) the impactof bamboofloweringon behavior

and food resources(Johnsonet al. 1988a; Reid et al.

1989; Taylor and Qin 1987, 1988c, 1989a,b, 1992,

1993a,b; Taylor et al. 1991a,b), (2) panda habitatselection and habitat structure(Taylor and Qin 1987,

1988a,b, 1993b;Reid and Hu 1991; Reid et al. 1991c),and (3) community ecology focussing on potential

competitive interactionsof giant pandas with Asiatic

black bears (Ursus thibetanus) (Schaller et al. 1989,Reid et al. 1991a) and red pandas (Ailurusfulgens)

(Johnsonet al. 1988b, Reid et al. 1991b).

Population SurveyFrom 1985 to 1988 a team of 35 biologists, led by

Ken Johnson, Shao Kaiqing, Liu Zhixue, and Jiang



EVALUATIONOFGIANTPANDACONSERVATION Reid 25

Fuquan,undertooka groundsurveyof the entirepanda

rangeto mapthe abundanceanddistributionof pandas,their bamboo and forest habitats, and the activitiesof

people in and around these habitats. The team used

line-transect sampling of bed-site densities as a

populationestimation echnique. Thiswas derivedfromthe detailedbehavioralstudiesin Wolong Reserve. The

initial and revised managementplans (MoF and WWF

1989, MoF 1992) used the dataon habitatdistribution

and preliminarypopulationestimates as the basis for

detailed recommendations on land tenure changesnecessaryfor pandaconservation.

Management PlanningThe managementplanning exercise began in 1986

when a team of approximately 10 biologists and

managers ed by Dr. JohnMacKinnonand Bi Fengzhoustarted integrating information from the biological

research,thepopulation urvey, investigations f habitat

distributionand change using remotesensing (De Wulf

et al. 1988), and the legal and institutional ramework

for wildlife conservation in the PRC. They producedthe Management Plan for the Giant Panda and its

Habitat (MoF and WWF 1989). The initialdocument

has been modified somewhat or formalapprovalby the

Chinesegovernment MoF 1992). However, the majorrecommendations (Table 1) are not changed.Successfulpandaconservationwill dependon judicious

implementation f the latest MoF plan.

TrainingForeignfield biologistsbecame nvolved in thepanda

project to teach Chinese counterpartsradio telemetryand ecological samplingskills. Theyhave acted as role

models for scientists in China, where wildlife biologyis a rare and poorly understoodvocation. In 1986 thiselement of the training expanded to include 4 other

programs: (1) a lecture and field course in protectedarea managementdesigned for reserve directors and

higher officials (Machlis and Marsh 1988); (2) field

courses in biophysicalinventorytechniquesfor reservebiologists; (3) field courses in patrollingreserves for

wardens; and (4) sending young Chinese biologists tooverseas universities for graduate level degrees inresourcemanagement.

THEFOCUS OF THEBIOLOGICALESEARCH

Given limitedtime and resourcesand the intricaciesof ecological processes, conservationresearcherswiththe responsibilityof ensuring persistence of a species

mustcarefullyanswerthe question: on whatbiological

processesshouldwe focus our research? Two schemes

help provide a comprehensiveand systematicapproachto identifyingprocessesaffectingpopulationpersistence.These are populationviability analysis (Shaffer 1981,

1990; Gilpin and Soule 1986), and life-history theory(Southwood 1988). Many biologists would implicitly

apply these general approaches, at least in part, in

answering the above question. An explicit,

comprehensive scheme of options should enable

researchers o betterdefine researchpriorities, specify

objectives, and evaluateprogress.

PopulationViabilityAnalysisGeneral.-Population Viability Analysis (PVA) is a

comprehensiveapproachidentifying demographicand

genetic factors that put a population at risk of

extinction, and describing how these factors operate

(Shaffer1981, 1990; Gilpinand Soule 1986). The key

parametersto be quantified are population carrying

capacity (K), population growth rate (r) or the

differencebetween ratesof birth(b) anddeath(d), and,most important,he variance n r (Vr) (Goodman1987).When estimating K, biologists must differentiate

between saturationof the habitatwith respect to food,and saturationwith respectto other conditions such as

social spacing. Substantiallydifferent estimates mayresult. Vr affects extinction probabilitydramatically

because it incorporatesall the stochasticenvironmentaleffects on birthand deathrates.

Systematic factors affecting birth and death rates,such as human-inducedhabitat loss or mortality, are

assumedincorporatedn b and d as measured,and are

generallynotconsidered urther n PVA (Shaffer1981).In practice this does not seem reasonable since both

systematic and stochastic environmentalchanges are

incorporated n empirical measures of b, d and Vr.

Unless dataare from a pristineenvironment,it will be

very hardto disentangle he two. Managementactionsare best suitedto addressingsystematicfactors.

To apply PVA, researchers must quantify thedistribution n time and space of all populationsof a

species (the metapopulation tructure),the flux withineach of thosepopulations a numberof separatePVAs),and the flux between populations. Published PVAshave requireda long time series of demographicdata,andthe researchers tressed he dangersof extrapolatingbetween populationsor assuming that the data were

sufficiently comprehensive (Shaffer 1983, Menges1990, Murphy et al. 1990). The task of acquiringsufficient data for a PVA in the time available is

virtually impossible for long-lived, secretive animals



26 Int. Conf.BearRes. andManage.9(1) 1994

Table 1. Goals, methods, and adequacy of knowledge for proposed giant panda management plan interventions in 5 sets of

institutions. ? means the issue was not addressed in the plans. Adequacy of knowledge is subjectively judged as good (G),

moderate (M), or poor (P).

Institution Goal Methods Typeand adequacyf knowledge

Landtenure Increasereserve size Relocate or phase-outtimber Locations Gcuttingunits Compensation economics M

Establishnew reserves Resettle communities Resettlementeconomy P

Resettlementculture P

Harvesttrees by selection Legislation/regulations Harvest economics P

logging Technology ? Forest ecology M

Economic Develop alternativeenergy Hydro-electricor solar ? Engineering P

Foster common property Collectiveownership? Communityorganizationand P

resourcemanagement Sustainedharvest? traditions

Education? Poachingeconomics P

Ecotourism Tourist economics M

CulturalFoster sustaineduse Collective

ownership?

Communityorganizationand P

Traditionalvalues ? traditions

Education

Legal Establishdisincentivesfor Legislation Legislation G

poaching, habitat Law enforcement Sociology and economics of P

degradation patrolling

Political Fosterpoliticalwill Education Power structures M

Diplomacy Culturalattitudes P

Personaltrust Developmenteconomics P

Economic incentives

such as the giant panda, and mostursids.However, parametervalues for r, K and Vr can be

approximated from allometric relationships among

mammals Belovsky 1987). Populationprojectionscan

thenbe madewithoutfield study. Althoughthesemust

be viewed as tentative projections, the sensitivity of

persistence to changes in parametervalues can be

estimated from repeated model projections (Shaffer

1983, Bunnell and Tait 1980). Populationviability

analysis is a useful conceptual frameworkbecause it

forces researchers to focus on all demographic

parameters

n buildinga model, andresearchers anuse

empiricaldatadirectlyin projections.Panda Project.--The cooperative research has

gatheredsome datapertinentto a PVA. The dataon

metapopulation tructureand change are the best. The

survey team mapped the distributionof all panda

populations,estimateddensity for each, and quantified

habitat oss acrossthe entirerangesince the mid 1970s.

There are 24 distinctpopulations,most of which have

less than 50 animals (MoF 1992). Such small

populations,even if left intact,have low probabilityof

persistence or demographic Belovsky 1987, Goodman

1987) and genetic reasons (Lande and Barrowclough

1987).The population evel demographicdataare generally

poor because of low sample size, resultingfrom poor

live-capture success. Future efforts must focus on

novel techniquesandbaits. Density estimatesfrom the

Wolong study area give 1.9 km2/pandaat peak food

availabilitybefore bamboo die-back (Schaller et al.

1985), and 2.1 km2/pandaafter loss of 82% of the

standingcrop of preferredbamboo in a die-back, at

which point panda feeding exceeded the regeneration

capacity of the bamboo (Reid et al. 1989). Carrying

capacity (K) is apparentlymediated by interference

competition some exclusive space use and marking)at

peak food availability, and exploitative competition

when food is more scarce. Absolutedensity estimates

are difficult to make and often lack precision. For

monitoringspecific populations,managerswill have to

rely on indicesof relativeabundancebased on spoor.

In 5 of the 11 parturition easons (involving 5 adult

females) it was unclearwhetherthe femalehad actually

given birth. In each case she was not accompaniedby

a yearling,buther movementsnever suggestedshe was

using a maternityden. Since timing of parturition s

quitevariable n this species, female(s) could have lost



EVALUATIONOF GIANTPANDACONSERVATION Reid 27

cub(s) soon after birth. A pregnancytest, for use on

wild pandas, would help clarifythe birthrates.

Certainreproductivehormones are found in higherconcentrations n feces during mid to late pregnancythan at other times. This is true of estrogen in feral

mares(Kirkpatrick t al. 1990), cattle(Bos taurus)andmuskoxen (Ovibos moschatus) (DeSaulniers et al.

1989), andof progestinin caribou(Rangifertarandus)

(Messier et al. 1990) and some primates Wasseret al.

1988). These assay techniques should be tested on

pandas.Neonatalmortality s high in the wild. Only 1 of 6

knownjuveniles survivedits firstyear. The fates of 4

were never determined.Sinceyoung pandasare altricial

at birth, not moving in a coordinated, independentmanner until approximately90 days (Schaller et al.

1985), they could be subject to predation and

hypothermic stress when unaccompanied. One was

killed by a predator. This may have been a case of

infanticide Catton1987). Leopards Pantherapardus)

occasionally eat young pandas (Schalleret al. 1985).These risks will depend in part on the availabilityof

suitablematernitydens (in hollow conifers or caves).Similar evidence for low neonate survival comes

from a life table analysis, based on aged pandacarcasses found in the wild, which shows poor

survivorship of young to 1 year of age (Wei et al.

1989). Survivalof captive-born oung is alsopoor. Of

37 litters (51 young) born in zoos between 1963 and1983, only 19 young survived beyond 2 months

(Schalleret al. 1985:159). Futurework should focus

on monitoring cub survival, perhaps through radio

telemetry, and the experimental enhancement of

maternityden abundance,especially in habitatswhere

hollow trees no longer exist.

High neonatal mortality is also associated with

inbreedingdepression n some species(RallsandBallou

1983). Data on geneticprofilesof captiveor wild giant

pandas are lacking. This must be rectifiedespeciallysince most populations are already small and

fragmented.Seven of 11 radio-collared adults died, 2 from

poaching, 2 from parasite infestation, and 3 of other

naturalcauses. The most troublesomecause of adult

death is poaching, which has increased dramaticallysince 1985, with 56 pandasknownpoachedthrough he

species' entire range from 1984 to 1988 (Qiu 1990,Schaller 1990). Increased poaching seems directlyrelated to economic liberalization n the PRC, which

replacedstate-controlled ricingandmarketing f manynaturalresources, such as timber(Repetto1988), withfree access to domesticandexportmarkets. The export

marketfor panda pelts, though illegal, now operates

throughmiddlemen n Sichuan. National legislation to

deterpandapoaching, includingthe deathpenalty, is in

place (Wildlife Conservation Law, 1988), but

enforcement s insufficient(Zhu 1989, Schaller 1993).

Poaching is a critical concern because decreased

survivorshipof adults, especially females, in a specieswith a low reproductive ate will have a dramaticeffect

on populationpersistence(Shaffer 1983).

Captive populations have not produced much

information for PVA. They are not self-sustaining.Few captiveadults,especiallymales, arereproductively

active, and neonataland adult survival is poor (Schalleret al. 1985). There is much room for researchhere.

Genetic profiles are necessary to determine current

levels of heterozygosity,and to directmating attempts.At least as important s the need to experiment with

husbandryalternativesratherthan the currentpracticeof housing pandas ndividually. Pandashousedin pairs,

notablyin zoos outsideChina,have had relatively high

matingsuccess (O'Brienand Knight 1987), suggestingthatfamiliaritymay facilitatemating. Aggregationsof

mating males around oestrus females in the wild

(Schalleret al. 1985) probablyprovidenecessarysocial

stimulationand learningof mating performance.The project lacks data on variance in rate of

populationgrowth(Vr) and on dispersaland gene flow

betweenpopulations.

Life-HistoryTheoryGeneral.-When direct measurements of

demographic parameters are too difficult or time-

consuming, researcherscan shortcut a complete PVA

and use other theoreticalapproachesto investigatethe

sensitivity of those parametersto ongoing ecological

processes. Life-historytheory, the integrationof age-andsex-specificbirth anddeathrates, is one theoretical

approach. Currentsynthesis presents life histories as

compromisesbetween 3 types of selection:competition

(traditionallyK-selection), disturbance(traditionallyr-

selection), and adversity or stress (Southwood 1988,Partridge and Harvey 1988). Southwood (1988)

categorizes hosecompromisesas differential nvestment

between 5 categories of survival tactics: food

harvesting, reproductive activities, defense to avoid

death, escape in space and time, and physiologicaladaptationso inclementphysicalconditions.

This is a comprehensivescheme directing a searchfor adaptations ffectinglife historyand, consequently,PVA parameters.Researchers houldenquirehow eachof the 5 survival tactics is expressed in the studyanimal'sanatomy,physiology, or behavior. Specialized



28 Int. Conf.BearRes. andManage.9(1) 1994

adaptationsenerally uggest strongselection,and arelativenflexibilityo changes n thoseenvironmentalconditions that select for optimal expression or

functioning f the trait(s). Key adaptationsre bestidentifiedby empirical tudyof the organism.Strong

clues also come fromallometricomparisons,ndbyinference rom taxonomicallyelatedspecies (Peters1983,Belovsky1987).

PandaProject.-In this section use the 5 survival

tacticsSouthwood988)asaframeworkordiscussing

progressn ecological esearchn pandas nd llustratehow conservationecommendationsesult.

Thepanda's bviousanatomicaldaptationsnhance

food harvestingradial esamoid nlargement,obust

molars [Davis 1964]), but its lack of obvious

adaptationso enhance igestion f a fibrous ood with

resultingow

digestiveefficiency(Dierenfeldt al.

1982), led Schaller t al. (1985)to lookcloselyat the

many behavioraltactics for survival on such a

specialized ietof bamboo. Patternsf dietselection,

movement, and activity all appearedadaptedto

maximizeenergy intake rates. Tactics for food

harvesting ominatedhesuiteof adaptationsxhibited

by pandas.However, t is notyetclearwhether andasthathaveaccess o moreof thepreferredrowthorms,nutritionalevels, anddispersion atterns f bamboos,alsohaveimproved eproductiveateor survivorship.Suchresponseso a relaxationf thelimiting ffectsof

foodareassumed ndarelikely,butideally houldbeinvestigatedmore closely. Opportunitiesor such

investigationxistintheMinshan tpresent, nd n the

Qionglaishann 5 to 10 years, as bambooclones,

regenerating fter a dieback,produceculms large

enough for panda foraging. The resultingrapidincreasen thestandingropof foodshouldmake ood

acquisitionasier orpandas.Mostbiological esearchnthepanda rojecturther

investigatedomeaspectof thepanda'sactics orfood

harvesting,nd heprocesses ffectingoodavailability.Pandasselectedtheir feedingsites on relatively lat

terrainwherethe bamboogrowthprovided he mostefficientlyngestible iomass ReidandHu 1991). In

theWolong tudyarea, hebestgrowing onditionsor

bamboopreferred y pandaswere undera selectivelycut forest, rather han a clearcutor uncut(climax)forest(Reidet al. 1991c). A mature orestcanopysimilaroclimax anregeneratenselectivelyutstands

butwill veryrarelydo so in clearcutsTaylor ndQin

1988a,b). Thesestudies ndicatehatselective uttingis the best timberharvestregimefor pandahabitat

retentionn this bambooecosystem Taylorand Qin

1989a).

The extremely poor regenerationof Bashania

fangianabamboo rom seed in clearcuts omparedomature orest(TaylorandQin 1988c; 1993a) meansthat managerscannot rely on seral processes to

regenerateabitatn clearcuts nd willhaveto actively

restore these areas by first replanting anopy treeseedlingsand aterbamboo. Thebestopportunityor

plantingtree seedlings comes soon after bamboo

dieback,whencompetitionrom heslowlyregeneratingbamboos minimalTaylorandQin 1989b,Tayloret

al. 1991a).Schalleret al. (1985) summarizedhe biology of

panda reproduction,he second life-historytactic,

stressingthe altricial nature of the neonate, the

specializednatureof maternity ens, the late age at

reproductivematuration,nd the low litter size. A

specieswith such tacticscan

only persistwith

highsurvivorshipnda relatively ncompromisedirth ate.

These facts emphasizeagain that managersmust

eliminatepoaching,maintain vermatureonifersfor

maternity ens (Schaller t al. 1985, TaylorandQin

1989a,MoFandWWF1989),and nvestigatehehighinfantmortalityate.

Pandasdo not havespecialized efensetactics thethird ife-historyopic) o avoidpredation,xcept heir

ability o climbtrees. Largeclearcuts, houghoften

supportingmuch bamboo, do not provide escape

opportunities, akinghemsuboptimalabitats.

Pandas pparentlyo behave o as to escapeadverseconditions.Theytendto moveto lower elevationso

avoid wintersnow and cold (Schalleret al. 1985).Bamboolowering nddieback,houghnfrequent,an

alsoproduce dverse onditions.Following loweringin the Qionglaishann 1983, about 80% of the standing

crop of Bashaniafangiana, the panda'sprimaryfood,

died (Reid et al. 1989). Pandas responded by

increasinghomeranges,decreasingdiet selectivity,and,

most important, shifting to an adjacent, unflowered

bamboo food resource (Johnsonet al. 1988a, Reid et

al. 1989). The option of feeding on a second or third

bamboowhenone flowers s essential or populationpersistence. A number of panda populationsdo not

now have such choice. Their persistence depends on

habitatrestoration MoF and WWF 1989).

Pandas do not have any obvious physiological

adaptations o inclementphysical conditions, the fifth

life-history tactic, though they currently occupy the

margins f theirhistoricalange, ivingataltitudeswith

a marked seasonal climate, and in mountains with

rugged topography. They do not hibernate,and their

energybudgetsshow only slight surplusof intakeover

costs (Schaller et al. 1985). They may compensate



EVALUATIONOFGIANTPANDACONSERVATION Reid 29

behaviorally. For example, they select feedingsites on

relatively level ground (Reid and Hu 1991). It isunclearwhethersurvivorshipor reproductive utputare

compromised by the energetic costs of surviving incurrent habitats, compared to other historical range.

The timing of adult mortalities and the causes of

juvenile mortality might provide clues. Given the

uncertainty and the lack of alternative range,

managementmust focus on making remaining pandahabitatsas suitable as possible for the animals.

THE ROLE OF BIOLOGICALRESEARCH

Sufficiency of Biological ResearchThe previous review of factorsaffecting,or inferred

to affect,demographic parameters

demonstratesthathuman activity profoundly impacts panda populationpersistence. The most significantresultsare increased

mortalityfrom poaching, reducedeffective populationsize and rates of gene flow through habitat

fragmentation, and decreased habitat quality (withinferred decreases in birth rate and density, andincreases in death rate) through timber harvesting.Evenwhen a bamboodiebacksubstantiallyeducesfood

availability, the extent to which humans have

manipulatedforests containing alternativebamboos isthe key to understandingwhether pandas will starve

(Reid et al. 1989).Biological research is most often insufficient to

implementa conservationgoal. Biological knowledgemustbe integratedwith knowledgeof humanbehaviorsto find suitableand successfulways of changinghumanbehavior. The managementplansimplicitly dentifytheneed for intervention in at least 5 sets of humaninstitutions (Table 1). However, the plans do notcontain sufficient information for successful

implementationof most of the proposed interventions

(Table 1). In particular he planslack depthin 2 areas:

(1) detailedeconomic analysesof proposedchanges in

timber harvesting regimes and resettlementschemes;and (2) understanding of the cultural traditions,community organization, and economic status of

agriculturalistsliving in or around reserves. Theeconomic analysis is essential to evaluatewhetherandhow a shift in timber harvest from clearcuttingtoselectivecuttingin pandahabitatoutsidereservescanbe

accomplished. Knowledge of local communityorganization and traditions is essential to makeresettlement chemesworkpoliticallyandeconomically,to distributebenefitsfrom ecotourism o local people asthey see fit, and to protectcommonpropertyresources

such as wildlife with local interest, consent,and involvement.

Most local people living in and near reserves are

poor subsistence farmers and members of ethnic

minority ribes,whereasmostreserveadministratorsre

Han Chinese from an urban background. There is

frequentlyantagonismbetween the 2 groups. Localsoften destroy habitat to acquire necessary wood forbasic energy needs. They poach wildlife, including

pandas,to supplementa subsistence iving throughsaleof animal pelts, organs, and glands. Very basiceconomicneeds drive theirbehaviors, which are at thesame time the key threats to panda persistence.

Managers frequently act weakly in enforcingregulations,even in reserves, because of the difficulteconomic conditionsof locals (Schaller 1993). Giventhe obvious

need for strong reserve protection andchanges in harmfulbehaviors, the pandaproject mustdirect strongerpatrolling and enforcement, and must

develop sustainable alternative energy and incomeschemesfor locals.

Economic Development.-Integrating social andeconomic development of local people withconservationactivitieshas become a central theme in

protectedareamanagement McNeely 1989, McNeelyand Miller 1984). The hurdles are diverse. Projectsmust includepersonnelwho understand he lessons of

previous internationalrural development experiences.

Biologistsrarelyhave such training. Unfortunately hepandaprojectdid not includesocial scientists to addresseconomic and anthropological ssues. Projects must

explicitly outline the linkages between betterment oflocal social andeconomicconditionsandthe changes inharmfulbehaviorsof local people that are assumedtofollow. These linkagesmustbe strongfor conservationbenefitsto result(Wells and Brandon1992).

The managementplans identify opportunities forecotourismwithprofits stayinglocally, and the need tomake reserve patrolling a financially rewarding andvaluedjob. The contributoryvalues (Norton 1989) of

pandaconservation o other ecosystem processes, suchas forest conservation, watershed stabilization, andmaintenance f diversefaunaland floralgene pools formedicinesand food, all need to be explored in greaterdetail. Removal of mountain forests, the panda'shabitats,has increasedsoil erosion and flooding, andmade river flows more erratic with drastic

consequencesfor irrigationand farmingregimes in the

Chengdu Basin (Han 1989). Panda habitat retentionand restoration an contributesubstantially o local and

provincial economies. This must become a central

argument in support ofmanagement

plan



30 Int. Conf.BearRes.andManage.9(1) 1994

implementation, and a central topic of more

detailedinvestigation.Political Will.-Conservation initiatives cannot be

implementedwithout the will of the politicians,decision

makers, and local people. Yet policy makers,

politicians,andmanagersoften have to makepoliticallydifficult decisions. Conservation professionals,

althoughthey are rarely politicaldecisionmakers,must

actively fosterpoliticalwill.

Fostering political will depends on developingboth

a sharedconservationgoal anda commontrustbetween

collaboratinggroups. In the pandaprojectthe major

groups are the Chinese government, the foreign

conservationcommunity,and agricultural ommunities

in and aroundpandahabitat. Pandaconservationstill

lacks a commitmentby all partiesto a sharedgoal of

conservationin the wild. North American,and more

generally Western, cultures have developed a strong

cultural affinity with wild animals in wild habitats.

Such an affinity is rare in currentChinesecultureand

experience. Consequently, Chinese often hold the

attitude hat animalsare able to lead betterlives when

kept in captivityrather hanleft in the wild. The poor

success of captivepandahusbandryproves this wrong.

Froma Westernperspective,conservingpandasonly in

zoos is insufficient. The Westernethicaljustification

for conservation ies in the psychological, spiritual,and

economic values the animals and their habitats, with

associated species, all enhance. Managers mustincreasetheirefforts for the conservationof wild panda

populations.When parties do not immediatelyshare a common

goal, they must derive one through mutuallyagreed

policies. Economic incentive and education are

establishedapproaches,and are recognized in general

termsin the plans (MoF andWWF 1989, MoF 1992).

These approacheshave more chance of success if the

partiesshareculturaland ethicalvalues. Unfortunately

this is not immediatelythe case in China. A social

consensus that wildlife is an open access resource

requiringcollectivegovernmentalprotection romover-

exploitation by private interests has strong roots in

North America in this century (Gilbert and Dodds

1987). By contrast in China, the opening of market

economies for the private exploitation of common

propertyresources,throughdirectharvest,seems to be

drivingthe increased hreatson pandaand otherwildlife

populations(Schaller 1993). There is little legacy of

regulatory control because markets for animal parts

were monopolized by the state prior to the 1980s.

Whereregulationshave recentlybeenputin place(e.g.,

the Wildlife ConservationLaw 1988), thereappears o

be littlelegacy of culturalor ethicalconstraints,such as

traditionalcommunalmanagement(Berkes and Feeny

1990), compelling people to abide by the law (Zhu

1989).Educationcan play a short-termrole by informing

peoplelocallyandregionallyaboutconservationefforts.In the long term it can play a more potent role by

changing people's value systemsregardinganimalsand

ecosystemprocesses. However, educationcan only be

effective when it recognizes existing cultural values.

These still need to be better explored in the panda

context,whereethnicminoritiesare those living closest

to wild habitats.

Educationof locals cannot counteractthe extreme

economic incentivefor pandapoaching. Poacherscan

obtain more than twenty times an annual salary for a

pandapelt,becausethere are

wealthybuyersin

Japan,Taiwan, and Hong Kong (Schaller 1993). Foreign

greed drives poaching. This problem can only be

addressedby betterreservepatrollingandenforcement,

along with effective enforcementof CITES regulations

and developmentof social sanctionsagainst acquiring

pandapelts in the wealthysoutheastAsian nations.

Politicalwill flows also from mutualtrust. Trust

can be fostered at many scales in personal or social

interactions, from the role model a field biologist

provides for colleagues not experienced in such a

lifestyle, to the media attention and international

reputationa whole nation can attractby implementingmeasuresaffectingthe fate of a well-loved animalsuch

as the giant panda. Trust is built on a consistent

commitment on the part of each member of the

intervening group, normally the foreign agency, to

trying to understandthe cultural traditions, attitudes,

and languageof the other groups. It is a questionof

caring and communicatingwell while also trying to

implementchange.

CONCLUSION

ProjectComponentsThe 4 project components-biological research,

population urvey,management lanning,andtraining-

were all necessary, but have been insufficientbecause

panda populations continue to decline. An extra

component, socioeconomic research, was largely

overlooked, but will be essential for successful

implementation f the managementplan.

The Focus of BiologicalResearchInformationgatheredby researchershas fallen short



EVALUATIONFGIANT ANDACONSERVATIONReid 31

of an idealresult,especiallyin quantifyingdemographic

parametersand genetic profiles. The focus of the

research was not incorrect, but constraintsof time,

logistics and politics precludedfollowing all avenues.

There is no obvious block of research hat shouldhave

been avoided in favor of other options. Researchers

from Beijing University are now successfully

quantifyingdemographicparametersor wild pandas n

the Qinlingshan(Lu and Pan 1989).Most of the biological research has been conducted

in only 2 of the 6 mountainsystems, andwith only 5 of

the approximately 33 bamboo species (Yi 1985)

growing in pandarange. This geographical ocus was

necessary, and study areas were wisely chosen,

including areas already impacted by human activity.

However, researchfindings can rarelybe extrapolatedbetween

ecosystems,and some studies will have to be

replicated in other bamboo forest habitats. The

comprehensivework of the populationsurveyteamhas

partly compensated for the lack of data from other

systems.I have used PVA and life-history frameworks to

comprehensively evaluate, post facto, the biologicalresearch. These frameworksoverlap in subjectmatter,but complement each other. Population viability

analysis stresses patterns of data for predictive

modelling in a probabilistic sense (Shaffer 1990),

though no model yet exists to integrate both

demographic and genetic data. The life-historyapproach highlights processes, including long term,naturalselection as evidencedby adaptation,and short

term, ecological and behavioralprocessesaffectingand

reflecting the adaptations. I believe that most

conservation research programs will benefit from

conscious initialandperiodicevaluationusingPVA and

life-historyframeworks.

The Role of Biological ResearchThe pandaprojectemphasizedbiological information

above all other data, and inadequately pursued

acquisitionof socioeconomic and anthropologicaldata.As is evident from this review, and more generally(McNeely 1989), conservation is primarily a social

problem because social forces driving harmfulhumanbehaviors are at the root of virtually all conservationthreats. Now that a managementplan(MoF 1992) is inthe early stages of implementation,the conservationeffort should incorporatesocial science research, and

stronger liaison with local peoples as integralprojectcomponents. To do this, the project should addChineseandforeign social scientistssuchas economistsandanthropologists o the teamof specialistspreviously

dominatedby biologists. Incorporating ocial scientists

in a team approachwill benefitconservationefforts for

many species and ecosystems on all continents.

LITERATURE CITED

BELOVSKY, .E. 1987. Extinctionmodelsandmammalian

persistence. Pages 35-57 in M.E. Soule, ed. Viable

populationsor conservation. CambridgeUniv. Press,

Cambridge.

BERKES,., ANDD. FEENY.1990. Paradigms ost: changing

views of the use of common property resources.

Alternatives 17:48-55.

BUNNELL,F.L., ANDD.E.N. TAIT. 1980. Bears in models

andreality-implicationso management. nt. Conf. Bear

Res. andManage.4:15-23.

CATTON,. 1987. Hua Hua eschews

cuddly reputation.BBCWildl.December1987.

DAVIS, . 1964. Thegiantpanda:a morphologicaltudyof evolutionary echanisms.Fieldiana:Zool.Mem.3:1-

339.

DESAULNIERS,.M., A.K. GOFF,K.J. BETTERIDGE,. E.

ROWELL,NDP.J. FLOOD.1989. Reproductivehormone

concentrationsn faeces duringthe oestrouscycle and

pregnancyn cattle(Bos taurus)and muskoxen Ovibos

moschatus).Can. J. Zool. 67:1148-1154.

DE WULF,R.R., R.E. GOOSSENS,.R. MACKINNON,ND

S.C. Wu. 1988. Remote sensing for wildlife

management: iantpandahabitatmappingromLandsatMSSimages. Geocartont. 1:41-50.

DIERENFELD,E., H. HINTZ,J. ROBERTSON,P. VAN SOEST,

AND0. OFTEDAL.1982. Utilization of bamboo by the

giantpanda. J. Nutr.112:636-641.

GILBERT,.F., ANDD.G. DODDS. 1987. The philosophyandpractice f wildlifemanagement. obertE. KreigerPubl.Co., Malabar,Fla. 279pp.

GILPIN,M.E., ANDM.E. SOULE. 1986. Minimum viable

populations: rocesses f speciesextinction.Pages19-34

in M.E. Soule,ed. Conservationiology: thescienceof

scarcityanddiversity.SinauerAssoc. Inc., Sunderland,

Mass.

GOODMAN,. 1987. Thedemographyf chance xtinction.

Pages11-34in M.E. Soule,ed. Viablepopulationsorconservation.Cambridge niv. Press,Cambridge.

HAN,C. 1989. Recent hanges ntherural nvironmentnChina. J. Appl.Ecol.26:803-812.

JOHNSON,K.G., G.B. SCHALLER,ND J. HU. 1988a.

Responses of giant pandas to a bamboo die-off. Natl.

Geogr. Res. 4:161-177.

___ , , and . 1988b. Comparative behavior of

red and giant pandas in the Wolong Reserve, China. J.

Mammal. 69:552-564.



32 Int. Conf. Bear Res. and Manage. 9(1) 1994

KIRKPATRICK,.F., S.E. SHIDELER,NDJ.W.TURNER.990.

Pregnancydetermination n uncaptured eral horses based

on steroid metabolites in urine-soaked snow and free

steroids in feces. Can. J. Zool. 68:2576-2579.

LANDE,R., ANDG.F. BARROWCLOUGH.987. Effective

population size, genetic variation, and their use in

populationmanagement. Pages 87-123 in M.E. Soule, ed.

Viable populations for conservation. Cambridge Univ.

Press, Cambridge.

Lu, Z., ANDW. PAN. 1989. The giant pandasof Qinling:

distribution,populationestimate and future. Abstr. Fifth

Int. Theriological Congr., Rome.

MACHLIS,G., ANDJ. S. MARSH. 1988. Protected area

managementand planning in China: a trainingmanual.

Int. Union for the Conserv. of Nat. and Nat. Resour.,

WorldWide Fund, and Minist. of For., People's Republic

of China. 381pp.MCNEELY,.A. 1989. Protected areasand humanecology:

how nationalparkscan contribute o sustainingsocieties of

the twenty-firstcentury. Pages 150-157 in D. Westernand

M.C. Pearl, eds. Conservation for the twenty-first

century. Oxford Univ. Press, Oxford.

AND K.R. MILLER,eds. 1984. National parks,

conservationanddevelopment: the role of protectedareas

in sustaining society. Smithsonian Institution Press,

Washington, D.C.

MENGES, .S. 1990. Population viability analysis for an

endangered plant. Conserv. Biol. 4:52-62.

MESSIER,., D.M. DESAULNIERS,.K. GOFF,R. NAULT,R.

PATENAUDE,NDM. CRETE.1990. Cariboupregnancy

diagnosis from immunoreactiveprogestins and estrogens

excreted in feces. J. Wildl. Manage. 54:279-283.

MINISTRYFFORESTRY. 992. Plan for the protectionof

the giant panda and its habitat. China Wildl. Conserv.

Assoc., Beijing. In Chinese with English summary.

MINISTRYOF FORESTRY ND WORLDWIDEFUND FOR

NATURE. 1989. Nationalconservationmanagementplan

for the giant panda and its habitat. Minist. of For.,

Beijing, People's Republic of China, and WWF, Gland,

Switzerland. 157pp.

MURPHY, .D., K.E. FREAS,ANDS.B. WEISS. 1990. An

environment-metapopulation approach to population

viability analysis for a threatened nvertebrate. Conserv.

Biol. 4:41-51.

NORTON,.G. 1989. The culturalapproach o conservation

biology. Pages 241-246 in D. Western and M.C. Pearl,

eds. Conservation for the twenty-firstcentury. Oxford

Univ. Press, Oxford.

O'BRIEN, .J., ANDJ.A. KNIGHT.1987. The future of the

giant panda. Nature 325:758-759.

PARTRIDGE,., ANDP.H. HARVEY.1988. The ecological

context of life history evolution. Science 241:1449-1455.

PETERS, .H. 1983. The ecological implications of bodysize. CambridgeUniv. Press, Cambridge.329pp.

QIU,M. 1990. Ecological and social implications for the

managementof the giant panda. M.S. Thesis, Univ. of

Idaho, Moscow. 56pp.

RALLS,K., ANDJ.BALLOU.1983. Extinction: lessons from

zoos. Pages 164-184 in C.M. Schonewald-Cox, S.M.

Chambers, B. MacBryde, and W.L. Thomas, eds.

Geneticsandconservation.TheBenjamin/CummingsPubl.

Co., Inc., Menlo Park, Calif.

REID, D.G., ANDJ. HU. 1991. Giant panda selection

between Bashaniafangiana bamboo habitats in Wolong

Reserve, Sichuan, China. J. Appl. Ecol. 28:228-243.

, J. Hu, S. DONG,W. WANG,ANDY. HUANG. 1989.

Giant panda (Ailuropoda melanoleuca) behavior and

carrying capacity following a bamboo die-off. Biol.

Conserv. 49:85-104., M. JIANG,Q. TENG,Z. QIN,ANDJ. HU. 1991a.

Ecology of the Asiatic black bear (Ursus thibetanus) in

Sichuan, China. Mammalia 55:221-237.

_ , J. HU, ANDY. HUANG. 1991b. Ecology of the red

panda(Ailurusfulgens) in the Wolong Reserve, China. J.

Zool., London. 225:347-364.

, A.H. TAYLOR,J. HU, AND Z. QIN. 1991c.

Environmental nfluences on bamboo(Bashaniafangiana)

growthand implicationsfor giant panda

conservation. J. Appl. Ecol. 28:855-868.

REPETTO,. 1988. The forest for the trees? Government

policies and the misuse of forest resources. WorldResources Institute, Washington, D.C.

SCHALLER,.B. 1990. Saving China's wildlife. Int. Wildl.

20:30-40

1993. The last panda. Univ. of Chicago Press,

Chicago, Ill. 291pp.

, J. HU, W. PAN,AND ZHU. 1985. The giantpandas

of Wolong. Univ. of Chicago Press, Chicago, Ill. 298pp.

, Q. TENG,K.G. JOHNSON,. WANG,H. SHEN,ANDJ.

Hu. 1989. The feeding ecology of giant pandas and

Asiatic black bears in the Tangjiahe Reserve, China.

Pages212-241 inJ.L. Gittleman,ed. Carnivorebehavior,

ecology and evolution. Cornell Univ. Press, Ithaca, N.Y.

SERVHEEN,. 1990. The status and conservation of the

bears of the world. Eighth Int. Conf. Bear Res. and

Manage., Monogr. Ser. No. 2. 32pp.

SHAFFER, .L. 1981. Minimumviable populationsizes for

species conservation. BioScience 31:131-134.

1983. Determiningminimumviable populationsizes

for the grizzly bear. Int. Conf. Bear Res. and Manage.

5:133-139.

1990. Population viability analysis. Conserv. Biol.

4:39-40.

SOUTHWOOD,.R.E. 1988. Tactics, strategies, and



EVALUATIONFGIANT ANDACONSERVATIONReid 33

templets. Oikos 52:3-18.

TAYLOR, .H., ANDZ. QIN. 1987. Culmdynamicsanddry

matter production of bamboos in the Wolong and

Tangjiahe giant pandareserves, Sichuan, China. J. Appl.

Ecol. 24:419-433.

__ , AND . 1988a. Regenerationpatternsin old-

growth Abies-Betula forests in the Wolong Natural

Reserve, Sichuan, China. J. Ecol. 76:1204-1218.

, AND . 1988b. Tree replacementpatterns in

subalpineAbies-Betula forests, Wolong NaturalReserve,

China. Vegetatio. 78:141-149.

, AND . 1988c. Regeneration from seed of

Sinarundinariafangiana bamboo, in the Wolong giant

panda reserve, Sichuan, China. Am. J. Bot. 75:1065-

1073.

, AND . 1989a. Structure and composition of

selectivelycut and uncut Abies-Betula forest in

WolongNatural Reserve and implicationsfor pandaconservation

in China. Biol. Conserv. 47:83-108.

_ , AND . 1989b. Bambooflowering andpandas.

Nature 341:111.

, AND . 1992. Tree regenerationafter bamboo

dieback inAbies-Betula forests, Wolong NaturalReserve,

China. J. Veg. Sci. 3:253-260.

, AND . 1993a. Bamboo regenerationafter a

bamboo dieback in Abies-Betula forests, Wolong Giant

Panda Reserve, China. Biol. Conserv. 63: 231-234.

__ , AND . 1993b. Structure and dynamics of

bamboos in the Wolong NatureReserve, China. Am. J.

Bot. 80:375-384.

, D.G. REID,Z. QIN,ANDJ. Hu. 1991a. Bamboo

dieback: an opportunity to restore panda habitat.

Environ. Conserv. 18:166-168.

, D.G. REID,Z. QIN, ANDJ. Hu. 1991b. Spatial

patterns nd environmental ssociates of bamboo(Bashania

fangiana Yi) after mass flowering in southwesternChina.

Bull. TorreyBot. Club 118:247-254.

WASSER, S.K., L. ROSLER, AND R.A. STEINER. 1988.

Excretedsteroidsinprimate eces over the menstrualcycle

and pregnancy. Biol. Reprod. 39:862-872.

WEI,F., J. HU, ANDG. HU. 1989. A study of the life table

of wild giant pandas. Acta Theriologica Sinica 9:81. (In

Chinese.)WELLS,M., ANDK. BRANDON.1992. People and parks:

linkingprotectedareamanagementwithlocal communities.

The World Bank, World Wildl. Fund, and U.S. Agency

for Int. Dev. 99pp.

YI, T.P. 1985. Classification and distributionof the food

bamboosof the giantpanda. J. BambooRes. 4:20-45. (In

Chinese.)

ZHU,J. 1989. Natureconservation n China. J. Appl. Ecol.

26:825-833.

focus and role of research in panda conservation - reid

Documents