You are on page 1of 10

Society for Conservation Biology

Natural Die-Offs of Large Mammals: Implications for Conservation

Author(s): Truman P. Young
Source: Conservation Biology, Vol. 8, No. 2 (Jun., 1994), pp. 410-418
Published by: Blackwell Publishing for Society for Conservation Biology
Stable URL:
Accessed: 13/09/2010 16:10

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact

Blackwell Publishing and Society for Conservation Biology are collaborating with JSTOR to digitize, preserve
and extend access to Conservation Biology.
NaturalDie-Offsof Large
Louis Calder Center
Box K
Armonk,NY 10504, U.S.A.

Abstract:The viabilityofpopulations is a central concern Declinaciones naturalesrepentinasy pronunciadasen

of biological conservation.The occurrenceof catastrophic mamiferosgrandes:implicacionespara la conservacion
die-offsmay greatlyreduce the long-termviabilityofpopu-
Resumen: La viabilidad de las poblaciones es una preocu-
lations.Theoreticalextinctionmodels and viabilityanalyses
pacion centralen la biologia de conservacz'n.La ocurrencia
requireinformationon thefrequencyof die-offsand on the
de "die-offs"(declinaciones repentinas y pronunciadas)
distributionof die-offseverities.A reviewof literatureiden-
catastr6ficosreducenen gran medida la viabilidad a largo
tified96 natural die-offsin large mammal populations,
plazo de las poblaciones. Modelos teoricos de extinciony
with a die-offbeing definedas a peak-to-troughdecline in
analisis de viabilidad requierende informaci6nsobre de la
estimatedpopulation numbersof at least 25%. If such die- frecuencia de los "die-off'y sobre la distribucionde la se-
offsare common,population viabilityanalyses that ignore
riedadde los mismos.Una revisionde la literaturaidentifico
themmay be overlyoptimistic.The severitiesof thenatural
96 "die-offs" naturales en poblaciones de mamiferos
die-offsof large mammals presentedhereare not uniformly
grandes,definiendoseal "die-off'como una declinacion de
distributed.Thereis a relativeoverabundance of die-offsin
por lo menos un 25% en el tamafiopoblacional. Si tales
the 70-90% range, and an underabundance of die-offs
"die-offs"son comunes,los analisis de viabilidadpoblacio-
greaterthan90%. This may indicate thepresence of buffers
nal que los ignoranseri'ansumamenteoptimistas.La sever-
against population extinction.The reportedcauses of large
idad de los "die-offs"naturales de mamiferosgrandes aqui
mammal die-offsweresignificantlyrelated to trophiclevel:
presentadosno estd distribuidauniformemente. Existe una
herbivoredie-offswere more oftenattributedto starvation,
sobreabundancia relativa de "die-offs"en el nivel del 70 al
while carnivoredie-offswere more oftenattributedto dis-
90%, y una superabundancia de "die-offs"mayores que
ease. Populations subject to large-scalephenomena such as
90%. Esto ultimo podria indicar la presencia de "buffers"
droughtand severewintersmay not be protectedfrom die-
contra las extincionespoblacionales. Las causas reportadas
offsbypopulation subdivision. On the otherhand,popula-
para los "die-offs"de grandes mamiferosestuvieronsignifi-
tions subject to catastrophicdisease epidemics may be pro-
cativamenterelacionadas con el nivel tr6fico;los "die-offs"
tectedby subdivision,and threatenedby corridorsbetween
de berbivorosse atribuyeronma'sfrecuentemente a la inani-
conservationareas and by translocationefforts.
cion, mientrasque los "die-offs"de carni'vorosse atribuy-
eron mas ferecuentemente a enfermedades.Las poblaciones
sujetas a fenomenos a una escala espacial grande, tales
como sequfas e inviernosseverosno podri'an serprotegidas
de los "die-offs"mediante la subdivisionpoblacional. Por
otro lado, las poblaciones sujetas a epidemias catastr6ficas
de enfermedadespodrian serprotegidas mediante la subdi-
visionypuestas enpeligropormediode corredores entreareas
de conservaciony por medio de esfuerzosde translocacio6n.

and Methods be manipulatedto ensure some (high) probabilityof
long-termpopulationsurvival(Soule 1987). Limitson
A centraltask of biological conservationis estimating
theviabilityofpopulationsare hypothesizedto be set by
theviabilityofpopulations,and how variousfactorsmay variousfactors(Lande 1988; Simberloff 1988), includ-
Paper submittedSeptember4, 1992; revised manuscriptaccepted ing inbreeding,loss of genetic variation(Lande 1988;
June 1, 1993. Simberloff 1988), demographicstochasticity(Richter-

ConservationBiology,Pages 410-418
Volume 8, No. 2, June 1994
Young Natural
ofLargeMammals 411

Table 1. Naturaldie-offs
of largemammals.
Species % Die-off Primary Cause Reference Biome
Gray Kangaroo 55 starvation (drought) Caughley et al. 1985 temperate
Red Kangaroo 59 starvation(drought) Caughleyet al. 1985 temperate
Red Kangaroo >50 starvation(drought) Newsome 1975 temperate
Red Kangaroo 85 starvation(drought) Corbet & Newsome 1987 temperate
Warthog >90 starvation(drought) Scholes 1985 tropicalard
Warthog >90? starvation(drought) Walkeret al. 1987 tropicalarid
Warthog 38 starvation(drought) Walkeret al. 1987 tropicalarid
CollaredPeccary 64 ? Glanz 1982 tropicalforest
Mule Deer 42 starvation(winter) Robinetteet al. 1952 temperate
Moose 50-87 starvation(winter) Mech 1966 temperate
Caribou 89 starvation(winter) Milleret al. 1977 arctic
Caribou 60 starvation Davis 1978 arctic
Caribou 78 starvation Davis 1978 arctic
Caribou 44 starvation Davis 1978 arctic
Caribou 69 starvation Davis 1978 arctic
Caribou 71 starvation(winter) Gates et al. 1986 arctic
Grant'sGazelle 46 starvation(drought) Stewart& Zaphiro 1963 tropicalarid
Thomson'sGazelle 65 predation(competition) Borneret al. 1987 tropicalarid
Springbok >85 starvation(drought) Viljoen & Bothma 1990 tropicalarid
Gemsbok >85 starvation(drought) Viljoen & Bothma 1990 tropicalarid
Impala 75 starvation(drought) Scholes 1985 tropicalarid
Impala 72 starvation(drought) Walkeret al. 1987 tropicalarid
Impala 36 starvation(drought) Walkeret al. 1987 tropicalarid
Impala 87 disease (anthrax) Prins& Weyerhauser1987 tropicalarid
Impala 90 disease Prins& Douglas-Hamilton1990 tropicalarid
Impala 49 predation(starvation) Hirst1969 tropicalarid
Waterbuck 80 starvation(drought) Walkeret al. 1987 tropicalarid
Waterbuck 30 starvation(drought) Walkeret al. 1987 tropicalarid
Waterbuck 93 disease (brown ear tick) Melton 1987 tropicalarid
Reedbuck 90 starvation(multiple) Ferrar& Kerr 1971 tropicalarid
Reedbuck 91 disease Prins& Douglas-Hamilton1990 tropicalarid
Kongoni 81 starvation(drought) Hillman& Hillman1977 tropicalarid
Kongoni 69 starvation(drought) Stewart& Zaphiro 1963 tropicalarid
Wildebeast 88 starvation(drought) Scholes 1985 tropcialarid
Wildebeast 90 starvation(drought) Walkeret al. 1987 tropicalarid
Wildebeast 62 starvation(drought) Walkeret al. 1987 tropicalarid
Wildebeast 32 starvation(drought) Hillman& Hillman1977 tropicalarid
Wildebeast 46 starvation(drought) Stewart& Zaphiro 1963 tropicalarid
Wildebeast 31 predation Hirst1969 tropicalarid
Wildebeast 100 predation(drought) Prins& Douglas-Hamilton1990 tropcialarid
Kudu 50 starvation(drought) Walkeret al. 1987 tropicalarid
Kudu 40 starvation(drought) Walkeret al. 1987 tropicalarid
Kudu 51 starvation Hirst1969 tropicalarid
Eland 50 starvation(drought) Stewart& Zaphiro 1963 tropicalarid
Giraffe 27 starvation Hirst1969 tropicalarid
BighornSheep 76 disease Anonymous1987 temperate
BighornSheep >50 disease (scabies) Lange et al. 1980 temperate
BighornSheep 88 starvation(disease) Uhazyet al. 1973 temperate
BighornSheep >75 ? Stelfox1971 temperate
BighornSheep 85 disease (starvation) Stelfox1971 temperate
BighornSheep 52 ? Stelfox1971 temperate
BighornSheep 85 starvation(winter) Stelfox1971 temperate
Dall Sheep 70-90 starvation(winter) Murie 1944 arctic
Dall Sheep 56 starvation(winter) Walterset al. 1981 arctic
MuskOxen 86 starvation(winter) Milleret at. 1977 arctic
Bison 43 ? Meagher1973 temperate
Bison 63 ?Meagher 1973 temperate
Bison 74 ?Meagher 1973 temperate
AfricanBuffalo 87 starvation(drought) Walkeret al. 1987 tropicalarid
AfricanBuffalo 34 starvation(drought) Walkeret al. 1987 tropicalarid
Pronghorn 62 starvation(winter) Martinka1967 temperate
Burchell'sZebra 90 starvation(drought) Walkeret al. 1987 tropicalarid
Burchell'sZebra 73 starvation(drought) Walkeret al. 1987 tropicalarid

Volume 8, No. 2, June 1994
412 ofLargeMammals
NaturalDie-Offs Young

Table 1. Continued.

Species % Die-offj Primary Causeb Reference Biome

Burchell's Zebra 80 starvation (drought) Scholes 1985 tropical arid
Burchell's Zebra 28 starvation (drought) Stewart & Zaphiro 1963 tropical arid
Burchell's Zebra 50 starvation (drought) Hillman & Hillman 1977 tropical arid
AfricanElephant 29 starvation (drought) Corfield 1973 tropical arid
AfricanElephant 29 disease (drought) Prins & Douglas-Hamilton 1990 tropical arid
Vervet Monkey 72 habitat change (predation) Isbell et al. 1990 tropical arid
Baboon 95 habitat change Altmann et al. 1985 tropical arid
Siamong 67 disease Palombit, personal communication tropical forest
Tamarin 84 habitat change (?) Glanz 1982 tropical forest
Cebus Monkey 85 ? Glanz 1982 tropical forest
Howler Monkey 48 disease (yellow fever?) Collias & Southwick 1952 tropical forest
Langur "massive" disease (viral) Work et al. 1957 tropical forest
Bonnet Macaque "massive" disease (viral) Work et al. 1957 tropical forest
Three-Toed Sloth 76 ? Glanz 1982 tropical forest
Bottlenose Dolphin 50 disease Harwood & Hall 1990 marine
Coyote 50 disease (parvivirus) Pence et al. 1983 temperate
Coyote 87 starvation (hare cycle) Clark 1972 temperate
Wild Dog 84 disease (distemper) Borner et al. 1987 tropical arid
Wild Dog 72 disease (competition) Frame et al. 1979 tropical arid
Wild Dog >63 disease (?) J. Malcolm, personal communication tropical arid
AfricanLion 75 disease (fly outbreak) Fosbrooke 1963 tropical arid
Northern Sea Lion 39 ? Merrick et al. 1987 marine
Crabeater Seal 85 disease (crowding, starvation) Laws & Taylor 1957 marine
Harbour Seal 85 disease (?) Pitcher 1990 marine
Harbour Seal 75 disease (distemper) Bjorge 1991 marine
Caspian Seal 80 disease Lauckner 1985 marine
Fur Seal 62 starvation (El Nino) Trillmich & Limberger 1985 marine
Fur Seal 33 starvation Glynn 1988 marine
Sea Otter 33 ? Lubina & Levin 1988 marine
Sea Otter 68 starvation Kenyon 1969 marine
Sea Otter 68 starvation Kenyon 1969 marine
Sea Otter 52 starvation Kenyon 1969 marine
Coati 59 starvation (?) Kaufmann 1962 tropical forest
a Estimatesof "percentagedie-offt'
were made on samples of various sizes and methodologies.b,"Primary
cause" was takenfromthe original
reports.Parentheticalcauses are secondaryor associated causes.

Dyn & Goel 1972), environmentalvariability(Leigh would thereforebe usefulto examinethe literatureon
1981; Goodman,1987a, 1987b; Dennis et al. 1991), and naturaldie-offs.
"catastrophes."Catastrophescan be definedas local ex- Largepopulationfluctuations have been documented
tinctionsof a metapopulation(Ewens et al. 1987) or as in manypopulationsofinvertebrates (Williamson1972;
rare,severe environmental events(Hanson & Tuckwell Wolda 1982,1992), birds(Stout & Cornwell1976), and
1978) such as drought,disease,or habitatchange.There smallmammals(Williamson1972; Hanski1987). I have
is growingunderstandingthatnongeneticfactors,and limitedthissurveyto largemammalsforseveralreasons.
particularly catastrophicevents,may be more likelyto First,because oftheirecological impact,huntingpoten-
limitthe viabilityof populations than genetic factors tial,and conservationattention,thereare numerousre-
(Lande 1988). peated population estimates of many populations of
Catastrophicdie-offshave the potentialfor putting largemammals.Second,largelong-livedanimalsexhibit
great constraintson minimumviable population size less severe population fluctuationsthan do smaller,
(Hanson & Tuckwell 1978; Mangel & Tier 1993). But shorter-livedspecies (Williamson 1972; Pimm 1991)
how common and how severe are such die-offsin na- and so serve as conservativesamplesofpopulationdie-
ture?Theoreticalmodels requireinformation on thefre- offsin nature.Third,justifiablyor not, large mammals
quency of die-offsand on the distribution of die-offse- receive a disproportionateshare of attentionin conser-
verity(Mangel & Tier 1993). It is also importantthatwe vationactivityand research.
identify the naturalfactorsthatput populationsat occa- I identified96 naturaldie-offsof large mammalsin a
sional risk.Populationsthatare subject to catastrophic surveyof the literature(Table 1). Large mammalsare
disease epidemics are likelyto be manageddifferently definedhere as thosefromall mammalianordersexcept
than populationssubject to catastrophicstarvation.It rodents,shrews,bats, and lagomorphs.A die-offis de-

Volume 8, No. 2, June 1994
Young ofLargeMammals
NaturalDie-Offs 413

finedas a monotonicdrop in populationnumbersthat Table2. Numbersoflargemammal specieswhosedie-offs

occurs between two or among more thantwo popula- attributed
tion surveys.In long-termdata sets, these die-offsare andprimates.
peak-to-trough populationchanges.I have includedonly SuggestedCause
die-offsin which therewas at least a 25% reductionin Habitat
populationsize. Group Starvation Disease Predation Change Unknown
I have not included in the analysisany die-offsthat Herbivores 22 (48) 5 (7) 3 (4) 4 (7)
were clearly related to human activities (over- Carnivores 4 (7) 7 (8) 2 (2)
Primates 4 (4) 1(1) 3 (3) 1(1)
exploitation,habitatdestruction),nor have I included
introducedpopulations.I attemptedto differentiate be- Numbersin parenthesesare thenumbersofpopulations.
tween natural epizootics and diseases introducedby
man. For example,I have not included die-offsassoci-
atedwithrinderpestin AfricanungulatesorPasteurella "Periodicdie-offsof deer on denselypopulated ranges
in bighornsheep, which are almost certainlyrecently became commonplace-some caused by parasitesor
introduceddiseases,but I have includeddie-offs associ- disease,othersby outrightstarvation,
but all manifes-
ated with scabies in bighornsheep and distemperin tations of local over-population"(Longhurs et al.
wild dogs, even thoughthe originof these is not com- 1952:22-23).
Berger (1990) documented extinctionsof isolated Resultsand Discussion
populationsof bighornsheep,witha greaterextinction
rates among smallerpopulations.The large numberof These die-offs were widelydiverseetiologically(Table
local bighornextinctionsstronglyimpliesthatthe glob- 2), ecologically(Table 3), and taxonomically(Table 4).
al populationwas not in equilibrium.Eitherextinction Virtuallyall types of large mammalsare represented.
rateshave recentlyincreaseddramatically comparedto Ungulates and carnivores account for most of the
earliertimes(perhaps throughreductionofpopulation records,but primates,marsupials,edentates,and pro-
sizes below criticallimits),or recolonizationrateshave boscids are also represented.The seven differentmam-
lowered. In either case, die-offsin these populations malian orders did not differsignificantlyin the mean
maybe relatedto humanactivity. severity of their die-offs(F = 1.04, d.f. = 7,p = 0.41).
Even thoughI have includedall ofthe naturaldie-offs Die-offsin the fivedifferent
habitattypesdid not dif-
thathave come to my attention,it is reasonableto as- fersignificantlyin theirmean severity(F = 0.33, d.f.=
sumethatthedie-offs reportedin Table 1 representonly 4, p = 0.86). Semiarid(savanna) regions are numeri-
a small fractionof the severe mammaliandie-offsthat cally stronglyrepresented.There are several possible
have occurredin naturein recentdecades. Marinemam- reasonsforthis.First,these regionsharborlargepopu-
mals are difficultto census unless theycome periodi- lationsoflargemammals.Second,thesehabitatsmaybe
cally to limited terrestrialbreeding and calving sites more prone to lethallydryyears than are more mesic
(Thompson & Harwood 1990). Censusingof forestan- ecosystems.Third,animalsin these open habitatsare
imalsis especiallydifficult (Glanz 1982; Milton 1990). relativelyeasy to count.Tropicalforestsare sometimes
Even in open terrestrialhabitats,detailed population consideredrelativelyconstantenvironments. However,
censuses have onlyrecentlybeen carriedout,and only severalmammalspecies in tropicalforestshave experi-
fora limitednumberof species and sites. enced documenteddie-offs.
Not included here are numerousnonquantitative re- Die-offsattributedto differentcauses did not differ
ports of catastrophicdie-offs(such as Longhurset al. significantlyin their mean severity (F = 1.40, d.f. = 4,
1952; Simmonds1992). For example, Longhurset al. p = 0.24). The most commonlyreportedsources of
(1952) report 180 local die-offsof deer in California mortalitywere disease and starvationdue to droughtor
betweenthe late nineteenthcenturyand the mid twen- severe winter.Althoughthe mean severitiesof die-offs
tiethcentury.Over a thirdof these were attributedto were independentof guild(herbivores,carnivores,pri-
disease and over halfto starvation,
winters.Interestingly, these proportionsare similarto
those of die-offs
attributedto starvationand disease for Table3. Numbersofpopulations
withreported for
all herbivoresin my survey(see Table 2). Examples: different
Habitat Frequency
"The winterof 1889 killed most of the deer and ante- TropcialArid 46
lope" (Longhurset al. 1952:14); Temperate 15
"[In 1921] a severe epizootic nearlywiped out a heavy Arctic 9
population of black-tailsalong the McCloud River" Marine 12
(Longhurs1952:21); Tropicalforest 8

Volume8, No. 2,June1994
414 ofLargeMammals
NaturalDie-Offs Young

Table4. Numbers
different groups(mostly
mammalian orders).
Group # Species # Populations
Marsupialia 2 4
Ungulata 24 62
Artiodactyla 23 57
Perissodactyla 1 5
Carnivora 10 17
Primates 7 7
Proboscidae 1 2 30 40 50 60 70 80 90 100
Edendata 1 1 Percent Die-off
Cetacea 1 1
Figure 1. The distributionof 92 reportednatural die-
offsin large mammal populations (From Table 1).
mates; F = 0.09, d.f. = 2, p = 0.92), the frequencies of The dottedline representsa uniformdistribution.
reportedprobable causes of die-offs were significantly
differentbetween herbivoresand carnivores(Table 2). populationsize mayhave more profoundeffectson ex-
Herbivoredie-offs were more oftenattributedto starva- tinctionprobabilitiesthansuggestedby these data.
tion,and carnivoredie-offswere more oftenattributed The numberof die-offssmallerthan 70% is also less
to disease,when analyzedeitherat the species level (X2 thanthose 70-90%. Ifthisis real,it maybe a threshold
= 7.36, d.f. = i,p < 0.001) or at the populationlevel effect,where populationstend to withstandintermedi-
(X2 = 11.54, d.f. = 1,P < 0.001). ate levels of enviornmental stressbut responddramati-
There are severalpossible reasonsforthisdifference cally to extremeevents.I believed thatit more likely
in the reportedcauses of die-offsof predatorsand her- representsthe relativeunder-reporting of less severe
bivores. First,most of the carnivoresin Table 1 are die-offs.
highlysocial and live in more intimatecontactthando A similarpatternalso occurs amongdie-offs in cyclic
most social herbivores,which may promotethe trans- populations.Figure 2 shows the distributionof carni-
missionof disease. Second, the diets of carnivoresmay vore die-offsfrom the well-known lynx/harecycle
facilitatethe acquisition and transmissionof disease. (Keith 1963; Williamson1972). These data come from
Third,herbivoresas a trophiclevel maybe more often trappingdata and may underestimateminimumnum-
food-limitedthan are carnivores.Fourth,the food re- bers and thereforeoverestimatepopulationreductions.
sources of carnivoresmay be less subject to environ- These data show an abundance of die-offintensitiesin
mentalfluctuations. Herbivores(food forpredators)de- the range of 78-92%. Strongdensitydependence is
cline more slowlythanfood plantsduringa droughtor thoughtto protectthesecyclicpopulationsfromextinc-
over a severewinter.Finally,these differences could be tion.The similarity between the large mammaldie-offs
due to the biases of observersif those studyingcarni- in Figure1 and the carnivorecrashesin Figure2 imply
vores were more likelyto attributedie-offsof mixed that catastrophicdie-offsof large mammalsmay also
etiologyto disease. have a density-dependent component. Those die-offs
Althoughmost studies reportedonly a single likely thatwere due to density-dependent factorscould be the
cause of the die-off,multiplecauses are likely.Over- results of either overcompensation (Grenfell et al.
populationmaymanifestitselfboth in starvationand in 1992) or ofchangesin carryingcapacitydue to external
disease susceptibility(Longhurset al. 1952), and habitat factors(such as droughtor severe winters;Western
loss can increasefoodlimitationand predation(Isbell et 1975; Caughley& Gunn 1993). Unfortunately, it is dif-
al. 1990). Althoughmultiplefactorsmay be common, ficultto demonstratedensitydependence fromtime-
only a minorityof the studies directlyaddressed con- series data of naturalpopulations (Vickery & Nudds
tributing factors(see Table 1). 1991; Caughley& Gunn 1993).
The distributionofdie-offintensitieswas not random. These data do not allow calculationof the frequency
Die-offsof 70-90% were over-abundant,and die-offs of die-offsforindividualpopulations.However, severe
greaterthan 90% were under-abundant (Fig. 1; X2 = droughtsof the kindassociatedwith the recorded die-
19.1, d.f. = 6, p < 0.005; among die-offsmore than offsin easternand southernAfricahave occurred sev-
30%, in 10% intervals).The paucityof naturaldie-offs eral timesthiscentury.Longhurset al. (1952) reported
greaterthan90% mayindicatea demographic"bound- numerousdie-offsof deer in Californiaduringthe late
aryeffect"thattendsto protectpopulationsfromlocal 1800s and early 1900s. Harwood and Hall (1990) con-
extinction.Of course, these populationsare the survi- cluded that"marinemammalpopulationsare occasion-
vors of long-term, large-scaleprocesses thatcan be ex- allysubjectto eventsthatmayremove 50% or more of
pected to have alreadyeliminatedpopulationssuscep- the individuals.In the case of seals in the UnitedKing-
tible to extinction. Human-caused reductions in dom, such eventsseem to occur with a periodicityof

Volume 8, No. 2, June 1994
Young NaturalDie-Offs
ofLargeMammals 415

L = Lynx
F = Fox
C = Coyote
M = Fisher


0% -10% -20% -30% -40% -50% -60% -70% -80% -90% -100%

Figure2. Distributionof cycliccarnivorepopulation crashesfrom the lynx-harecycle.Thesedata come from

trappingand may underestimateminimum numbersand therefore overestimatepopulation reductions.

about 50 years."Similarly, threemajor die-offsof Alas- dict less long-termviabilitythan analysesthatdo not.
kan caribou have occurred in the past hundredyears On the other hand, viabilityanalysesthatincorporate
(Miller et al. 1977; Davis 1978). It appears thatmajor limitsto the severityof die-offswill tend to predict
die-offsare not uncommonamongsome largemammal greaterlong-term viabilitythananalysesthatdo not.
species. More problematicis thatthese die-offs mayoccur so
The data are even less able to demonstratewhether rarelythat they are statisticallyunpredictable.Other
such die-offsoccur regularlyor sporadically.Current stochasticsources of variation("demographic,""envi-
models ofenvironmental variationand catastrophicdie- ronmental")can be modelled if the variationoccurs
offsconsider their effectsto be temporallyrandom over timeperiods thatare shortrelativeto the predic-
(Hanson & Tuckwell 1978; Goodman 1987a, 1987b; tiverangeof the viabilityanalysis.Severe and relatively
Mangel & Tier 1993). It is likelythatif die-offswere unpredictabledie-offsmay limitthe usefulnessof pop-
more regularlyspaced, theywould representless of a ulationviabilityanalysis.
threatto populationviabilitybecause theywould allow While these die-offsmay not oftenthemselvesdrive
populationstimeto recoverbetweendie-offs. However, naturalpopulations to extinction,such demographic
the existence of densitydependence as a cause of die- bottlenecksmayput largepopulationsregularlyat risk
offsdoes not guaranteetheirregularityif carryingca- fromthe factorsthat affectsmall populations (Keith
pacityitselffluctuatesrandomlyand strongly (Caughley 1963; Simberloff 1988; Berger1990), includinggenetic
& Gunn 1993). and demographiceffects;in addition,theyare likelyto
be negativelyaffected by humaninterference. As human
ImplicationsforConservation influencesin natureincrease,majorpopulationcrashes
are likelyto become more threateningto species sur-
The patternsreportedherehave severalimplicationsfor vival,even iftheyare naturalin origin.
biologicalconservation.First,severepopulationcrashes Third,different sources ofmassmortality are likelyto
in populationsof large mammalsare apparentlywide- affectsubdividedpopulationsin different ways.Popula-
spreadand notuncommonin nature.The frequencyand tionsubdivision(metapopulationstructure)and the ex-
severityof such die-offsmayput severe restrictions on istenceand qualityofcorridorsamongpopulationshave
the viabilityof populations of large mammals,which become importantconsiderationsin the study of re-
alreadyrequire large areas per individual.Population serve design (Simberloff1988). Populationssubject to
viabilityanalysesshould incorporatethe kindsofpopu- large-scalephenomenasuch as droughtand severewin-
lationvariationindicatedby these die-offs (see Mangel tersmay not be protectedfromdie-offsby population
& Tier 1993). subdivision(Quinn & Hastings 1987). On the other
Second,theremaybe limitsto the severityof natural hand, populationssubject to catastrophicdisease epi-
die-offsamongextantpopulations.The rarityofextreme demics maybe protectedby subdivision(Simberloff &
die-offsrepresentsa formof densitydependence that Cox 1987) and threatenedby corridorsbetween con-
may help bufferpopulations againstextinction.It ap- servationareas and by translocationefforts. It is impor-
pears thatthefactor(s)thatcause severedie-offs are not tantformanagersto considerthe relativelikelihoodof
effectivein eliminatingindividualswhen population thesedifferent potentialsources of catastrophicdie-offs
sizes are greatlyreduced. Possible explanationsinclude beforemakingdecisionsabouthabitator populationma-
refuges,immigration, thedecreased efficiency ofdisease nipulations.
transmission at low populationdensity,or the recovery The list of die-offsin Table 1 is undoubtedlyincom-
offood resources. plete. It is hoped thatthispreliminary reviewwill serve
Therefore, theseresultswill affectpopulationviability as a basis forfurthersurveywork on catastrophicdie-
analysisin conflicting
ways.Viabilityanalysesthatincor- offsin nature.Uncommoneventsare more likelyto be
porate occasional catastrophicdie-offswill tend to pre- recorded(Weatherhead1986) and understood(Young

Conservation Biology
Volume 8, No. 2, June 1994
416 NaturalDie-Offs
ofLargeMammals Young

& Isbell, 1994) when studies are conducted over the Ewens, W.J.,P.J. Brockwell,J.M. Gan, S. I. and D. Resnick.
long term.The patternsrevealed in thisrevieware fur- 1987. Minimumviable populationsize in the presence of ca-
tastrophes.Pages 59-68 in M. E. Soule, editor.Viable popula-
therevidence forthe value of long-term fieldresearch.
tions for conservation.Cambridge UniversityPress, Cam-
Acknowledgments Ferrar,A.A., and M.A. Kerr. 1971. A populationcrash of the
reedbuckRedunca arundinum (Boddaert) in Kyle National
I thankL. Isbell, T. Caro, C. Southwick,M. Mangel,J. Park,Rhodesia.Arnoldia(Rhodesia) 5(16):1-19.
Ginsberg,M. Peacock, F. Smith,D. W. Macdonald, R.
May,and D. Ehrenfeldforhelp withthe manuscriptand Fosbrooke,H. A. 1963. The stomoxysplague in Ngorongoro,
theideas therein,and the CenterforPopulationBiology 1962. East AfricanWildlifeJournal1: 124-126.
and the Departmentof Botanyof the University of Cal-
iforniaat Davis forlogisticaland intellectualsupport. Frame,L. H., J.R. Malcolm,C. W. Frame,and H. van Lawick.
1979. Social organizationofAfrican
wild dogs (Lycaonpictus)
on theSerengetiPlains,Tanzania(1967-1978). Zeitschriftfur
Cited Tierpsychologie50:225-249.
Altmann,J.,G. Hausfater,and S. A. Altmann.1985. Demogra- Gates,C. C.,J.Adamczewski,and R. Mulders.1986. Population
phy of Amboseli baboons, 1963-1983. AmericanJournalof dynamics,winterecology and social organizationof Coates
Primatology8:113-125. Island caribou.Arctic3:216-222.
Anonymous.1987. Audobon 1987(11):18.
Glanz,W. E. 1982. The terrestrial mammalfaunaofBarroCol-
orado Island:census and long-term changes.Pages 455-468 in
Berger,J. 1990. Persistenceofdifferent-sized
populations-an E. G. Leigh,A. S. Rand,and D. M. Windsor,editors.The ecology
empiricalassessmentof rapid extinctionsin bighornsheep. of a tropicalforest:Seasonal rhythms and long-termchanges.
ConservationBiology41:91-98. SmithsonianInstitutionPress,Washington,D.C.
Bjorge,A. 1991. Statusoftheharbourseal Phoca vitulina L. in
Glynn,P. W. 1988. El Nino-southern oscillation1982-1983.
Norway.Biological Conservation58:229-238.
AnnualReview of Ecology and Systematics19:309-345.
Borner,M.,C. D. Fitzgibbon, M. Borner,T. M. Caro,W. K Lind-
Goodman,D. 1987a. Considerationof stochasticdemography
say, D. A. Collins, and M. E. Holt. 1987. The decline of the
in the designof biologicalreserves.NaturalResource Model-
SerengetiThomson'sgazelle population.Oecologia 73:32-40.
Caughley,G.,G. C. Caughley,and L. Smith.1985. The effectof
droughton kangaroopopulations.JournalofWildlifeManage- Goodman,D. 1987b. The demographyof chance extinction.
ment43:679-685. Pages 11-34 in M. E. Soule, editor.Viablepopulationsforcon-
servation.CambridgeUniversity Press,Cambridge,England.
Caughley,G., and A. Gunn. 1993. Dynamicsof large heribi-
vores in deserts:Kangaroosand caribou. Oikos 67:47-55. Grenfell,B. T., 0. F. Price,S. D. Albon,and T. H. Clutton-Brock.
1992. Overcompensationand population cycles in an ungu-
Clark,F. W. 1972. Influenceof jackrabbitdensityon coyote late. Nature355:823-826.
populationchange.Journalof WildlifeManagement36:343-
356. Hanski,I. 1987. Populationsof small mammalscycle-unless
theydon't.Trendsin Ecology and Evolution2:55-56.
Collias,N.,and C. Southwick.1952. A fieldstudyofpopulation
densityand social organizationin howlingmonkeys.Proceed- Hanson,F. B., and H. C. Tuckwell. 1978. Persistencetimesof
ingsof the AmericanPhilosophicalSociety96:143-156. populationswith large randomfluctuations. TheoreticalPop-
Corbet,L. K, andA. E. Newsome. 1987. The feedingecologyof
the dingo. III. Dietary relationshipswith widely fluctuating Harwood,J.,and A. Hall. 1990. Mass mortality
in marinemam-
preypopulationsin aridAustralia:An hypothesisof alteration mals: Its implicationsforpopulation dynamicsand genetics.
ofpredation.Oecologia 74:215-217. Trendsin Ecologyand Conservation5:254-257.

Corfield,T. F. 1973. Elephantmortality

in Tsavo NationalPark, Hillman,J.C., and A. K Hillman.1977. Mortalityofwildlifein
Kenya.East AfricanWildlifeJournal11:339-368. NairobiNationalPark,duringthe droughtof 1973-1974. East
Davis,J.L. 1978. Historyand currentstatusofAlaskacaribou
herds. Biological Papers of the Universityof Alaska,Special Hirst,S. M. 1969. Predationas a regulatingfactorof wild un-
Report3:1-8. gulatepopulationsin a Transvaallowveldnaturereserve.Zoo-
logica Africana4:199-230.
Dennis,B., P. L. Munholland,and J.M. Scott. 1991. Estimation
of growthand extinctionparametersforendangeredspecies. Isbell,L. A.,D. L. Cheney,and R. M. Seyfarth.
1990. Costs and
Ecological Monographs61:115-143. benefitsof home rangeshiftsamongvervetmonkeys(Cerco-

Conservation Biology
Volume 8, No. 2, June 1994
Young Natural ofLargeMammals 417

pithecusaethiops) in AmboseliNatioalPark,Kenya.Behavior- 1972-74. ReportSeriesNumber40. CanadianWildlifeService,

al Ecologyand Sociobiology27:351-358. Ottawa,Ontario,Canada.

Kaufmann, J.H. 1962. Ecologyand social behaviorofthecoati, patternsof a mammalcom-

Milton,K 1990. Annualmortality
Nasua narica on BarroColorado Island.University ofCalifor- munityin centralPanama.JournalofTropicalEcology6:493-
nia Publicationsin Zoology 60:95-222. 499.
of Wis-
Keith,L. B. 1963. Wildlife'sten-yearcycle. University Murie,A. 1944. The wolves of MountMcKinley.Fauna of the
consin Press,Madison,Wisconsin. NationalParksof the United States.Fauna Series 5, National
Kenyon, K W. 1969. The sea otter in the eastern Pacific
Ocean. NorthAmericanFauna Series #68. U.S. Government Newsome,A. E. 1975. An ecological comparisonof the arid-
PrintingOffice,Washington,D.C. and theiranomalousprosperityin
zone kangaroosofAustralia,
the introductionof ruminantstock to their environment.
Lande,R. 1988. Genetics and demographyin biological con- QuarterlyReview of Biology50:89-425.
servation.Science 241:1455-60.
Pence, D. B., L. A. Winberg,B. C. Pence, and R. Sprowls.1983.
Lange,R. E.,A.V. Sandoval,and W. P. Meleney.1980. Psorotic The epizoologyand pathologyof sarcopticmangein coyotes,
scabies in bighornsheep (Ovis canadensis mexicana) in New Canis latrans, from south Texas. Journal of Parasitology
Mexico. Journalof WildlifeDisease 16:77-82. 69:1100-1115.
Lauckner,G. 1985. Diseases of mammalia:Pinnipedia.Pages
Pimm,S. L. 1991. The balance ofnature.University
683-693 in 0. Kinne,editor.Diseases of marineanimals,vol.
VI,part 2. BiologischeAnstaltHelgoland,Hamburg,Germany.

Laws,R. M.,and R.J.F. Taylor.1957. A massdyingofcrabeater Pitcher,K W. 1990. Major decline in the numberof harbour
seals, Lobodon carcinophagus (Gray). Proceedings of the seal, Phoca vitulina richardsi,on Tugidak Island, Gulf of
Zoological Societyof London 129:315-324. Alaska.MarineMammalScience 6:121-134.

Leigh,E. G. 1981. The average lifetimeof a population in a Prins,H. H. T., and F. W. Weyerhauser.1987. Epidemics in
varyingenvironment. Journalof TheoreticalBiology90:213- populationsofwild ruminants: Anthraxand impala,rinderpest
239. and buffaloin ManyaraNationalPark,Tanzania.Oikos 49:28-
Longhurs.W. M.,A. S. Leopold, and R. F. Dasman. 1952. A sur-
vey of Californiadeer herds. Game Bulletin #6. California Prins,H. H. T., and I. Douglas-Hamilton.1990. Stabilityin a
Departmentof Fish and Game, Sacramento,California. multispeciesassemblage of large herbivoresin East Africa.
Oecologia 83:392-400.
Lubina,J.A.,and S. A. Levin.1988. The spreadof a re-invading
species: Rangeexpansionin the Californiasea otter.American Quinn,J.F., and A. Hastings.1987. Extinctionin subdivided
Naturalist131:526-543. habitats.ConservationBiology1:198-208.

Mangel,M., and C. Tier. 1993. Dynamicsof metapopulations Richter-Dyn,N., and N. S. Goel. 1972. On the extinctionof a
with demographicstochasticityand environmentalcatastro- colonizing species. Theoretical Population Biology 3:404-
phes. TheoreticalPopulationBiology44:1-31. 433.
Martinka,C. J. 1967. Mortalityof northernMontana prong- Robinette,W. L., 0. Julander,
J.S. Gashwiler,and J.G. Smith.
horns in a severe winter.Journalof WildlifeManagement 1952. Wintermortalityon mule deer in Utah in relationto
31:159-164. rangecondition.JournalofWildlifeManagement16:289-299.

Meagher,M. M. 1973. The bison ofYellowstoneNationalPark. Scholes,R.J. 1985. Droughtrelatedgrass,tree and herbivore
ScientificMonographSeries 1. NationalParkService,Washing- mortalityin a southernAfricansavanna.Pages 350-353 inJ.C.
ton,D.C. TothillandJ.J.Mott,editors.Ecologyand managementof the
world's savannas.AustralianAcademy of Science, Canberra,
Mech, L. D. 1966. The wolves of Isle Royale. Fauna of the Australia.
NationalParksof the United States.Fauna Series 7, National
ParkService,Washington,D.C. D. 1988. The contributionofpopulationand com-
munitybiology to conservationscience. Annual Review of
Melton, D. A. 1987. Waterbuck (Kobus ellipsipyrmnus) pop-
Ecologyand Systematics19:473-512.
ulationdynamics:The testingof a hypothesis.AfricanJournal
of Ecology 25:133-145.
D., and J. Cox. 1987. Consequences and costs of
Merrick,R. L.,T. R. Loughlin,and D. C. Calkins.1987. Decline conservationcorridors.ConservationBiology1:63-71.
in abundanceofthenorthernsea lion,Eumetopiasjubatus, in
Alaska1956-86. FisheryBulletin85:51-365. Simmonds,M. 1992. Mass mortalityof mediterraneandol-
Miller,F. L., R. H. Russell,and A. Gunn. 1977. Distributions,
movementsand numbersof Pearycaribou and muskoxenon Soule,M. E.,editor.1987. Viablepopulationsforconservation.
western Queen Elizabeth Islands, Northwest Territories, CambridgeUniversity Press,Cambridge,England.

Conservation Biology
Volume 8, No. 2, June 1994
418 Nahtual
ofLaqe Mammals Young

Stelfox,J.G. 1971. Bighornsheep in the Canadian Rockies:A Walters,C.J.,M. Stocker,and G. C. Haber. 1981. Simulation
history1800-1970. Canadian Field Naturalist85:101-122. and optimizationmodel for a wolf-ungulatesystem.Pages
317-337 in C. W. Fowlerand T. P. Smith,editors.Dynamicsof
Stewart,D. R. M.,and D. R. P. Zaphiro. 1963. Biomassand den- largemammalpopulations.JohnWileyand Sons,New York.
sityofwild herbivoresin different EastAfricanhabitats.Mam-
malia 27:483-496. Weatherhead,PJ. 1986. How unusual are unusual events?
Stout,I.J. and G. W. Cornwell. 1976. Journalof WildlifeMan-
40:681-693. Western,D. 1975. Wateravailabilityand its influenceon the
structureand dynamicsof a savannahlarge mammalcommu-
Thompson,P. M., and J.Harwood. 1990. Methodsforestimat- nity.East AfricanWildlifeJournal13:265-286.
ingthepopulationsize ofcommonseals,Phoca vitulina Jour-
nal ofApplied Ecology 27:924-938. Williamson,M. 1972. The analysisof biological populations.
F.,and D. Limberger.1985. DrasticeffectsofEl Nino
on Galapagospinnipeds.Oecologia 67:19-22. Wolda,H. 1982. Pages 319-330 in E. G. Leigh,A. S. Rand,and
D. M. Windsor,editors.The ecology of a tropicalforest:Sea-
Uhazy,L. S.,J.C. Holmes,andJ.G. Stelfox.1973. Lungwormsin sonal rhythms
and long-term changes.SmithsonianInstitution
the RockyMountainbighornsheep of westernCanada. Cana- Press,Washington,D.C.
dianJournalof Zoology 51:817-824.
Wolda,H. 1992. Trendsin abundanceoftropicalforestinsects.
Vickery,W. L., and T. D. Nudds. 1991. Testing for density- Oecologia 89:47-52.
dependenteffectsin sequentialcensuses. Oecologia 85:419-
423. Work,T. H., H. Trapido,D. P. NarasimhaMurthy,R. Laxmana
Rao, P. N. Bhatt,and K G. Kulkarni.1957. Kyasanurforestdis-
Viljoen,P.J.,and J.du P. Bothma. 1990. Daily movementsof ease. III. A preliminary
reporton the natureof the infection
desert-dwellingelephants in the northernNamib Desert. and clinicalmanifestations in humanbeings.IndianJournalof
SouthAfricanWildlifeResearch20:69-72. Medical Science 11:619-645.

Walker,B. H., R. H. Emslie,R. N. Owen Smith,and R.J.Scholes. Young,T. P., and L.A. Isbell. 1994. Minimumgroup size and
1987. To cull or not to cull: Lessonsfroma southernAfrican otherconservationlessonsexemplifiedby a decliningprimate
drought.Journalof Applied Ecology 24:381-401. population.BiologicalConservation68:129-134.

Conservation Biology
Volume 8, No. 2, June 1994