Estimating Grizzly Bear Density in Relation to Development and Exploitation in Northwest
Alaska
Author(s): Warren B. Ballard, Kathryn E. Roney, Lee Anne Ayres, Douglas N. Larsen
Source: Bears: Their Biology and Management, Vol. 8, A Selection of Papers from the Eighth
International Conference on Bear Research and Management, Victoria, British Columbia,
Canada, February 1989 (1990), pp. 405-413
Published by: International Association of Bear Research and Management
Stable URL: http://www.jstor.org/stable/3872944
Accessed: 03/01/2009 20:43
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. 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
http://www.jstor.org/action/showPublisher?publisherCode=iba.
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 organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].
International Association of Bear Research and Management is collaborating with JSTOR to digitize, preserve
and extend access to Bears: Their Biology and Management.
http://www.jstor.org
ESTIMATING
GRIZZLYBEAR DENSITYIN RELATIONTO DEVELOPMENT
AND
IN NORTHWESTALASKA
EXPLOITATION
WARREN B. BALLARD,Alaska Department of Fish and Game, P.O. Box 1148, Nome, AK 99762
KATHRYNE. RONEY, National Park Service, Northwest Alaska Areas, P.O. Box 1029, Kotzebue, AK 99752
LEE ANNE AYRES, National Park Service, Northwest Alaska Areas, P.O. Box 1029, Kotzebue, AK 99752
DOUGLAS N. LARSEN, Alaska Department of Fish and Game, P.O. Box 689, Kotzebue, AK 99752
Abstract: Grizzly bear (Ursus arctos) densities within a 1,862 km2 study area surrounding a lead/zinc mine in northwest Alaska were estimated using mark-recapture
methods during late May and early June 1987. Radio collars were used to mark bears and assess population closure. Density estimates were 1 bear/66 km2 for adults
(>3-years-old) and 1 bear/5 1 km2 for bears of all ages. Some of the biases and problems associated with the mark-recapture method were discussed. Density estimates
were used to estimate population size within and near the bear study area, and this estimate was compared with reported and suspected annual harvests. Estimated annual
harvest rates in recent years ranged from 8 to 16%. Current bear density and population estimates will be compared with estimates obtained after the mine is developed
to assess impacts on the bear population.
Int. Conf. Bear Res. and Manage. 8:405-413
Conservationof brown/grizzlybears(Ursusarctos)in
Alaskadependson the availabilityanduse of assessment
methodsthat allow game managersto monitorstatusof
populationsregularly. Historically,managershave relied on crudeanalysesof harvestdataandmiscellaneous
observationsto assess bearpopulationtrendsandeffects
of harvest.However,thebasisfor use of harveststatistics
for monitoringpopulationstatusis not well documented
and appearsimpreciseandunreliable(Harris1984; Harris and Metzgar 1987a,b). In areas where unreported
harvestsarepotentiallylarge, reportedharvestsmay not
be representativeof harvestmortality,andconsequently,
problemsassociatedwith use of harvestdatafor assessing populationtrends may be insurmountable. Fortunately, bearpopulationsappearhealthyand abundantin
many areas of Alaska. If viable populationsare to be
maintained,appropriatemethods must be developed so
thatmanagerscan accuratelyidentifyandremedypopulation declines.
Increasinghumanpopulationshave significantlyreduced the abundanceanddistributionof grizzly bearsin
NorthAmerica (Cowan 1972). Althoughcurrentabundance and distributionof bears in Alaska is similar to
historical levels, alterationof importanthabitatscould
significantly alter productivityand survival of affected
bearpopulations. Currentunderstandingof grizzly bear
populationdynamicsin relationto humandevelopments
is inadequatefor providingeffective guidelinesfor minimizing andmitigatingimpactsto bearpopulations.This
inadequacyexists because such impacts are often long
term, research is usually of short duration,and many
impactsare relativelyrecent (Peek et al. 1987).
This studywas conceived in responseto variationsin
estimatesof bearabundance,andconcernaboutpotential
adverseimpactsfrom developmentand operationof the
Red Dog Mine in northwestAlaska. This studysoughtto
evaluate effects of human harvestsby comparingbear
density with known reportedharvests, and to provide
baseline dataon beardensity, structure,movementsand
reproductiveparametersbefore large scale mine development. Significantchanges in bear density due to the
Red Dog mine will be assessed later by repeatingthe
studyusingidenticalstudymethods. Backgroundforthis
study was providedby Ballard(1987) and Ballardet al.
(1988). This reportpresentsand discusses use of markrecapturemethodsfor estimatingpre-miningbeardensities and estimatingcurrentminimumharvestrates.
The following individuals deserve recognition for
theirassistanceduringthis study:L. Adams,J. Coady,A.
Eliason, D. James,V. Karmun,R. Kemp, A. Lovaas, S.
Machida,M. McNay, R. Nelson, S. Patten,D. Reed, J.
Rood, F. Sandegren,J. Schoen, M. Shaver,R. Sheldon,
andP. Walters. C. Heplerpreparedfiguresandmaps. S.
Millerprovidedvaluableadvicein use of mark-recapture
methods. Constructivecriticismof this manuscriptwas
providedbyA. Cunning,S. Machida,S. Miller,J. Schoen,
and 3 anonymousreviewers. The study was fundedby
theNationalParkService,the AlaskaDepartmentof Fish
andGame(ADF&G),andseveralFederalAid in Wildlife
RestorationProjects.
STUDY AREA
Dynamics, movements, and habitat use of grizzly
bears were studied during 1986 through 1988 within a
6,700 km2 area (Noatak River Study Area [NRSA])
(Fig. 1). The NRSA was locatedwithinGame Management Unit (GMU) 23, an area of approximately
111,370 km2,hereinreferredto as northwestAlaska.
The Red Dog Mine projectwill consist of an open pit
lead/zincmine locatedon Red Dog Creek 131 km north
406
BEARS-THEIR
BIOLOGY AND MANAGEMENT
Fig. 1. Boundaries of the Noatak River Study Area in northwest Alaska where
grizzly bears were studied from 1986 through 1988.
of Kotzebue,Alaska (Fig. 2). In additionto the mine, the
project will include tailing ponds, a mill, power plant,
workerhousing,a saltwaterport,waterreservoir,over 90
km of gravel road,and several gravel borrowsites (U.S.
EnvironmentalProtectionAgency andU.S. Dep. of Interior 1984). The facilitieswill occupy at least 35 km2.The
projectis expected to last 40 years and longer if 18,000
other mining claims are developed. The site will be
occupiedby 225-250 employees. A transportationcorridor from the mine site to the coast may accommodatea
railroadin futureyears. Improvedaccess will result in
increasedhumanuse and additionalmining exploration
and development.
The NRSA boundarieswere selectedto encompassan
area receiving a moderateamountof bear harvestpressure. Because this area was too large for an intensive
mark-recapture
program(hereinreferredto as a census),
a smaller area was selected based upon movements of
radio-collaredbearsin 1986 andlocationof the mine and
roads(Fig. 2). This smallerareais referredto as the mine
census area (MCA).
The MCA was divided into 10 sample units (SUs)
rangingin size from 161-202 km2andtotalling 1,862km2
(Fig. 2). Naturallandmarkssuch as streamsand ridges
were used as boundariesbetween SUs.
The MCA was characterizedby steep, mountainous
terrain traversed by several major rivers and creeks.
Vegetation types rangedfrom riparianstandsof willow
(Salix spp.), birch (Betula nana andB. glandulosa), and
cottonwood(Populusbalsamifera)alongthe streamsand
rivers,gradingintoclosed tall shrub,low shrub,open low
shrub, tundra, and then bare rock and ice at higher
elevations. Relativelythickstandsof white spruce(Picea
glauca) occurredwithin the southernhalf of SUs 3,4, 8,
and 10. Elevationsrangedfrom 60 m along the southern
boundaryto 1,190 m along the northernboundary. Less
than 5% of the MCA containedareas>915 m elevation.
All of the MCA contained usable bear habitatand was
used in calculationsof density estimates.
The NRSA is characterizedby a polar maritimeclimate along the coast and a continental type climate
inland. Summertemperaturesrangefrom 2 to 32 C and
wintertemperaturesrangeas low as -26 to -47 C. Annual
precipitationaveragesfrom 25 cm along the coast, to 5176 cm in the mountains,with half occurringduringJuly
through September. Snow cover usually occurs from
mid-Octoberto mid-May. Caribou(Rangifertarandus),
moose (Alces alces), and Dall sheep (Ovis dalli) occur
within the study area and serve as carrion or prey for
grizzly bears. Arctic char (Salvelinusalpinus), grayling
(Thymallusarcticus), and chum salmon (Oncorhynchus
keta)all occurwithinthemajorriversandtheirdrainages,
and are an important source of bear food. Salmon
migrationusually occurs from July throughSeptember
each year.
Fig. 2. Boundaries of sample units used to census grizzly bears in relation to the
Red Dog Mine in northwest Alaska from 29 May through 4 June 1987.
BEARDENSITY* Ballard et al.
GRIZZLY
METHODS
Bearswere capturedfor markingusing standardhelicopter immobilizationprocedures(Sprakeret al. 1981,
Ballardet al. 1982, ReynoldsandHechtel 1985, Milleret
al. 1987). Bears were immobilized with a mixture of
tiletaminehydrochlorideand zolazepam hydrochloride
(Zoletil 100, Wildlife Laboratories,P.O. Box 8938, Fort
Collins, Colo. 80525), which was deliveredfromeithera
projectiledartor by hand injection(Tayloret al. 1989).
Each capturedbear was sexed, weighed, measured,and
individuallymarkedwith 1-3 lip tattoos,ear-tagged,and
radio-collaredifjudgedto be >5-years-old.Threesubadult
(3.5-4.5-year-olds)bearswere radio-collaredduringthe
census withcollarsdesignedto fall off within 1 year. Premolarswere extractedfromeach immobilizedbear>1.0year-old. Extracted teeth were aged using methods
describedby Goodwin and Ballard(1985).
Methodsused for calculatingmark-recapture
density
estimateswereidenticalto thosedescribedby Milleret al.
(1987). This involved use of mark-recapturemethods
with radio telemetry to correct for populationclosure.
Fixed-wing aircraftwere used to search(withoutaid of
telemetry) individual SUs thoroughly until a bear or
groupof bearswas spotted.Oncesighted,radiotelemetry
was usedto determinewhethertheanimal(s)was marked,
i.e., radio-collared. Sightings of bears with functioning
radiocollars were consideredrecapturesof markedindividuals except for total population estimates, young
accompaniedby their motherswere consideredto have
the same status as their markedor unmarkedmothers.
Adultbearsthatdid not possess functioningradiocollars
wereconsideredunmarked.If unmarked,the animalwas
markedandavailableasa recaptureinsubsequentsearches.
All observed unmarkedadults were captured,with the
exception of 1 adultfemale accompaniedby 1 yearling.
The census occurred during the breeding season, and
consequentlyadultswere sometimesobservedtogether.
These sightings were treated as independentobservations. Because sexualmaturityvariesin each population
we also providedestimates of adult bears >2- and >3years-old, respectively, so that density estimates from
otherstudiescan be compared.Equationsforcalculating
populationsize, density,andassociatedconfidenceintervals wereprovidedby Milleret al. (1987). The bear-days
estimatorwas usedratherthanstandardLincoln-Petersen
estimates.
Twentypeopleparticipatedin thecensusfrom29 May
through 4 June 1987. Six fixed-wing aircraft and 1
helicopter(Bell Jet Ranger206B) were used duringthe
census. Fixed-wing aircraftincluded: 3 PA-18's, 1 PA12, 1ArcticTern,and 1Cessna 185. TheCessnawas used
407
each day for radio-trackingto determinedegreeof population closure (numberand identificationof individual
radio-collaredbearsthatwere eitherin or out of individual SUs). For2 days it was also used for surveying. Both
times, population closure was assessed after assigned
SUs were searched. During other days, radio-tracking
occurredsimultaneouslywith surveys. Trackingaircraft
also maintainedvisual contactwhen surveyaircraftspotted unmarkedbearsthatneededto be capturedandradiocollared. The remainingfixed-wing aircraftwere used
exclusively for surveys.
Surveyaircraftpilot-observerteamsandassignedSUs
were rotateddaily. Pilot-observerteams were directed
not to discuss the locationof sightedbearsduringor after
the census so that searchefforts would not be biased in
succeedingdays.
RESULTS AND DISCUSSION
Population Estimates and Density
During 29 May through4 June 1987, 6 fixed-wing
aircraftflew 198 hourssearchingfor grizzlybearswithin
the MCA. Search effort averaged 0.9 min/km2/day.
Searcheffort per SU variedfrom 0.8 min/km2/dayfor a
SU characterizedby relativelyflat terrainandlow elevational relief where sightability should have been optimum(SU 2), to 1.1 min/km2/dayfor a rugged,mountainous area in the north (SU 9) where observabilitywas
difficult. In retrospect,a larger area could have been
surveyed by reducing search effort or having tracking
aircraftparticipateearlierin the survey. Average search
effort per airplane was 5.6 hours/day, not including
commutetimeorassistingduringimmobilization.Search
efforts were not longer because of fatigue. Consensus
amongsurveyorsindicatedthis was the maximumeffort
that should be attemptedwith 6 aircraft.
Twelveradio-collared
grizzlies(8 femalesand4 males),
which had been capturedand radio-collaredin 1986,
wereavailableas marksduringthecensusin 1987. Home
rangesof the 12 bearsoverlappedthe MCA, and 7 bears
dennedwithinits boundaries.Threemales and6 females
marked in 1986 were resighted at least once during
surveydays 2 through7. No markedbearswereobserved
duringthe first day of the census.
Five adults originally capturedin 1986 were recapturedto replaceradiocollarsbeforeor duringthe census,
and 7 adults were radio-collaredoutside but near the
peripheryof the census area in an effort to increase
potentialmarks.An additional6 adultmalesand 12 adult
females previously unmarkedwere capturedand radiocollaredwithinthe MCA as partof the surveyeffort. Of
408
BEARS-THEIR
BIOLOGY AND MANAGEMENT
the 12 adultfemales, 8 were unaccompaniedby young, 1
was accompaniedby 3 cubs-of-the-year(COY), 2 were
accompaniedby 3 yearlings,and 1 was accompaniedby
3 2.5-year-olds. Yearlingsand COY composed 31% of
the populationin 1987 (Table 1). Ratio of adult (>5years-old)males to females was 61/100.
One of the key assumptionsin mark-recaptureestimatesis thatall individualshaveanequalchanceof being
captured(Otis et al. 1978, White et al. 1982). This
assumptionmayhave beenviolatedin thisstudy. Several
studies have suspected differences in sightability between sows with COY and otherage-sex classifications
(Sprakeret al. 1981, Miller andBallard1982, Ballardet
al. 1982, Miller et al. 1987). Although we did not
statistically test differences in sightability (numberof
times seen divided by numberof times within the area)
amongsex andage classes becauseof smallsamplesizes,
thereappearedto be a sightabilitybias againstsows with
COY. Two radio-collaredsows with COY were within
the census areaon 11 of 14 possible beardays but were
only observedtwice (Table2), thusprovidingsupportfor
the hypothesis of low sightabilityfor sows with COY.
Sightability among other groups was similar ranging
from 29%for females accompaniedby young (>1-yearold) to 34% for single females. Sightabilityof all bears
was 31%. Males that had been captured and radiocollaredbeforethe census had an averagesightabilityof
29% whereas those capturedduringthe census had an
average sightability of 38%. Single females captured
before the census had an average sightability of 40%
whereasthose capturedduringthe census averaged24%.
A preliminaryanalysisof sightabilityin severalAlaskan study areas indicatedthatthere were no significant
differences (P > 0.05) in capturesightabilityof marked
bears by family and age class, or area (Becker 1988).
Therewereno differencesin capturehomogeneityby day
or area for this study (P = 0.316) or among 4 Alaskan
estimates
studyareas(P = 0.449) where mark-recapture
Table 1. Sex and age structureof grizzlybears in and adjacentto the Red Dog
Minecensus area in northwestAlaskaduring29 Maythrough4 June 1987.
Age
Males
Females
Unknown
Total
%
0.5
2
5
9
13
12.7
2
2
11
1
2
1.5
4.2
3
6
3
9
12.7
2.5
3.5 - 4.5
5.5- 10.5
18.3
11
16
27
38.0
>11.0
3
7
10
14.1
Totals
19
37
71
100.0
15
Table 2. Sightability of radio-collared grizzly bears by age, sex and family class
during a census of the Red Dog Mine study area in northwest Alaska from 29 May
through 4 June 1987.
%
No. days
observed sightability
Sex and family
class
Age
(yrs)
No.
bears
No. days
in area
Single females
<5
>5
5
9
14
2
3
12
3
47
11
14
13
16
2
4
25.0
37.1
34.0
18.2
28.6
19
72
22
30.6
All males
3
7
10
9
28
37
2
10
12
22.2
35.7
32.4
Total bears
29
109
34
31.2
Subtotal
Females w/COY
Females w/young
> -yr-old
All females
Males
>5
>5
<5
>5
3
have beenmade(southcentralAlaska- Milleret al. 1987,
Miller 1990; northwestAlaska - this study; Admiralty
Island - Schoen and Beier 1987; and KarlukLake on
KodiakIsland- Barneset al. 1988). Theseresultssuggest
that bear sightabilityamong areas and sex-age classes
may not be as variableas previouslysuspectedby Miller
and Ballard(1982) and Miller et al. (1987).
Two types of populationestimates were developed
fromthis study: (1) numbersof adultbears>3-years-old
and (2) total numbersof bearsincludingCOY and other
offspring. The formerestimatewas the most statistically
valid because it violated fewer assumptions. The adult
(>3-year-olds)populationestimatewas 28 bearsandthe
total populationestimate was 37. The 80% confidence
interval(CI)of theadultestimatewas25-35 (95%CI = 2338), while the 80% CI of the total estimate was 33-43
(95%CI = 31-46). Density estimateswere 1/66 km2for
adults(80%CI = 53-74) and 1/51 km2(80%CI = 44-57)
for total bears. The adult (>3-years-old) estimate was
close to the total number of individual radio-collared
adult and subadultbears (29) that were known to have
beenpresenton 1or moreoccasionswithintheMCAduring the 7-day searcheffort. The totalestimate(37) using
mark-recapturemethods was slightly lower than the
numberof radio-collaredanduncollaredadults(40) and
young thatwere knownto be withinthe areaon >1 days
duringthe survey. If we correctlyaged 3 2.5-year-olds
based on body size, that accompanied1 adult sow, the
estimatednumbersof adultbears(>2.0-years-old)was 32
with an 80 and95%CI of 29-40 and27-44, respectively.
Similarto otherbearpopulationestimates (Miller et
al. 1987, Reynolds et al. 1987. Schoen and Beier 1987,
Barneset al. 1988, SmithandVan Daele 1988), CIs con-
GRIZZLY BEAR DENSITY * Ballard et al.
vergedas surveyeffortprogressed.Populationestimates
and associatedCIs leveled off by day 6 (Fig. 3).
Because grizzly bear populations have been extirpatedor are threatenedwith extinctionin many areasof
the United States,and Alaskacontainsabout65%of the
continentalpopulation(Peek et al. 1987), particularcare
should be taken to reduce and minimize development
impactson grizzlybearpopulations.Historically,declining trendsin grizzly bearpopulationshave been difficult
to detect because of our inabilityto monitorpopulation
statusaccuratelyandina timelyandcost-effectivemanner.
Typically, by the time a change in status of a bear
populationhas been identified,needed remedialactions
are severe and often ineffective. For these reasons, we
recommendthatthe 80%CI be used to evaluateimpacts
of developmentson grizzlybearpopulations.This would
partiallypreventmaking a Type II errorof falsely concluding thattherehas been no change in the population
(SnedecorandCochran1973) as a resultof development.
Theriskof thisapproachis thatmanagementactionsmay
be takenwhen, in fact, no change has actuallyoccurred.
However, if errors are made in the other direction, a
valuableandformerlyrenewableresourcemay be sacrificed.
A large portionof the expense of conductinga markrecapturestudyon grizzly bearshas been associatedwith
markingnew individualsduringthe census. We comparedthe adult and total bear populationestimates and
respective CIs for this study if no new individualshad
beenradio-collared.Theresultingpopulationestimateof
adults would have been only 2% less than the estimate
obtained by marking new individuals. However, the
Adult Bear Pop. Estimate and 95% CI
Red Dog Mine Census Area
20C
!
m
o
d
z
+
Upper 95% limit
Survey Doy
0
Lower 95% limit
Fig. 3. Daily adult (>3-years-old) grizzly bear population estimates and associated 95% confidence intervals derived from using bear-days estimator during a
census in northwest Alaska during 29 May through 4 June 1987.
409
resultingCI wouldhavebeen muchwiderif no new bears
had been marked(95% CI = -29 to +64% of estimatein
comparison to -17 to +39% of estimate obtained by
additional marking). In contrast, the total population
estimate had no new bears been capturedand marked
would have been 30% largerthanthe estimateobtained
by using new marks. Similarto the adultestimate,the CI
on the totalestimatewouldhave been muchwiderhadno
new bears been capturedand marked(-31 to +67% of
estimate in comparisonto -16 to +26% of the estimate
obtained during this study). We concluded that the
primarybenefitof capturingandmarkingnew bearswas
a reductionin the width of the CIs, and perhaps,a more
accurate total estimate. However, because the total
population estimate violates independence of capture
assumptions(i.e., treatingyoung as same statusas adults
- markedor unmarked),the latterconclusion should be
viewed with caution. Similarresults were reportedby
Miller et al. (1987).
Totaloperationalcost(excludingsalaries)of theNoatak
bearsurveywas $64,713 (U.S.). Nearlyhalf the expense
involvedcapturingandradio-collaring25 adultbears. If
we hadnotbeeninterestedin permanentlymarkingbears,
costs could have been reducedseveral thousanddollars
by exclusively using break-awaycollars or some other
temporarymethod of attachment. Expenses for the
density estimation procedurecould have been higher
withoutthe benefit of a contractfor helicoptercosts and
use of government-ownedor leased aircraft.Withcommercial aircraftat commercial rates, expenses for the
census could have been as high as $108,000.
Otiset al. (1978) andWhiteet al. (1982) list 4 assumptions that must be met for capture-recapture
population
estimation methods to be valid: (1) the populationis
closed, (2) animals do not lose their marks during the
experiment, (3) all marks are correctly noted and recordedat eachtrappingoccasion,and(4) each animalhas
a constant and equal probability of capture on each
trappingoccasion. This also implies that captureand
markingdo not affect the catchabilityof the animal.
In this study these assumptionswere either met or
were sufficientlyreducedto allow confidentuse of markrecapturemethodsfor estimatinggrizzlybearpopulation
size in a relatively small area. Use of radio collars to
monitorwhichbears(bear-daysestimate)werepresentor
absent in the MCA eliminatedor substantiallyreduced
violationsof populationclosure. Assumption2 was met
even when an animallost its markbecause loss of radio
collarscould be detecteddaily. Forexample,duringthis
study 1 bearshed its collar on the 6th day of the census.
This loss of markwas identifiedon the day of occurrence
410
BEARS-THEIR
BIOLOGY AND MANAGEMENT
and the bear was subsequentlytreatedas an unmarked
individual.We believe assumption3 was metin all cases.
The largestpotentialproblemof use of mark-recapturein this studywas potentialviolationof assumption4,
which has hamperedall mark-recapturestudies (Otis et
al. 1978). If Becker's (1988) preliminaryanalyses that
therewereno significantdifferencesin capturesightability amongareasor sex-age classes arevalid andaccurate,
and if substantiatedby future replications, they have
significant ramificationsfor the use of mark-recapture
methodsfor estimatingbearnumbers.Perhapsthe statement by White et al. (1982) thatequal catchabilityis an
unattainableideal in naturalpopulations may require
modification for grizzly bears in certain areas under
specific sets of conditions.
One additionalassumptionnot mentionedabove was
that all observationsare independent. Because this assumptionis violatedwhenunmarkedyoung aregiven the
same statusas theirmothers(markedor unmarked),the
total population estimate, which includes bears of all
ages, mustbe used with caution. Similarproblemscould
also occurduringthe matingseason when a second adult
is sighted because of the first observation. A problem
with includingthese sightings and/orage classes in the
estimateis thatit will inflatethe samplesize andcausethe
varianceof theestimateto be biasedtowardsthe low side,
althoughpoint estimates should be similar (E. Becker,
Alas. Dep. Fish and Game, pers. commun.).
Use of mark-recaptureproceduresin this study was
successfulpartiallybecause>50%of the populationwas
markedandbeardensitieswererelativelyhigh. At lower
bear densities, the method has a numberof biases and
samplesize problemsthatmay be overcomewith further
refinement(Reynoldset al. 1987, Miller 1990).
Recently,Eberhardt(1989) evaluatedthemark-recaptureprocedureused in this studyandconcludedthatuse
of the mean of the daily Petersenestimatesmay be preferableto the "bear-days"estimator. If correct,the bear
populationestimatescalculatedin this studycould actually be larger. Forexample,using Eberhardt'sapproach,
the adult(>3-years-old)populationestimatewould have
been 35 with a 80%CIof 22 to 48 (95%CI= 13-57) while
the total estimate would have increased to 49 ? 20
(80% CI).
Density Comparisons
Our reportedtotal density estimate lies within the
rangeof publisheddensityestimatesfor arcticstudyareas
in North America (Table 3). Reynolds (1982) reported
that for North Slope Alaskan populations, high bear
densities in optimumhabitatapproached1 bear/50km2
Table 3. Comparison of reported grizzly bear densities in arctic areas of North
America.
Area
Density (km2/bear)
NorthernYukon
33-39
NorthernYukon
48
WesternBrooks Range,
AK
NW Alaska
42-44
51(44-57)a
EasternBrooks Range,
AK
83-304
NorthwestTerritories
211-262
Source
Nagy et al. 1983a
Pearson 1976
Reynolds 1984
This study
Quimby 1974
Quimbyand Snarski 1974
Curatoloand Moore 1975
Reynolds 1976
Nagy et al. 1983b
a 80%confidence
interval.
and low density in lower quality habitatswas about 1
bear/207 km2. Most grizzly bear density estimates are
based on the additive total numbersof bears observed
over severalyearsof studyand,consequently,containno
measureof precision and no objective estimate of area
occupiedby theestimatedpopulation.A high proportion
of ourcensus areawas composedof denninghabitatand
may not be representativeof average bear densities in
northwestAlaska. Ninety percent of the marked and
unmarkedbearsobservedduringthe surveyperiodwere
located in the mountainousportions of the study area
(Fig. 2: SUs 5-10). Only 10% of the bears observed
during the surveys were found in the lower elevation,
southernSUs (1-4), and 80%of those observationswere
withinSU 4. Typically,bearsmove out of the mountainous terrainand inhabitlower lying areas as spring and
summerprogress(Ballardet al. 1988). A similardistributionof bears was evident during 1986 when we capturedbearsfor studies of movementand demography.
Duringspring1986,we captured48 bears,31 of which
were subsequentlyradio-collared,in partto aid in defining a census area boundaryfor 1987 and to minimize
potentialcapturebiases for sows with COY. Duringthat
captureeffort,we attemptedto searchall portionsof the
NRSA equally. Thirty-onebears were capturedin the
mountainousportionsof theNRSA and 17,or45 %fewer,
were capturedin the southernhalf. We concludethatour
reportedbeardensityestimatesareprobablyrepresentative of high qualitydenninghabitatin an arcticecosystem.
Assessment of Harvest Impacts
One objectiveof this studywas to resolve conflicting
views overthe statusof grizzlybearsin northwestAlaska.
GRIZZLYBEARDENSITY* Ballard et al.
Some local residents have expressed concerns about
losses of property and potential threats to human life
(Larsen 1988). Some residents also believe bears are
currentlynumerous and more abundantthan observed
historically (Loon and Georgette 1989). Many believe
there are too many bears and would prefer a smaller
population(Loon andGeorgette1989). Because of these
concerns and because grizzly bears are classified as a
subsistence use species (defined by Federal Law as
customaryand traditionaluses by ruralAlaska residents
for directpersonalor family consumptionas food, shelter, fuel, clothing, tools, or transportationand for the
makingand selling of handicraftarticlesfor barter,customary trade, and sharing) in northwest Alaska, some
local residentshave advocatedliberalizinggrizzly bear
huntingseasons and bag limits.
Alaskan hunting regulations currently require that
hide and skull of all grizzly bearsharvestedbe presented
to officials of ADF&G within 30 days of the date of kill
for sealing. Sealing of bear hides and skulls is required
butcompliancein someareas,especiallynorthwestAlaska,
has been low. Annualreportedharvestsof grizzly bears
in northwestAlaska have graduallyincreased over the
years (Fig. 4) rangingfrom 8 in 1962 to a high of 57 in
1979. Since 1979, annualreportedharvestshave ranged
between22-48. Annualreportedharvestswithinthe bear
studyareahave fluctuatedsimilarlyto those of northwest
Alaska but an increasing proportionof the total area
harvesthas come from NRSA (Fig. 5).
Use of grizzly bearsfor food is reportedlywidespread
in portions of northwest Alaska (Loon and Georgette
1989). Based on key respondentinterviews in selected
villages, Loon and Georgette(1989) estimatedthatonly
Annual Harvest of Grizzly Bears
=
GMU 23 (y
+ 1.12X, r = 0.67)
-56.8
60
D
50 -
O
40 -
C,I0
m
[
El
30
?
D
o
0
^^^
z
20 -
0??
O
10 -
I
0
1962
1967
l
I
l.
I
I
l.
I
1972
I
I
I
.
I
1977
.
1982
1987
Fig. 4. Reported annual harvest of grizzly bears within GMU 23 of northwest
Alaska from 1962 through 1987.
411
of GMU 23 Harvest from Study Area
1962-1987
(y = -91.1
+ 1.58X, r = 0.74)
70
1962
1967
1972
1977
1982
1987
Fig. 5. Proportion of reported GMU 23 grizzly bear harvests occurring within the
Noatak River Study Area in northwest Alaska from 1962 through 1987.
14-18%of actualharvestsof grizzly bearsarereportedto
ADF&G. Most of the reportedharvestswereby nonlocal
Alaska residentsandnonresidents(Larsen1988). Compliance with sealing regulationsby guides and nonlocal
residents is thought to be high. Although Loon and
Georgette's (1989) estimates contain no measure of
accuracy or precision, if assumed correct, then actual
annual harvests could be from 103-142% larger than
reported. Use of harveststatisticsfor assessing population statusis at best marginaleven when the sex and age
structureof a high proportionof the kill is known (Harris
1984;HarrisandMetzgar1987a,b). The use of suchdata
when >50% of the harvestis unreportedwould be even
less reliable. Because of these problems,othermethods
were used to evaluatethe statusof the populationandthe
potentialfor allowing higher harvests.
To assess the potentialimpactsof humanharvestson
thisbearpopulation,thebeardensityestimatefromMCA
was extrapolatedto a much larger area, and compared
with known minimumharvests. We estimatedthe total
bearpopulationwithinthe NRSA andadjacentareasthat
encompassed nearly all of the home ranges of radiocollaredbearsfrom this study. Based upon the distribution of bearswithin the study areain 1986 and 1987, we
assumedbeardensitiesin the mountainousportionsof the
NRSA were similarto those in the census area(1/51 km2).
Forthe lower lying southernareas,we assumeddensities
were 50%lower or about 1/101 km2. The latterassumption was baseduponthe distributionof bearsightingsand
capturesin 1986 and 1987. These densities were then
extrapolatedto the study areabased on our stratification
into 1 of 2 densitystrata.We classified 5,947 km2as high
densityhabitatand6,932 km2as low densityhabitat.The
412
BEARS-THEIR
BIOLOGY AND MANAGEMENT
extrapolatedbearpopulationfor the 12,879 km2areawas
188 bearsin 1987.
Minimumreportedannualharvestswithinthe NRSA
from 1959through1987have rangedfrom0 to 23. From
1983 through1987, reportedharvestshave rangedfrom
11 to 23. If estimatedunreportedharvestsfromcommunities withinor adjacentto the NRSA (100%of Noatak,
100%of Kivalina,and25%of Kotzebuekills [Loonand
Georgette1989])wereaddedto knownreportedharvests,
then the estimated annual harvest rates during 1983
through1987 wouldincreaseto 8-16%. These ratesmay
also be low becausesome bearswereknownto havebeen
killedandnotretrievedorreported(unpubl.data)in Loon
andGeorgette's(1989) sample.If we hadusedEberhardt's
(1989) approachfor estimating density, our total bear
populationestimate would have been 248 and harvest
rateswould have rangedfrom 6 to 12%.
Although our harvest rate estimates are admittedly
crude,comparisonwith harvestratesreportedfromelsewhere in NorthAmerica (LeFrancet al. 1987) suggests
that currentharvestsapproachor exceed the maximum
allowableharvest. All of our estimatesare in excess of
theconservativeexploitationratesof 2-4%recommended
fornortherlylatitudesbyReynolds(1976),andSidorowicz
andGilbert(1981). Evenif ourestimatesareonly a rough
approximationof actualharvestrates,they suggest that
hunting seasons and bag limits cannot be liberalized
withoutcausing a reductionin the bearpopulation.
SUMMARY
Despiterealandpotentialproblemsandbiases associmethoddescribed
atedwiththe use of themark-recapture
by Miller et al. (1987) for estimatingbear density, the
methodallows managersto estimatepopulationsize and
density quickly and objectively within relatively small
areas. More importantly,the resulting estimates are
repeatableand include a measure of precision. Other
methodsto datehavereliedlargelyon the experienceand
expertiseof the investigator,have been expensive, time
consuming,andusuallycontainno measureof precision.
By applying density estimates obtainedfrom mark-recaptureprocedures,currentannualharvestratesin relation to humanexploitationcould be assessed.
LITERATURECITED
W.B. 1987. Demographyof Noatakgrizzly bearsin
BALLARD,
relationto humanexploitationand mining development.
Alas. Dep. Fish andGame,Fed. Aid in Wildl. Restoration
Prog.Rep. Proj.W-22-5andW-22-6.Juneau.45pp.
_
1982. Home range,
ANDT.H. SPRAKER.
, S.D. MILLER,
daily movements,andreproductivebiology of brownbear
in southcentralAlaska. Can. Field-Nat. 96(1): 1-5.
, K.E. Roney, D.N. Larsen, and L.A. Ayres. 1988.
of Noatakgrizzlybearsinrelationto human
Demography
exploitationandminingdevelopment.Alas. Dep. Fishand
Game,Fed. Aid in Wildl.RestorationProg.Rep., Proj.W22-5 and W-22-6, Job 4.20. Juneau. 100pp.
BARNES,V.G., JR., R.B. SMITH,ANDL.J. VAN DAELE. 1988.
Density estimates and estimated population of brown
bears on Kodiak and adjacent islands, 1987. Kodiak
brown bear research and habitat maintenancetrust, 15
April 1988. 34pp. (mimeo).
E. 1988. AppendixC. In S.D. Miller. 1989. Impacts
BECKER,
of increasedhuntingpressureon thedensity,structure,and
dynamics of brown bear populationsin Alaska's Game
ManagementUnit 13. Alas. Dep. FishandGame,Fed.Aid
in Wildl.Restoration.FinalRep.,W-22-6,Job4.21. 67pp.
I.M. 1972. The status and conservation of bears
COWAN,
(Ursidae)of the world - 1970. Int. Conf. Bear Res. and
Manage. 2:343-367.
J.A., ANDG.D. MOORE.1975. Home range and
CURATOLO,
populationdynamicsof grizzly bearin the easternBrooks
range,Alaska. Chapter1. In R.D. Jakimchuk,ed. Studies
of large mammalsalong the proposedMackenzieValley
Gas Pipeline Route from Alaska to British Columbia.
Arctic Gas Biol. Rep. Ser. Vol. 32.
L.L.
EBERHARDT,
1989.
Using radiotelemetry for mark-
recapturestudies with edge effects. J. Appl. Ecol. 27:In
press.
PROTECTIONAGENCY AND U.S. DEP. OF INTEENVIRONMENTAL
RIOR. 1984. Impactstatement- Red Dog Mine Project.
U.S.Environ.Protection
Agency.Seattle,Wash.290pp.
GOODWIN, E.A., AND W.B. BALLARD. 1985.
Use of tooth
cementumfor age determinationof graywolves. J. Wildl.
Manage. 49:313-316.
HARRIS,R.B. 1984. Harvestage-structureas an indicatorof
grizzly bear population status. M.S. Thesis, Univ. of
Mont., Missoula. 204pp.
, ANDL.H. METZGAR. 1987a. Harvest age structures as
indicatorsof decline in smallpopulationsof grizzly bears.
Int. Conf. Bear Res. and Manage. 7:109-116.
_
, AND
. 1987b. Estimating harvest rates of bears
fromsex ratiochanges. J. Wildl. Manage. 51(4):802-811.
D.N. 1988. Brown/grizzly survey and inventory
LARSEN,
progressreport,FY 1988. Pages65-70in S.O.Morgan,
ed.. Annual reportof survey-inventoryactivities, Vol.
Bear.Alas.Dep.Fishand
XVIX,PartV. Brown/Grizzly
Game,Fed.AidinWildl.Restoration
Prog.Rep.,Proj.W22-5 and W-22-6, Job 4.0.
LEFRANC,M.N., JR., M.B. Moss, K.A. PATNODE,ANDW.C.
SNUG,III, EDS. 1987. Grizzly bear compendium. The
Natl. Wildl. Fed. Washington,D.C. 540pp.
LOON, H.B., AND S.E. GEORGETTE.1989.
Contemporary
brownbearuse in northwestAlaska. Alas. Dep. Fish and
Game,ANILCAFed.AidFunds.Tech.Pap.163. Div.of
Subsistence,Kotzebue. 48pp.
1990. Detection of differences in brown bear
density and populationcomposition caused by hunting.
MILLER, S.D.
Int.Conf.BearRes.andManage.8:393-404.
, AND W.B. BALLARD. 1982.
Density and biomass
estimatesfor an interiorAlaskanbrownbearpopulation.
Can.Field-Nat.96(4):448-454.
, E.F. BECKER, AND W.B. BALLARD.1987. Black and
brown bear density estimates using modified capture-
GRIZZLY
BEARDENSITY* Ballard et al.
recapturetechniquesin Alaska. Int. Conf. Bear Res. and
Manage. 7:23-35.
NAGY, J.A., R.H. RUSSELL,A.M. PEARSON,M.C. KINGSLEY,
AND B.C. GOSKI.1983a. Ecological studiesof the grizzly
bearin arcticmountains,NorthernYukonTerritory,1972
to 1975. Can. Wildl. Serv. 104pp.
___,
___.
___
____,
AND C.B. LARSEN. 1983b.
A
studyof grizzlybearson thebarrengroundsofTuktoyaktuk
Peninsula and Richards Island, Northwest Territories,
1974 to 1978. Can. Wildl. Serv. 136pp.
G.C. WHITE,ANDD.R. ANDERSON.
OTIS,D.L., K.P. BURNHAM,
413
tion of a grizzly bear population. Alas. Dep. Fish and
Game,Fed. Aid in Wildl. RestorationFinalRep.. Juneau.
, ANDJ.L. HECHTEL.1985. Population structure, repro-
ductivebiology, andmovementpatternsof grizzlybearsin
thenorthcentralAlaskaRange. Alas. Dep. FishandGame,
Fed. Aid in Wildl. RestorationProg. Rep., Proj.W-22-1.
Job4.16R. Juneau. 29pp.
,
,___
ANDD.J. REED. 1987. Population dynamics of
a huntedgrizzlybearpopulationin thenorthcentralAlaska
Range. Alas. Dep. Fish andGame,Fed.Aid in Wildl.RestorationProg. Rep., Proj.W-22-5. Juneau. 55pp.
1978. Statisticalinference from capturedata on closed
animalpopulations. Wildl. Monogr.62. 135pp.
PEARSON,A.M. 1976. Populationcharacteristicsof the arctic
mountaingrizzly bear. Int. Conf. Bear Res. and Manage.
3:247-260.
SCHOEN,J.W., ANDL.R. BEIER. 1987.
PEEK, J.M., M.R. PELTON, H.D. PICTON, J.W. SCHOEN, AND P.
SIDOROWICZ,
G.A., ANDF.F. GILBERT.1981. The management
ZAGER.1987. Grizzly bear conservationand management:a review. Wildl. Soc. Bull 15:160-169.
R. 1974. Grizzlybear. ChapterII. In R.D. Jakimchuk,
QUIMBY,
ed. Mammalstudiesin northeasternAlaskawithemphasis
withinthe CanningRiverdrainage.ArcticGas Biol. Rep.
Ser., Vol. 24.
, AND D.J. SNARSKI. 1974.
A study of fur-bearing
mammalsassociated with gas pipeline routes in Alaska.
Chapter2. In R.D. Jakimchuk,ed. Distributionof moose,
sheep,muskoxenandfurbearingmammalsin northeastern
Alaska. Arctic Gas Biol. Rep. Ser., Vol. 6.
H. 1976. North slope grizzly bear studies. Alas.
REYNOLDS,
Dep. Fish andGame,Fed Aid in Wildl. Restoration.Final
Rep.Proj.W- 17-6and7. Jobs4.8R, 4.1 1R. Juneau.20pp.
__ 1982. Alaska range grizzly bear studies. Alas. Dep.
Fish andGame,Fed. Aid in Wildl.RestorationProg.Rep.,
Proj.W-21-1, Job 4.1R. Juneau. 10pp.
. 1984. Grizzly bear population biology in the western
Brooks Range,Alaska. AppendixA, pages 2-19 in H.V.
Reynolds, and J.L. Hechtel. Structure,status,reproductive biology, movement,distribution,and habitatutiliza-
Brown bear habitat
preferencesandbrownbearlogging and mining relationshipsin southeastAlaska. Alas. Dep. Fish andGame,Fed.
Aid in Wildl. RestorationProg. Rep., Proj. W-22-5, Job
4.17R. Juneau. 48pp.
of grizzly bearsin the Yukon, Canada. Wildl. Soc. Bull.
9(2):125-135.
SMITH,R.B., AND L.J. VAN DAELE. 1988.
Terror Lake
hydroelectricproject,KodiakIsland,Alaska. Alas. Dep.
Fish and Game, Spec. Rep. to Alas. Power Authority.
Anchorage. 182pp.(mimeo).
SNEDECOR, G.W., AND W.G. COCHRAN.
1973.
Statistical
methods. Iowa State Univ. Press, Ames. 593pp.
SPRAKER, T.H., W.B. BALLARD, AND S.D. MILLER. 1981.
Game ManagementUnit 13 brown bear studies. Alas.
Dep. Fish andGame,Fed. Aid in Wildl. RestorationFinal
Rep. Proj. W-17-10, W-17-11, and W-21-1. Job 4.13R/
57pp.
TAYLOR,W.P., JR., H.V. REYNOLDS, and W.B. BALLARD.
1989. Immobilizationof grizzly bears with tiletamine
hydrochlorideand zolazepam hydrochloride. J. Wildl.
Manage. 53:978-981.
WHITE, G.C., D.R. ANDERSON, K.P. BURNHAM, AND D.L. OTIS.
1982. Capture-recapture
and removal methodsfor sampling closed populations. Los Alamos Natl. Lab. LA8787-NERP,UC- I. 235pp.