OF EXTANT
AND RECENTLY
DIETAND MORPHOLOGY
EXTINCT
NORTHERN
BEARS
DAVIDJ. MATTSON, Forest and Rangeland Ecosystem Science Center, Department of Fish and Wildlife Resources, University
of Idaho, Moscow, ID 83844, USA, email: matt7281@ novell.uidaho.edu
Abstract: I examined the relationshipof diets to skull morphology of extant northernbears and used this informationto speculate on diets
of the recently extinct cave (Ursus spelaeus) and short-faced(Arctodussimus) bears. Analyses relied upon published skull measurementsand
food habits of Asiatic (U. thibetanus)and American (U. americanus) black bears, polar bears (U. maritimus), various subspecies of brown
bears (U. arctos), and the giant panda (Ailuropodamelanoleuca). Principal components analysis showed major trends in skull morphology
related to size, crushing force, and snout shape. Giant pandas, short-facedbears, cave bears, and polar bears exhibited extreme features along
these gradients. Diets of brown bears in colder, often non-forestedenvironmentswere distinguishedby large volumes of roots, foliage, and
vertebrates, while diets of the 2 black bear species and brown bears occupying broadleaf forests contained greater volumes of mast and
invertebratesand overlapped considerably. Fractions of fibrous foods in feces (foliage and roots) were strongly related to skull morphology
(R2 = 0.97). Based on this relationship,feces of cave and short-facedbears were predictedto consist almost wholly of foliage, roots, or both.
I hypothesized that cave bears specialized in root grubbing. In contrast, based upon body proportionsand features of the ursid digestive
tract, I hypothesized that skull features associated with crushing force facilitated a carnivorousrather than herbivorousdiet for short-faced
bears.
Ursus 10:479-496
Key words: Ailuropodamelanoleuca,Arctodussimus, black bear, brown bear, cave bear, food habits, giant panda,polar bear, short-facedbear,
skull morphology, Ursus americanus, Ursus arctos, Ursus maritimus,Ursus spelaeus.
An animal's diet can reveal much about its relationship to the physical environmentand otherorganismsin
it. Forexample,Ewer (1973), Eisenberg(1981), and several authorsin Gittleman(1989) describe pervasive associationsamongthe physiology, morphology,behavior,
and diet of numeroustaxa, including carnivores. This
type of researchstronglysuggeststhatdiet shouldclosely
reflectthe constraintsandadaptationsof somaticfeatures
while influencing the evolutionary trajectoryof these
same traits. At the very least, we would expect correlation, subject to judicious interpretationof functionalrelationships (Radinsky 1985).
The common associationsbetween physical form and
diet of extant organismshave led researchersto speculate on the diets and other life habits of extinct species
based on sharedskeletaltraits,especially of the skull and
limbs (e.g., Kurten 1967, 1976; Radinsky 1981a; Van
Valkenburgh1985, 1988, 1989). Given that fossilized
or otherwise preservedbody parts are the only tangible
clues left to inform us about the behavior of vanished
species, establishingrelationshipsbetween, for example,
diet andskull morphologyof extantorganismsis ouronly
means of imbuingthese physical remainswith useful information(Radinsky 1985, Guthrie 1990). Knowledge
aboutthe behaviorof recentlyextinct forms can, in turn,
help us betterunderstandwhy survivingspecies that existed with these vanished animals behave the way that
they do.
Although recent work (e.g., Davis 1949, 1964;
Radinsky 1981a; Gittleman and Harvey 1982; Van
Valkenburgh 1985, 1988, 1989; Gittleman 1986a,b;
McNab 1986) has placed the diet, behavior,and skeletal
morphologyof bearsin context of othercarnivores,little
has been done to examinerelationshipsamongursidsbetween their physical form and foods. There are some
descriptive(e.g, Bromlei 1965, Schalleret al. 1989) and
analytical comparisons (Davis 1964, Stirling and
Derocher 1990), but thereare no quantitativeanalyses of
relationshipsbetweenmorphologicalanddietarypatterns.
Regardless,the diverseursidskullmorphology(Radinsky
198la) holds promiseof being relatedto eithermechanical or evolutionaryforces associatedwith diet.
I used quantitativeand other analyticaltechniquesto
explore relationshipsbetween diets and morphologyof
extant bear species and used this informationto speculate on the diets and relatedbehaviorof otherbears that
went extinct within the last 15,000 years. My analysis
uses existing published informationand is intended to
serve severalpurposes: elucidatemeta-patters, precipitate more informedhypotheses,and stimulatefurtherresearch. I also undertookthis analysiswith relativelylittle
concern for whetherthe apparentrelationshipsbetween
diet and physical form were genetically fixed or merely
the result of mechanicalforces duringdevelopment. In
eithercase, the analysiswouldhavebearingon bothshortand long-termmorphologicaladaptationsas well as diets
of extinct bears.
The InteragencyGrizzlyBearStudyTeamandNational
Biological Service employed me while completing this
paper. I appreciatereviews by S. Stringham,D. Johnson,
K. Elgmork,J. Peek, B. Van Valkenburgh,andan anonymous reviewer, and I especially appreciatethe support
480
Ursus 10:1998
of R.G. Wright. Conversationswith S. Stringham,S.
Herrero,and P. Matheus were also very helpful to my
insights and analysis, along with the years I spent studying Yellowstone's grizzly bears.
METHODS
Dataand DataStructuring
The paucity of informationon bears from tropicaland
subtropicallatitudes(i.e., the sun [Helarctosmalayanus],
sloth [Melursus ursinus], and Andean [Tremarctos
ornatus]bears) led me to focus on extant and recently
extinct bears of temperate, boreal, and arctic regions.
These constitutea logical assemblageof species thathave
to some extent co-existed, if not co-evolved. Of surviving bears,I examinedthe polar,brown,Americanblack,
and Asiatic black bearsand the giantpanda. Of recently
extinct bears, I examinedthe cave and giant short-faced
bears. The skeletalmorphologyof cave and short-faced
bears has been examined in relative detail by Kurten
(1955, 1958, 1967, 1976); more recently the short-faced
bear was studied by Emslie and Czaplewski (1985). I
followed the classification and nomenclature of
Wozencraft (1989) and the corroboratingevidence of
Goldman et al. (1989) in treatingthe giant panda as a
bear. The pandaandpolarbeardefined dietaryand morphological end-pointsfor the analysis.
Datapublishedby otherauthorswereused for all analyses. Commonskull measurementswere availablefor individual bears from differentstudies and areas (sources
listed in App. 1). These measuresare defined in Figure
1, along with a measureof snout or rostrumlength (the
rostralend of the orbit to the rostralend of the skullFL) thatI derivedfrom publishedphotographs. I used a
dentalindex (DENTA) thatwas the summedproductsof
mean length and width for the P4,M1,and M2teeth (total
length of this dental row [DENTL] is illustratedin Fig.
1). All these measures,to some extent,have been associated with eitherdiet or forces relevantto explainingdiet
in other taxa especially of primatesand ungulates(e.g.,
Moss 1968, Kay 1975, Corruccini1980, Preuschoftet al.
1986, Janis and Ehrhardt1988, Spencer 1995).
I averagedall measures,by taxon, based on individual
studymeans, for use in the exploratoryanalyses because
not all skull measureswere available from each source,
sample sizes differeddramaticallyfor each measureand
species, and some results were presentedas means and
othersby individualskull. As a consequence,multivariate analyses were strongly orientedto the equivalentof
taxa centroids. Except for brownbearsthese taxa corre-
spondedto species. Brown bear subgroupswere distinguished on the basis of subspecies from Pacific coastal,
European,and interiorcontinental(North America and
Asia) regions. This stratificationwas consistent with
major morphological dissimilarities among subspecies
identifiedby Ognev (1931) and Hall (1984).
I also includedmeasuresof body size andrelativeskeletal dimensionsfor comparisonamong species. Skeletal
dimensions were expressed as previously defined indices: FMT = femurlength/longestmetatarsallength (Van
Valkenburgh 1985); VHR = humerus + radius length/
length of thoracicvertebrae10, 11, and 12 (Davis 1964);
RH = (radius/humeruslength)x 100 (Davis 1964, Emslie
and Czaplewski 1985); and HW = (humeralwidth across
epicondyles/humerallength) x 100 (Hildebrand1985).
Increasingvalues of HW and decreasingvalues of RH
reflectedpotentialadaptationsto digging, while increasing VHR and decreasingFMT values reflectedpotential
adaptationsto increasedterrestrialmobility.
Informationfrom 85 studies of bear diet representing
all 6 extant species were used for this analysis (see App.
2 for references). Otherstudies reportedfood habits or
describedfeeding behavior,but in ways that I could not
use. Most studiesused in this analysispertainedeitherto
Eurasianbrownbears(n = 21), NorthAmericangrizzlies
(n = 22) or American black bears (n = 32). Diet was
representedas fecal contentfor the periodcorresponding
to late summerand fall hyperphagia.I assumedthatthe
bulk of high qualityfood was consumedannuallyduring
this period (Nelson et al. 1983, Mattson et al. 1991).
Nelson et al. (1983) furtherarguedthat hyperphagiais
the time of year most critical to survival and reproduction of black and brownbears. The diet duringhyperphagiathushas the greatestpotentialrelevanceto explaining
morphologicalpatterns. This period likely varies with
habitatconditions, so I judged the onset and end of hyperphagiaby descriptions,if available,of bear behavior
for each area and the adventof high quality mast in the
diet. By this approach,with the exception of polarbears
and pandas,hyperphagiabegan in July and ended in October or November.
Scat contents(eitherby percentvolume or some index
of percent volume) were averaged for the season,
weighted by sample sizes, and averaged across years.
Each year was weighted equally unless annual sample
sizes were extremelydisparate.A few (n = 14) key studies only provided frequency of diet item occurrence.
These were used (standardizedso as to sum to 100%),
realizing that, in contrastto volumes, frequencies may
inflate contributions, especially of invertebrates (i.e.,
where both frequency and volume had been recorded,
DIETANDMORPHOLOGY
OFNORTHERN
BEARS* Mattson
invertebrateswere consistentlya frequentbut low volume partof scats).
Diets were describedin termsof 6 broadcategories:
foliage, roots, soft mast, hardmast, invertebrates,and
vertebrates.These categoriesreflectedeithermarkedly
differentnutrientcomposition, physical structures,or
bearforagingbehavior,with potentialrelevanceto differences in bear morphology. Hardmast consisted of
fruits and seeds with a hard protective covering (includingboth acornsandpine [Pinusspp.] seeds), while
soft mastconsistedof fleshy fruitsor even strobili(such
as those of Juniperus)thatwere comparablysoft. These
categories thus did not have clear anatomicalor taxonomic distinction, but rather related to features of
greaterpotentialrelevanceto how bearsused them. For
some analyses, these 2 categorieswere pooled simply
as "mast." Foliage and roots were similarlypooled as
fibrous foods, recognizing that they shared the com-
481
mon traitof potentiallyhigh fiber content, althoughone
was procuredby grazing or browsing and the other by
the considerablymore demandingprocess of grubbing.
Invertebrateswere distinguishedby their small size and
typical chitinousexoskeleton,while vertebratesincluded
tissue acquiredby diverse behaviorincluding grubbing,
scavenging, and both terrestrialand aquaticpredation.
The primarystratificationI used for analysisof dietary
differences reflected both bear taxonomy and structure
of the study area vegetation. Food habits studies were
segregatedon the basis of the investigated species and,
for brownandAmericanblackbears,by whetherthe predominantstudy areavegetationwas non-forest(i.e., arctic or alpine), or coniferous, mixed, or pure broadleaf
forest. Determinationof vegetation structurewas based
on studyareadescriptionsandthe ecoregion(Bailey 1989)
within which the study areawas located. I hypothesized
that the availabilityand use of differentbear foods was
Fig. 1. Measurements used in principalcomponents (PC) analysis of ursid skulls.
482
Ursus 10:1998
fundamentallyrelatedto these broaddifferencesin vegetation structure(e.g., hard mast would be more abundant in broadleaf forests, soft mast relatively more
abundantin coniferousforests, and roots and vertebrates
relativelymore abundantin nonforestareas). Finally, to
matchthe analysis of skull shape, I stratifiedfood habits
studies by species, and for brown bears, additionallyby
those subspeciesalong the Pacific rim, in interiorregions
of Asia and NorthAmerica,and in Europe.
The stratificationsthat I used for this analysis mixed
intra-and interspecificvariation. Subspecies and groupings based upon vegetationstructurewere recognizedin
additionto species. This approachwas consistent with
my premises: that physical form, whether genetically
fixed or mechanicallyreinforced,was both a reflection
and determinantof diet; further,prevailing vegetation
structurewas a factor, in additionto bear morphology,
that influenced diet. In any case, taxonomy represents
relativelyartificialdistinctions,highlightedby the likelihoodthatpolarbearsaregeneticallyno moredistinctfrom
brown bears than brown bears are genetically diverse
among themselves (Croninet al. 1991).
Analysis
I used principalcomponents (PC) analysis to reduce
the numerousand somewhatarbitrarymeasuresof skull
and dental morphology (including both extant and extinct species) to fewer dimensionsthatwere theoretically
related to more fundamentalfeatures of skull morphology (Wilson 1976, Radinsky 1981b). Ratherthan control for allometric effects prior to PC analysis by, for
example,using standardizedresidualsfrom relationships
with skull length or body mass (Radinsky1981b), I simply used natural-logtransformationsof the raw values.
The firstPC thusdescribedsubstantialsize-relatedcorrelations of skull measurements(i.e., allometric effects),
while subsidiaryPCs were of primaryinterest because
they described variationin skull morphology independent of size. Loadings of skull variables on PCs were
used for functionalinterpretations,while species scores
were used to describe relationshipsbetween skull morphology and mean diet as well as morphologicalrelationships among species.
I used multipleanalysis of varianceand canonicaldiscriminantanalysis to test for and describedifferencesin
diet among bear species, subspecies, and bear populations groupedby dominantvegetationstructure.If diets
differedfor all beartypes (at a = 0.1), I used a multivariate analogof the protectedleast significantdifferenceprocedure based upon pair-wise Hotellings' T2 to identify
individual types that were not different (Johnson and
Wichern 1992). I used these tests, coupled with
Mahalanobisdistances,to characterizepair-wisedissimilaritiesbetweendiets of differentbeartypes. I interpreted
dietarydifferences between types by examining the canonicalcoefficientsof dietarycategoriesfor each canonical variable, conditioning my interpretationupon their
eigenvalues and the proportionof variationthatthey explained. I used rankedvalues in these multivariateprocedures as a nonparametricapproach that reconciled
parametrictechniqueswith non-normaldata(i.e., proportions) (Conoverand Iman 1980, 1981). I also weighted
each study in these analyses by the numberof years and
feces thathad been sampled(100 x [{numberof years +
numberof feces}/2], where years and feces were scaled
to be <1, which correspondedto the maximumvalue observed for each in any study). I deleted 1 dietary category (invertebrates)from these analysesto alleviatethe
linearnon-independencethatwouldhave otherwisearisen
from the proportionssummingto 1. In additionto these
multivariate procedures, I also report univariate tests
(analysis-of-variance[ANOVA] and multiple comparisons) for differencesamong means for each dietarycategory, again using rankeddata.
Because I wantedto predictdiets of 2 extinctbearspecies, I related skull morphology to diet through leastsquaresmultiplelinearregressionanalysis,with diet item
fractiondependenton skull morphologyof extantbears.
I used logits of mean fractions for each diet item (see
Aitchison's [1986] use of log-ratiotransformationsin the
analysis of compositionaldata) and species scores from
the first 3 PCs for the 7 groupingsof extantbear species
(pandas, polar bears, the 2 black bear species, and 3
groupingsof brownbearsubspecies). Although,apriori,
the dependenceof diet upon skull morphologyis equivocal and independentvariableswere measuredwith error,
this predictive approachis compatiblewith exploratory
analysis and hypothesis generation(Gilbert 1989:33). I
did not use canonicalcorrelationanalysisbecauseof prohibitively small sample sizes and my interest in elucidatingrelationshipsbetween skull morphologyand bear
use of specific diet items. The relationshipof fibrous
foods in the diet to skull morphologywas so strong(see
below) that this result would have been prominent,regardless of which statisticalmethod I used.
RESULTS
Diets
The diets of extantbears are diverse (Table 1). Giant
pandas and polar bears were distinguished,a priori, by
DIETANDMORPHOLOGY
OFNORTHERN
BEARS* Mattson 483
near exclusive folivory and carivory, respectively. Diets of the remaining 3 species were also different, globally (Wilks' A = 0.545; 10, 148 df; P < 0.001) and
from each other. Considering groupings that also reflected the dominantvegetation structureof each study
area, there was an aggregatedifference in diets (Wilks'
A = 0.199; 25, 265 df; P < 0.001) and pairwise differences among most groupings(Table 2). Only 2 pairwise
contrastswere not significant. Diets of Americanblack
bears that occupied broadleafforests did not differ from
diets of Asiatic black bears(also almostexclusively residents of broadleafforests), and diets of brownbearsthat
occupied arctic or alpine areas did not differ from diets
of brown bears that occupied coniferousforests. Otherwise, among the bears with more diverse diets, Asiatic
black bears and brown bears living in arctic or alpine
regions exhibiteddietaryextremescharacterizedby large
volumes of hard mast and by large volumes of vertebrates and roots, respectively (Tables 1 and 2).
The first 3 canonicalvariablesgeneratedby canonical
discriminantanalysis were significant and collectively
described96% of the variationin diets amongthe groupings basedon bearspecies andvegetationstructure(Table
3). The first canonical variable primarilydescribed a
gradientfrom diets rich in roots, vertebrates,and foliage
to one rich in mast, or basically a gradientfrom foods
typical of brownbearsin colder climates and more open
habitatsto foods typical of black bears in warmerforests (Fig. 2). The second and third canonical variables
described gradientsrelated more to trade-offs between
soft mast, and hardmast and roots (the lattertypified by
the diet of grizzly bearsin the interiorcontiguousUnited
States [Mattsonet al. 1991, Craigheadet al. 1982]), and
between vertebratesand roots and soft mast (the latter
typified by the diet of grizzly bears in much of interior
Canadaand Alaska), respectively.
Among the species with morediversediets, blackbears
were typifiedby the consumptionof mast, with hardmast
a more common food for black bears occupying broadleaf forests (Table 1, Fig. 2). Black bears living in coniferousforestsconsumedproportionatelymore soft mast
thanbear species living anywhereelse. With the exception of populations living in broadleaf forests, brown
bears were distinguishedby consuming greaterrelative
volumes of vertebratesand roots. Interestingly,the diet
of brown bears living in broadleafforests substantially
overlappedthe diet of black bears, suggesting that vegetation structure,as much as species, was a controlling
factor.
Not considering the self-evident differences in diets
of giant pandas and polar bears, the diets of other taxa
used for relating diet to skull morphology (i.e., American and Asiatic black bears, and Pacific coastal, European, and interiorcontinentalsubspecies of brownbear)
were also statisticallydistinct(Wilks' A = 0.160; 20, 223
df; P < 0.001). Aggregate diets differed primarilyby
root, soft mast, vertebrate,and hard mast volumes (in
order of loading on canonical variables). All taxa differedby pair-wisecomparisons,except the diet of coastal
brown bears did not differ from the diet of continental
brown bears. By individual diet category (Table 2),
Americanblack bears ate more soft mast than did continentalbrownor Asiatic black bears, while Asiatic black
bears ate more hard mast and continentalbrown bears
ate more roots than any othertype except coastal brown
bears.
Skulls
Principalcomponentsanalysis effectively reducedthe
9 measuredskull dimensions to 3, accountingfor 99%
of total variation (Table 4). Similar positive loadings
for all skull dimensionson PC1 reflectedsubstantialsizerelated correlationamong these measures (i.e., allometric relationships). Although there was relatively little
variation residual to PC1, this residual variation was
nonethelessof greatinterestbecauseit characterizedskull
shape and presumedassociationswith diet (e.g., adaptations), aside from size per se. Dental surface area, cranial and mandibular height, and zygomatic width
exhibitedpositiveloadingson PC2. These measureshave
either been positively associated with muscle size (primarily of the temporaland masseter) or with increased
dentalsurfaceareathattogetherprovidefor greatercrushing capability (Davis 1964, Kay 1975, Greaves 1978,
Radinsky198lb, Demes et al. 1986). The negative loading of nasal width (NW) and the positive loadings of
"face"and mandiblelengths (FL and ML) on PC3 indicate a comparativelynarrowand elongate snout at high
values and a broad and short snout at low values.
The main Ursus lineage was clusteredin PC2 x PC3
space separatefrom the giantpandaand short-facedbear
(Fig. 3). As expected (Davis 1964), the giant pandaexhibited the greatest presumed adaptationsfor exerting
crushing force at the occlusal plane. Also as expected
(Kurten 1967), the short-facedbear was distinguished
by the shortest and broadest snout of all taxa, together
with moderate apparent crushing capabilities. Taxa
within the genus Ursusvariedfrom the polarand American black bears, with relatively little crushingcapability
and shorter,broadersnouts,to the cave bear,which combined a relatively long snout with relatively greaterapparentadaptationsto crushing. The blackbearsexhibited
484
Ursus 10:1998
Table 1. Mean percent volume of diet items in feces of northern bears, averaged over individual studies. Means not different
at a = 0.1 (based on ranks) are followed by the same letters in columns, only including taxa with sample sizes n > 5. Brown
bears and American black bears are geographically stratified in 2 ways that are described in the text, with statistical groupings
for the 2 classifications separated by a ''. Values do not sum exactly to 1 across rows because values for each type of diet
item are averaged over individual studies.
Taxa/Geographic
groupings
Giant panda
Polar bear
Asiatic black bear
Americanblack bear (1)
Broadleaf forest
Coniferous forest
Americanblack bear (2)
Brown bear (1)
Broadleaf forest
Coniferous forest
Arctic/alpine
Brown bear (2)
European
Continental
Pacific coastal
No. studies
(n)
Fecescomposition
(%volume)
Foliage
Roots
Softmast
Hardmast
2
2
7
99.0
1.0
15.2b/b
0.0
0.0
O.Ob/b
0.1
1.0
34.8b/b
0.0
0.0
43.1a/a
0.0
0.0
5.4abcd/bc
16
16
32
22.8b/
23.6ab/
19.5 /ab
1.4b/
O.Ob/
0.1 /b
49.4ab/
61.2a/
55.3 /a
20.0a/
1.4b/
10.7 /b
5.9abc/
4.4abc/
5.1 /ab
15
19
9
22.8ab/
30.9a/
26.2ab/
1.4b/
9.2a/
14.0a/
46.9b/
39.5b/
38.5c/
7.8b/
7.0b/
O.Ob/
12.2a/
3.0bcd/
0.8cd/
9.0ab
7.5ab
20.6a
12
23
8
20.7 /ab
29.9 /a
28.4 /ab
0.2 /b
12.4 /a
4.3 /ab
48.4 /ab
37.1 /b
45.5 /ab
2.5 /b
5.6 /b
11.4 /ab
14.5 /a
2.0 /c
3.5 /abc
13.3 /a
10.9 /ab
6.9 /ab
a modal, conceivablymore generalizedskull shape. The
closely related polar, brown, and cave bears (Kurten
1964, 1976; Shields and Kocher 1991) all displayed a
substantial diastema separating the canines from premolars, although this gap was most pronouncedin the
cave bear (Fig. 3b).
Relationshipsof Dietto SkullMorphology
Therewas a stronglinearrelationshipbetween fibrous
foods (Y) and PC1 and PC2 (R2= 0.97; 2,4 df; F= 76.0;
P < 0.001):
Y = -0.122 + 0.331PC1 + 3.38PC2
Fibrous foods were comprised of foliage and
roots added together (', as a logit, where diet
fraction [p] = e t-0.0/[ 1+e t-001]) Standardized coeffi-
Invertebrates
Vertebrates
0.1
97.5
2.4b/b
4.1 ab/
3.5b/
3.8 /bc
cients for PC1 and PC2 were 0.272 and 1.004, respectively; both coefficients were also significantlydifferent
from zero. In other words, relative volumes of fibrous
foods in the feces of extant bears were related,first, to
apparentadaptationsof the skull to crushingand, second,
to largersize. A naive extrapolationfrom this relationship would predictthatcave bearfeces, on average,contained84% fibrousfoods and that short-facedbearfeces
contained 87% fibrous foods during hyperphagia(Fig.
4). No other diet category exhibited a statisticallysignificantrelationshipto PCs describingskull morphology.
Considerations
OtherMorphological
The short-facedbear skull exhibited some additional
interestingand potentiallydiagnosticfeatures. Its snout
Table 2. Pair-wise Mahalanobis distances (D2) between northern bears based on ranked dietary composition of feces, not
considering invertebrates. The giant panda and polar bear were excluded because of small sample sizes. Pairwise
comparisons denoted by an * were significantly different (a = 0.1, using Hotellings' T2).
groupings
Taxa/Geographic
groupings
Taxa/Geographic
Americanblack bear-broadleaf forest
Americanblack bear-coniferous forest
Brown bear-broadleaf forest
Brown bear-coniferous forest
Brown bear-arctic or alpine
Asiatic
blackbear
0.12
0.54*
0.35*
0.91*
1.14*
American
blackbearforest
broadleaf
0.34*
0.19*
0.67*
0.79*
American
blackbearconiferous
forest
Brownbearforest
broadleaf
Brownbearforest
coniferous
0.13*
0.54*
0.42*
0.42*
0.53*
0.16
DIETANDMORPHOLOGY
OFNORTHERN
BEARS* Mattson 485
0 ARCTICOR ALPINEBROWNBEAR
FORESTBROWNBEAR
a CONIFEROUS
Table 3. Standardized canonical coefficients for diet
variables and means for groupings of northernbears on the
first 3 canonical variables from canonical discriminant
analysis (n = 82). Eigenvalues, proportionof total variation,
and significance are also given for each canonical variable.
Invertebrates were excluded from the analysis to achieve
linear independence, and polar bears and giant pandas were
excluded because of small sample sizes.
(a)
1.0
FOREST
BROWN
BEAR
0 BROADLEAF
FORESTAMERCIANBLACKBEAR
* CONIFEROUS
* BROADLEAFFORESTAMERICANBLACKBEAR
* ASIATICBLACKBEAR
HARD MAST
ROOTS
C.'
w
CL)
variable
Canonical
CAN1
CAN2 CAN3
1-4 0.5
I
0 ?
4
Diet variables
0.328
Foliage
1.299
Roots
0.183
Soft mast
-0.337
Hardmast
0.391
Vertebrates
Bear groupings
-0.504
Asiatic black bear
Americanblack bear-broadleaf forest -0.386
Americanblack bear-coniferous forest -0.125
-0.170
Brown bear-broadleaf forest
0.415
Brown bear-coniferous forest
0.441
Brown bear-arctic/alpine
1.325
Eigenvalue
0.683
Proportionof variation
<0.001
Significance
-0.317
0.533
-0.448
0.675
-0.130
-0.142
0.364
0.359
-0.144
-0.919
0.251
0.145
-0.350
-0.100
0.141
-0.145
0.466
0.196
<0.001
0.031
0.107
0.042
-0.182
-0.052
0.214
0.197
0.083
0.018
0.0
z
0
z
I
U
-0.5
- SOFT MAST
-
ROOTS, VERTEBRATES, & FOLIAGE -
MAST
-1.0
-0.5
0.0
0.5
1.0
1.55
CANONICAL VARIABLE I
(b)
1.0
is relatively the widest and flattest of the examinedbear
species, visually most similar to that of the polar bear
and markedly dissimilar to the black and brown bears
(Fig. 5). Among extantbears (based primarilyupon interpretationsof the polarbearskull) andothercarnivores,
this feature is thought to facilitate delivery of a killing
bite to prey and consumption and dismembermentof
larger carcasses (Kurten 1964, Ewer 1973, Radinsky
1981b). The short-facedbear's skull is also remarkably
similarin both profile and top view to thatof the spotted
hyena (Crocutacrocuta)(Fig. 6), except thatits teeth are
proportionallysmaller and the cusps less developed.
Body sizes and skeletal proportionsalso offered insightintopossibledietsof the 2 extinctbearspecies (Table
5). The giant short-facedbear had relativelythe longest
metatarsals(low FMT) and forelegs (high VHR) of any
bear,with the lattermeasurein rangeof othercarnivores
that were pursuitratherthan ambush hunters. By contrast,the cave bear had short metatarsalsand legs comparedto all otherbears except the pandaand polarbear.
However, compared to other carnivores, all bears had
short metatarsals(Table 5). Relative to known specialized scratchdiggers (Hildebrand1985), the short-faced
bear had slender and the cave bear comparablyrobust
humeri, with a trend towards similarly short radii (low
RH) in the cave bear. Naively assuming that the short-
1l\
0;I
0.5
C4
SOFT MAST
ROOTS
I ''
0.0
*4
-X
\-
Mfy
U -0.5
z
0
z
U -1.0
VERTEBRATES
-1~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
<-
MAST
ROOTS, VERTEBRATES, & FOLIAGE->
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
CANONICAL VARIABLE 1
Fig. 2. Groupings of extant northern bear are plotted with
respect to the (a) first and second and (b) second and third
canonical variables (CANs), from canonical discriminant
analysis of fecal composition described in terms of 5 broad
diet categories for the period of hyperphagia. General
interpretationsof CANs are also given.
faced bear was a predator(Kurt6n1967), relationships
between sizes of extantpredatorsand theirprey (Vezina
1985) lead to a predictionthat the short-facedbear (335
?
486
Ursus 10:1998
Table 4. Species scores, eigenvectors of skull variables,
and eigenvalues for the first 3 principal components (PC)
from analysis of northern bear skulls. Skull variables are
explained in the text and Fig. 1.
(a)
Parameters
Species
Polar bear
Americanblack bear
Asiatic black bear
Brown bear-European
Brown bear-Pacific coastal
Brown bear-continental
Cave bear
Short-facedbear
Giant panda
Skull variables
DENTA
CH
MH
ZW
FL
ML
CBL
NW
Eigenvalues
Proportionof variation
PC1
-0.40
-2.79
-4.54
-0.49
1.96
0.42
3.77
3.19
-1.11
0.317
0.365
0.363
0.367
0.355
0.357
0.355
0.346
7.30
0.912
PC2
PC3
-1.04
-0.33
0.16
-0.32
-0.32
-0.06
0.16
0.28
1.47
-0.25
-0.31
0.05
0.32
0.34
0.29
0.36
-0.84
0.03
0.735
0.186
0.175
0.121
-0.313
-0.328
-0.402
-0.110
0.463
0.058
0.224
-0.111
0.081
-0.070
0.370
0.229
0.129
-0.848
0.162
0.020
U arctos
(Pacific coastal)
ELONGATE
0.5
NARROW
SNOUT
en
0
A
U arctos
z
z
0
CU spe/aeus
0
*
(European)
Carctos (continental)
-
0.0
0
0
urpod
Ailuropoda
U. thibetaus
U. maritimus
U
U*
U.
americanus
-0.5
z
Arctodus
a.
- 1.0 -
1.5
CRUSHING
-1.0
FORCE
-0.5
PRINCIPAL
>
0.0
0.5
COMPONENT
1.0
1.5
2
(b)
0.5
z
w
z
0
a.
0.0
0
U
*^>0~6
?S??9^^S
^^~_
o
<^o?G<;
-0.5
kg) preyedon animalsas large as roughly2200 kg, in the
range of most mastodons (Mammutamericanum) and
smaller, presumably younger, woolly mammoths
(Mammuthusspp.).
DISCUSSION
Diets of ExtantSpecies
It is not altogetherclearfromthis analysisthe extentto
whichtaxonalone, as opposedto prevailingenvironmental conditions, influenced what bears ate. Nonetheless,
several expected patterns were evident. Asiatic and
Americanblack bears ate more mast than any otherbear
species, and when living in broadleafforests, consumed
relativelylarge volumes of hardmast. Small body sizes
presumablyallowed them access to tree mast in the face
of intense competition from other frugivores (e.g.,
Bromlei 1965, Pelton 1982), while skull morphologyof
these samebearsprobablylimitedtheiruse of foods such
as roots, bamboo, and other graminoids that are more
fibrous. Certainly,the contrastin food habits of Asiatic
blackbearswith sympatricgiantpandasandbrownbears
supportthis interpretation(Bromlei 1965, Schalleret al.
1989). Bears living in cold, open environmentsalso ate
the vertebratesandfibrousfoods thatwere comparatively
0.
a.
^o(20^)
-1.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
PRINCIPAL COMPONENT 2
Fig. 3. Extant or recently extinct northern bear taxa are
plotted with respect to the second and third principal
components
(PC2 and PC3) from analysis of skull
measurements. (a) Functional or more general interpretations
are shown for each PC as well as (b) skull profiles and planar
views of the upper right dental row, shown proportional to
their relative size. The 3 Ursus arctos subgroupings in (a)
are represented by a simple skull in (b).
more abundantthere. My analysis and Herrero(1978)
strongly suggest that brown bears are better able to extract and process foods that are more fibrous, including
roots, and thus are betteradaptedto living in non-forest
environments.This interpretationis tenuouslysupported
by the few investigationsof sympatricAmericanblack
and grizzly bears (e.g., Aune 1994), that have observed
virtuallyno excavatedfoods in black bear feces in areas
where excavated foods were a majorpartof the grizzly
bear diet.
Brown bears are capableof living in temperateforests
like thoseof easternNorthAmericaandeatinga dietmuch
like thatof Americanblack bears. This is clear from the
DIET AND MORPHOLOGY
OF NORTHERNBEARS * Mattson
Panda
\
100
rj
0
0
w
40
80
UA
C)
0
0
60
z
40
ZZ
0
04
Pm
20
P4
487
val densities of humans, were effective at competitively
excluding brown bears under favorable environmental
conditions; black bears via scramble competition and
humansby directmortalityand interferencecompetition.
Asiatic black bears occupied Europepriorto the last glacial epoch (Erdbrink1953a) andwere probablyprevented
from reoccupyingEuropefrom theirSoutheastAsian refuge because habitat bridges (i.e., contiguous temperate
deciduous forests) did not develop across central Asia
duringthe Holocene (Khotinsky1984). Underthese conditions, Europeanbrownbears would have had comparatively unimpededaccess to expandingenvironmentsthat
might have otherwise favored black bears.
Diet and Morphology
0
Polar bear
BEAR TYPE
Fig. 4. Percent fibrous foods (foliage and roots) predicted in
the feces of short-faced and cave bears are shown, based
upon the relationship of fibrous foods to principal
components from the morphometric analysis, as well as
observed and predicted volumes of fibrous foods in feces of
extant northern bears. Standard error bars are shown for
observed and predicted values. Black bears are denoted by
bib and brown bears by brb.
long historyof brownbears living in pure or mixed broadleaf forests in Europe and by the extent to which diets
of these bears overlappedwith diets of American black
bears. However, in contrast to Europe, temperate deciduousforestsin NorthAmericaandAsia aremuchmore
closely associated with black bears. It seems fruitful in
the futureto pursuethe hypothesis that North American
black bears, in combinationwith relatively high prime-
Many featuresof skull and body morphologywere associated with the diets of extant bears. A few of these
associationswere discussed in the previous section. This
resultdoes not contradictthe expectationthatbody structure would shape as well as reflect food habits througha
complex interplay of specific evolved physical adaptations, phenotypic plasticity, and longer-termevolutionary selection. In the shortterm,associationsbetween diet
andmorphologylikely restupondevelopmentalresponses
(e.g., hypertrophy)to mechanical forces (Moore 1965)
and somatic adaptationsthat allow individual bear species to at least survive on, if not competitively use, certain foods. In eithercase, these relationshipscan be used
to judiciously speculate about the food habits of extinct
ursids as well as the niches of surviving bears.
Most variationin the skull morphologyof extantbears
is associated with the amount of fibrous foods in their
diet. This associationis plausibly interpretedin terms of
-.o
.
American
black bear
Brown bear
Polar bear
Fig. 5. Planar views of snouts are shown, emphasizing the nasal opening
representative extant bear species and the extinct short-faced bear.
Short-faced
bear
and bridge and rostral shape, from three
488
Ursus 10:1998
Spotted hyena
Short-faced bear
nipulateselectfibrousvegetalfoods. By contrast,a shorter
rostrumcombined with a broadernasal opening and retractednasal bridge (epitomizedby the polar and shortfaced bears) implies several things-for one, a broader
muzzle (Kurten1976), but more importantly,a rostrum
thatbetterresiststhe torsionalforces associatedwith, and
otherwise facilitates, grasping and manipulatinglarge
objects.
The substantialdiastemaof polarbearsis inconsistent
with the positive associationof this featurewithincreased
selection and manipulationof vegetal foods. Polarbears
are almost strictlycarnivorous.The diastemain this species is thus probablybetterunderstoodin light of its recent derivationfromthe brownbearlineage(Kurt6n1964,
Goldmanet al. 1989, Shields and Kocher 1991) and the
evolutionarilyconservativenatureof mostdentalfeatures.
It may simply be that the diastemain this species is an
artifactof history and has no strong relationshipto its
presentdiet.
The Cave Bear
of a short-facedbear(thicklines)
Fig.6. Superimpositions
andspottedhyena(thinlines)skullareshown,scaledto the
same condylobasal length, in planar view and profile.
Shadingdepicts areas of discrepancybetweenoutlinesof
the 2 skulls.
adaptationsthatincreasecrushingcapabilities- primarily throughincreasesin dentalcrushingsurface,accommodationsfor largertemporalandmassetermuscles (i.e.,
increasedzygomaticwidthandskullandmandibleheight),
and configurationsthatbring the associatedforces more
effectively to bearon the occlusal plane. The gradientin
snoutshape,along with developmentof the diastemabetween caninesandpremolars,has less obvious functional
interpretations.
An elongate and comparativelynarrowsnout is commonly interpretedas an adaptationto moreselective feeding (Preuschoft et al. 1986, Gordon and Illius 1988,
Greaves 1991). A diastemanaturallyarises from retention of crushingandgrindingsurfacesat the distal end of
the dentary (Greaves 1991), nearer the zygoma where
torsionalstressesaredecreasedandthe temporalmuscles
maximallyeffective (Preuschoftet al. 1986). A diastema
between canines and premolarshas also been associated
with greater manipulationof vegetal foods within the
mouth(Adrianet al. 1958). By eitherinterpretation,this
trendculminatingin the cave bearis likely associatedwith
increaseduse of the mouth to procure and furtherma-
My results are consistentwith previoushypothesesby
researcherssuch as Erdbrink(1953b) and Kurt6n(1976),
and confirming evidence from analyses of carbon and
nitrogenisotopes by Bocherenset al. (1994) thatthe cave
bear ate a diet largely comprised of fibrous foods.
Hilderbrand's(1996) analysisof isotopes suggestsa more
diverse diet, but is open to interpretation.Even so, the
relatively simple ursid digestive tract (Jaczewski et al.
1960, Davis 1964, Mealey 1975), combinedwith the relatively shortperiod of either low fiber content or abovegroundavailabilitytypicalat higherlatitudes,makea diet
comprised largely of foliage highly improbable. The
panda accomplishes this feat aided by specific adaptations (e.g., 'the thumb') and the availabilityof a forage
(bamboo)thatis used by few competitorsandhas remarkably constantbiomass and nutritionalvalue throughout
the year (Davis 1964, Schaller et al. 1985). A diet of
leaves and stems is even moreunlikely given the competition likely posed to grazing cave bears by the diverse
faunaof largerherbivoresin PleistoceneEurope(Kurten
1968).
These complicationscan be reconciled with a fibrous
vegetal diet by hypothesizingthe cave bear to be a specialized root grubber. A root diet would be consistent
with the extremewearevident in cave bearteeth (Kurten
1958) and massive blunt claws well-suited (Hildebrand
1985) and previously hypothesized as adaptations to
scratchdigging (Erdbrink1953b,Kurten1976). It would
also fit the stoutfrontlimbs andpowerfulshoulderarchitectureof cave bears(Erdbrink1953b,Kurten1976). This
BEARS* Mattson
OFNORTHERN
DIETANDMORPHOLOGY
489
Table5. Body mass and indices of relativeskeletaldimensionsfor northernbears and averagedfor othercarnivoresby
pursuitand ambushhunters(? 1 SD). Definitionsand sources are given in footnotes.
Skeletalindicesa
Species or group
Species
Short-faced bearb
Americanblack bearc
Asiatic black beard
Brown beare
Cave bearf
Polar bearg
Giant pandah
Group
Pursuithuntersi
Ambush huntersi
Mass (kg)
FMT
VHR
RH
HW
335
101
86
178
284
288
125
4.58
4.67
5.30
4.75
4.83
5.30
9.13
6.50
5.56
84.6
90.3
8.7
5.76
5.51
5.03
5.06
88.6
75.6
9.3
10.5
77.1
9.4
34 ? 17
37 ? 39
2.40 ? 0.1
2.59 ? 0.3
5.96
4.56 ? 0.3
a FMT = femur length/lengthof longest metatarsal;VHR = humerus+ radiuslength/lengthof thoracicvertebrae10 + 11 + 12; RH = radius
length/humeruslength as %; HW = narrowestwidth of humerusshaft/humeruslength as %.
b Kurt6n
(1967), Emslie and Czaplewski (1985).
c Davis (1964), Emslie and Czaplewski (1985), Van Valkenburgh(1985).
d Bromlei
(1965), Van Valkenburgh(1985).
e Erdbrink(1953b), Davis (1964), Kurt6n(1967), Emslie and Czaplewski (1985), Van Valkenburgh(1985).
f Erdbrink(1953b), Kurten(1967).
g Van Valkenburgh(1985).
h Davis (1964).
i Davis (1964), Van Valkenburgh(1985).
latterfeaturehas been associatedwith fixing the scapula
in resistanceto pulling forces along the long axis of the
fore-limbs (Davis 1949), more specifically among grizzly bears as an adaptation to digging (Herrero 1978,
Craigheadand Mitchell 1982).
The cave bears'largediastemawouldbe consistentwith
the root-grubbinghypothesis. The diastema appearsto
be associated with manipulation of both hedysarum
(Hedysarum spp.) and yampah (Perideridiagairdneri)
when grizzlies are consumingroots of these plants(pers.
obs.). Cave bears may have had access to many large
starchy-rooted species, such as Hedysarum spp. and
Conopodiummajus, that are common in boreal, arctic,
andalpineregionsof present-dayEurasia(Komarov1948,
Couturier1954). It is also possible that some of the numerous ancestralor surviving Eurasianspecies of pika
(Ochotonaspp.),groundsquirrel(Spermophilusspp.),and
marmot(Marnota spp.) wouldhave been availableto and
used by a cave bear that specialized in scratchdigging.
This possible relianceon roots and otherfibrousfoods
provides additionalcontext for understandingthe cave
bears' extinction. First,this type of diet would be consistent with the very low reproductiverate postulated by
Kurten(1958), similar to that observed for the herbivorous panda (Schaller et al. 1985). Although large body
size could partlyexplain low fecundity,this relationship
does not automaticallyfollow from greaterbody mass.
Lower fecundity and more specialized food habits predictablywould have madecave bearsvulnerablenot only
to a rapidlychanginglandscape,but also to the competition and possible addedmortalitybroughtby the advent
of 2 generalistomnivores (brownbears and modem humans) in Europeduringthe late Pleistocene. Arrivalof
modemhumansabout30,000 yearsago couldhave tipped
the scale against cave bears in what may have been an
alreadystringentcompetitive trianglethat included Neanderthals(Homosapiensneanderthalensis)andrecently
arrivedbrown bears. In any case, these results are consistent with the hypothesis that humanswere an importantcatalystin the demise of cave bears(Erdbrink1953b,
Kurt6n1976).
The Short-Faced Bear
Kurt6n(1967) and Emslie and Czaplewski (1985) offer strikinglydifferenthypotheses about the food habits
of short-facedbears. Kurten (1967) evoked an active
predatorthat could achieve comparativelyhigh speeds
(for a bear)because of relatively long legs and forwardaligned feet. He also speculatedthat the broadrostrum
and short face and neck facilitatedgrabbingand subduing prey, complemented by features of the lower first
molarthataidedmeat consumption. By contrast,Emslie
490
Ursus 10:1998
and Czaplewski (1985) hypothesized that short-faced
bears were omnivores, if not near-exclusiveherbivores.
Their argumentrested principallyon the species' large
body mass (the short-facedbearis largerthanany extant
carnivoreandcomparablein size to some herbivores)and
alternateexplanationsfor featuresthatKurten(1967) had
associatedwith carivory.
The results presented here contributeto judging the
relative meritsof these contrastingviews. An uncritical
extrapolationof the relationshipbetween fibrous foods
and skull morphologyof extantbearspredictsthat shortfaced bearsate an almostwholly fibrousdiet. Yet if this
were so, the bear's broadrostrum,broad incisor arcade
and short face suggest a relatively unselective diet of
coarse foliage (Hylander1975, Gordonand Illius 1988,
Janis and Ehrhardt1988). As with the cave bear, this is
highly unlikely given the simple ursiddigestive tractand
the abbreviatedseason of high foragequalityat northerly
latitudes. Moreover,the short-facedbear's long legs and
shortneck(Kurt6n1967)wouldhavecomplicatedgroundlevel grazing,unlessthis speciesfilled thebrowsingniche.
Morelikely, the apparentcrushingcapabilitiesof its skull
facilitated foraging behavior other than grinding roots
(Kurten1967, Emslie and Czaplewski 1985) or foliage.
Emslie and Czaplewski(1985) also suggestedthat the
short-facedbear's tall and slenderstature(1.0-1.5 m at
the shoulder)may have been eitheran adaptationto scanning over tall groundcover or a possible means of procuring foliage and berries from the limbs of trees and
shrubs. It is hardto imagine what benefits would be derived for these purposes in the Beringian steppe-tundra
or when most groundcover duringthe late Pleistoceneof
modem-day western United States was likely <1 m tall
(Thompson and Mead 1982). In any case, a bear that
browsedflowers, foliage, seeds, or berriesfromthe canopies of trees andshrubswould have almostcertainlybeen
severely handicappedby a rostrumand incisor arcadeas
large and as flattened as that of the short-faced bear
(Hylander 1975, Gordon and Illius 1988, Janis and
Ehrhardt1988). Again, presumedadaptationsto either
herbivoryor frugivoryfail to give a satisfactoryexplanation of the short-facedbear's body proportions.
Emslie and Czaplewski (1985) further contend that
large body size is evidence against carivory. Several
authors (e.g., Rosenzweig 1968, Hespenheide 1975,
Vezina 1985) have observed a relatively strongpositive
relationshipbetween the size of solitary predatorsand
their prey. Largerprey presumablyallow for the existence of largerpredators. Large body size would therefore only be a convincing argumentagainstcarnivorous
short-facedbears if there were no commensurate-sized
prey in Pleistocene NorthAmerica. In fact, a numberof
large herbivores existed, including mastodons, mammoths,andbison(Bison spp.;Kurt6nandAnderson1980),
thatcould have constitutedthe prey of a predatoras large
as the short-facedbear(see Results). It seems more than
coincidence that the largest short-faced bears (A. s.
yukonensis) sharedPleistocene Alaska with some of the
highest North American mammoth concentrations
(Agenbroad1984). Largebody size is not an inherently
strongargumentagainstcamivory, and in this case may
even be an argumentfor it.
Carnivorousfood habits still provide the most consistent and compelling explanationfor diagnostic features
of short-facedbear morphologyand are also consistent
with the constraintsof a simpleursiddigestive tract. This
interpretationis furthermoreconsistent with previously
observed associations of short-facedbear remains and
herbivorebones markedand fragmentedin ways characteristic of bear scavenging (Agenbroadand Mead 1986,
Voorhies and Comer 1986, Guthrie 1988, and Gillette
andMadsen 1992) andwith analysisof stablecarbonand
nitrogenisotopes from bone collagen (Matheus 1995).
Skeletal indices suggest a relativelymobile bear (Van
Valkenburg1985, GarlandandJanis1993)that,like many
otherspecies, perhapsachievedthis performancein spite
of an arrayof unique phylogenetic traits (e.g., a plantigrade or semi-plantigradeposture; Taylor et al. 1974,
GarlandandJanis 1993). The skull exhibitsfeaturesthat,
along with the short neck, are plausibly interpretedas
adaptationsto grasp,kill, and dismemberprey, probably
in concertwith the characteristicallyflexuous ursidpaws.
If this bear engaged in ambush-typepredation,it would
likely have been swift enough to catch its large-bodied
prey, perhaps grasping it from the rear with its paws
and collapsing the hind-quarterswith a combination
of weight and a crippling bite to the back, much like
contemporarybrown bears. A skull in many respects
strikingly similar to that of the spotted hyena may have
furthermore aided a short-faced bear's manipulation
and dismemberment of a very large carcasses that it
had obtained.
If the short-faced bear were adapted to prey on or
scavenge very large herbivores, then its' late-Pleistocene extinction is comparatively easy to understand.
Most likely, the short-faced bear went extinct because
its primaryfood went extinct (McDonald 1984, OwenSmith 1987). Kurtenand Anderson (1974, 1980) suggested that competition with newly arrived brown
bears was a major cause. In contrast, Matheus (1995)
and results presented here suggest that there would
have been relatively little niche overlap between the 2
DIETANDMORPHOLOGY
OFNORTHERN
BEARS* Mattson
species and that brown bears probably played a relatively minor role in the short-faced bears' extinction;
perhaps simply hastening its demise throughscramble
competitionfor carrionof smaller-bodiedherbivores.
Matheus (1995) hypothesizes that short-faced bears
wereprimarilyscavengers,discountingtheirpotentialrole
as predators. Whereasthe results presentedhere do not
constitute evidence for or against this proposition,it is
worth reflecting on the food habits of 2 other species,
spottedhyenas and brownbears,that, when eating meat,
are conventionallydeemed to be scavengers. In particular,field studieshave shownthat,even thoughtheirphysical form is consistentwith scavenging,these species can
be formidablepredatorsand can, in fact, derive much of
theirmeatfromkills (Kruuk1972,Mattson1997). I therefore arguefor caution againstover-interpretingbehavior
from physical form.
CITED
LITERATURE
G.M., F.H. KEMP, AND W.D.L. RIDE. 1958. A
radiographicanalysis of masticationand swallowing in the
domestic rabbit Oryctolagus cuniculus. Proc. Zool. Soc.
Lond. 130:257-274.
L.D. 1984. New World mammothdistribution.
AGENBROAD,
Pages 90-108 in P.S. MartinandR.G. Klein,eds. Quaternary
extinctions: a prehistoricrevolution. Univ. ArizonaPress,
Tucson.
, ANDJ.I. MEAD. 1986. Large carnivores from Hot
Springs mammoth site, South Dakota. Natl. Geogr. Res.
2:508-516.
J. 1986. The statisticalanalysis of compositional
AITCHISON,
data. Chapmanand Hall, New York, N.Y. 416pp.
AUNE,K.E. 1994. Comparativeecology of black and grizzly
bears on the Rocky MountainFront, Montana. Int. Conf.
Bear Res. and Manage.9:365-374.
BAILEY,R.G. 1989. Ecoregions of the continents. Environ.
Conserv. 16:MapSuppl.
BOCHERENS, H., M. FIZET, AND A. MARIOTTI. 1994. Diet,
physiology and ecology of fossil mammalsas inferredfrom
stable carbon and nitrogen isotope biogeochemistry:
implications for Pleistocene bears. Palaeogeogr.,
Palaeoclimatology,Palaeoecol. 107:213-225.
F.G. 1965. Bears of the south far-easternUSSR.
BROMLEI,
Zdatel'stvo"Nauka",Moskva-Leningrad.IndianNatl. Sci.
DocumentationCent.,New Delhi. (TranslatedfromRussian.)
BUNNELL,
F.L., ANDD.E.N. TArr.1981. Populationdynamicsof
bears implications. Pages 75-98 in C.W. Fowler and T.D.
Smith, eds. Dynamics of large mammalpopulations. John
Wiley and Sons, New York, N.Y.
CONOVER,
W.J., ANDR.L. IMAN.1980. The ranktransformation
as a methodof discriminationwithsomeexamples. Commun.
Stat. Theory and MethodsA9:465-487.
. 1981. Ranktransformationsas a bridge
, AND
between parametricand nonparametricstatistics. Am. Stat.
ADRIAN,
491
35:124-133.
R.S. 1980. Size and positioning of the teeth and
CORRUCCINI,
fossarelativetotaxonomicanddietaryvariation
infratemporal
in primates. Acta Anat. 107:231-235.
M. 1954. L'ours brun. Grenoble,France. 904pp.
COUTURIER,
(In French.)
1982. Grizzly bear Ursus
CRAIGHEAD,
J.J., ANDJ.A. MITCHELL.
arctos. Pages 515-556in J.A.ChapmanandG.A. Feldhamer,
eds. Wild mammalsof NorthAmerica. JohnsHopkinsUniv.
Press, Baltimore,Md.
, J.S. SUMNER,AND G.B. SCAGGS. 1982. A definitive
systemforanalysisof grizzlybearhabitatandotherwilderness
resources. Wildl.-WildlandsInst.Monogr.No. 1, Missoula,
Mont. 279pp.
CRONIN, M.A., S.C. AMSTRUP, G.W. GARNER, AND E.R. VYSE.
1991. Interspecificand intraspecific mitochondrialDNA
variationin North Americanbears (Ursus). Can. J. Zool.
69:2985-2992.
D.D. 1949. The shoulderarchitectureof bearsandother
DAVIS,
carnivores. Fieldiana:Zool. 31:285-305.
1964. The giant panda: a morphological study of
evolutionarymechanisms. Fieldiana:Zool. Memoirs Vol.
3. 339pp.
1986.
DEMES, B., N. CREEL, AND H. PREUSCHOFT.
Functional
significanceof allometrictrendsin thehominoidmasticatory
apparatus. Pages 229-237 in J.G. Else and P.C. Lee, eds.
Primate evolution. Vol. 1. Cambridge Univ. Press,
Cambridge,U.K.
J.F. 1981. The mammalianradiations: analysis of
EISENBERG,
trendsin evolution,adaptation,andbehavior.Univ. Chicago
Press, Chicago,Ill. 610pp.
N.J. CZAPLEWSKI.
1985. A new recordof giant
EMSLIE,
S.D., AND
short-faced bear, Arctodus simus, from western North
Americawith a re-evaluationof its paleobiology. Nat. Hist.
Mus. of Los Angeles County,Contrib.Sci. 371:1-12.
D.P. 1953a. A review of fossil and recent bears of
ERDBRINK,
the Old Worldwith remarkson theirphylogeny based upon
theirdentition. Vol. I. Deventer,Netherlands. 320pp.
1953b. A review of fossil and recentbearsof the Old
Worldwithremarkson theirphylogenybasedupondentition.
Vol. II. Deventer,Netherlands. 277pp.
ERICKSON,A.W., J. NELLOR,ANDG.A. PETRIDES.1964. The black
bearin Michigan. Mich. State Univ., Agric. Exp. Stn. Res.
Bull. 4. 102pp.
R.F. 1973. The carivores. CornellUniv. Press, Ithaca,
EWER,
N.Y. 494pp.
GARLAND,
T., JR., ANDC.M. JANIS.1993. Does metatarsal/femur
ratiopredictmaximalrunningspeed in cursorialmammals?
J. Zool. (Lond.) 229:133-151.
GILBERT,N. 1989. Biometricalinterpretation:makingsense of
statistics in biology. Oxford Univ. Press, Oxford, U.K.
129pp.
GILLETTE,
D.D., ANDD.B. MADSEN.1992. The short-faced bear
Arctodus simus from the late quaternaryin the Wasatch
Mountainsof centralUtah. J. Vertebr.Paleontology12:107112.
J.L. 1986a. Carnivore life history patterns:
GITTLEMAN,
492
Ursus 10:1998
allometric,ecological and phylogenetic associations. Am.
Nat. 127:744-771.
.1986b. Carnivorebrainsize, behavioralecology, and
phylogeny. J. Mammal.67:23-36.
, EDITOR.1989. Carnivore behavior, ecology, and
evolution. CornellUniv. Press, Ithaca,N.Y. 620pp.
1982. Carnivorehome-rangesize,
, ANDP.H. HARVEY.
metabolicneedsandecology. Behav.Ecol. Sociobiol. 10:5764.
D., P.R. GIRI,ANDS.J. O'BREN. 1989. Molecular
GOLDMAN,
genetic-distanceestimates among the Ursidae as indicated
by one- and two-dimensional protein electrophoresis.
Evolution 43:282-295.
GORDON,
I.J., ANDAW. ILLIUS.1988. Incisor arcadestructure
and diet selection in ruminants. FunctionalEcol. 2:15-22.
W.S. 1978. The jaw lever system in ungulates: a
GREAVES,
new model. J. Zool. (Lond.) 184:271-285.
.1991. A relationshipbetween premolarloss andjaw
elongation in selenodont artiodactyls. Zool. J. Linn. Soc.
101:121-129.
R.D. 1988. Bone litterfrom an AlaskanPleistocene
GUTHRIE,
carnivoreden. CurrentRes. in the Pleistocene 5:69-71.
.1990. Frozenfaunaof the mammothsteppe:the story
of blue babe. Univ. Chicago Press, Chicago, Ill. 323pp.
HALL,E.R. 1984. Geographic variation among brown and
grizzlybears(Ursusarctos)in NorthAmerica. Univ. Kansas
Mus. Nat. Hist., Spec. Publ. 13. 14pp.
S. 1978. A comparison of some features of the
HERRERO,
evolution,ecology andbehaviorof blackandgrizzly/brown
bears. Carnivore1:7-17.
H.A. 1975. Preycharacteristicsandpredatorniche
HESPENHEIDE,
width. Pages 158-180 in M.L. Cody and J.M. Diamond,
eds. Ecology andevolutionof communities. BelknapPress,
Cambridge,Mass.
M. 1985. Digging of quadrupeds.Pages 89-109
HILDEBRAND,
M.
in
Hildebrand,D.M. Bramble, K.F. Liem, D.B. Wake,
eds. Functionalvertebratemorphology.HarvardUniv. Press,
Cambridge,Mass.
G.V. 1996. Use of stable isotopes to determine
HILDERBRAND,
diets of living and extinct bears. M.S. Thesis, Washington
State Univ., Pullman. 29pp.
W.L. 1975. Incisor size and diet in anthropoids
HYLANDER,
withspecialreferenceto Cercopithecidae.Science 189:10951097.
Z., J. GILL,ANDS. KozNmwsKI.1960. Capacityof
JACZEWSKI,
the differentparts of the digestive tractin the brown bear.
Trans.Int. Union Game Biol. 4:146-154.
1988. Correlationof relative
C.M., ANDD. EHRHARDT.
JANIS,
muzzle width and relative incisor width with dietary
preferencein ungulates. Zool. J. Linn. Soc. 92:267-284.
1992. Appliedmultivariate
R.A., ANDD.W. WICHERN.
JOHNSON,
ed.
Third
Prentice-Hall,Englewood
statistical analysis.
Cliffs, N.J. 642pp.
KAY,R.F. 1975. The functionaladaptationsof primateteeth.
Am. J. Physical Anthropology43:195-216.
N.A. 1984. Holocene vegetation history. Pages
KHOTNSKY,
179-200 in A.A. Velichko, ed. Late Quaternary
environmentsof the Soviet Union. H.E.Wright,Jr.andC.W.
Baronsky, eds. Univ. Minnesota Press, Minneapolis.
(English-languageed.)
KOMAROV,V.L. 1948. Flora of the U.S.S.R. Vol. 13.
Leguminosae: Oxytropis, Hedysarum. Izdatel'stvo
Akademii Nauk SSSR, Moskva. Israel Prog. for Sci.
Translations,Jerusalem.455pp. (TranslatedfromRussian.)
KRorr,P. 1962. Beitrage zur kenntnisdes Alpenbars Ursus
arctos L. 1758. SaugetierkundlicheMitteilungen10:1-35.
(Translatedfrom German.)
KRUUK,H. 1972. The spottedhyena: a study of predationand
social behavior. Univ. Chicago Press, Chicago, Ill. 335pp.
B. 1955. Sex dimorphismand size trendsin the cave
KURTEN,
bear, UrsusspelaeusRosenmullerandHeinroth. Acta Zool.
Fennica 90. 48pp.
1958. Life and death of the Pleistocene cave bear.
Acta Zool. Fennica95. 59pp.
1964. The evolutionof the polarbearUrsusmaritimus
Phipps. Acta Zool. Fennica 108. 30pp.
1967. The Pleistocene bears of North America 2.
GenusArctodus,short-facedbears. Acta Zool. Fennica 115.
60pp.
.1968. Pleistocene mammalsof Europe. Aldine Publ.
Co., Chicago, Ill. 317pp.
1976. The cave bear story. ColumbiaUniv. Press,
New York, N.Y. 163pp.
, AND E. ANDERSON. 1974. Association of Ursusarctos
and Arctodus simus (Mammalia: Ursidae) in the late
Pleistocene of Wyoming. Brevoria426:1-6.
, AND
.
1980. Pleistocene mammals of North
America. ColumbiaUniv. Press, New York, N.Y. 448pp.
P.E. 1995. Diet andco-ecology of PleistoceneshortMATHEUS,
faced bearsandbrownbearsin easternBeringia. Quaternary
Res. 44:447-453.
D.J. 1997. Use of ungulatesby Yellowstone grizzly
MATrSON,
bears Ursusarctos. Biol. Conserv. 51:161-177.
ANDR.R. KNIGHT.1991. Food habits
, B.M. BLANCHARD,
of Yellowstone grizzly bears, 1977-87. Can. J. Zool.
69:1619-1629.
MCDONALD,J.N. 1984. The recordedNorthAmericanselection
regime and Late Quaternarymegafaunalextinctions. Pages
404-439 in P.S. Martinand R.G. Klein, eds. Quaternary
extinctions: a prehistoricrevolution. Univ. ArizonaPress,
Tucson.
MCNAB,B.K. 1986. The influence of food habits on the
ecological energeticsof eutherianmammals. Ecol. Monogr.
56:1-19.
MEALEY,S.P. 1975. The naturalfood habits of free ranging
grizzly bearsin Yellowstone NationalPark,1973-74. M.S.
Thesis, Mont. State Univ., Bozeman. 158pp.
MooRE,W.J. 1965. Masticatoryfunctionand skull growth. J.
Zool. (Lond.) 146:123-131.
Moss, M.L. 1968. Functionalcranialanalysis of mammalian
mandibularramalmorphology. Acta Anat. 71:423-447.
NELSON,R.A., G.E. FOLK,JR., E.W. PFEIFFER,J.J. CRAIGHEAD,
Behavior,
C.J. JONKEL, AND D.L. STEIGER. 1983.
biochemistry, and hibernationin black, grizzly and polar
DIETANDMORPHOLOGY
OFNORTHERN
BEARS* Mattson
bears. Int. Conf. Bear Res. and Manage. 5:284-290.
Yu.B. PUKINSKY,
E.K.
NovIKOv,G.A., A.E. AIRAPET'YANTS,
ANDI.M. FOKIN.1969. Nekotorye osobennosti
TIMOFEEVA,
populyatsii burykh medvedey Leningradskoi oblasti.
Zoologicheskii zhurnal 68:885-900. (Translated from
Russian.)
S.I. 1931. Mammals of easternEurope and northern
OGNEV,
Asia. Vol. 2. Carnivora (Fissipedia). GlavnaukaGosudarstvennoeIzdatel'stvo, Moskva. IsraelProg.for Sci.
Translations,Jerusalem.590pp. (TranslatedfromRussian.)
OWEN-SMITH, N.
1987. Pleistocene extinctions:
the pivotal
role of megaherbivores. Paleobiology 13:351-362.
M.R. 1982. Black bear. Pages 504-514 in J.A.
PELTON,
ChapmanandG.A. Feldhamer,eds. Wild mammalsof North
America. Johns HopkinsUniv. Press, Baltimore,Md.
R.I. 1933. The black and brownbears of Europeand
POcOCK,
Asia. Part II. The sloth bear (Melursus), the Himalayan
black bear (Selenarctos),and the Malayanbear(Helarctos).
J. Bombay Nat. Hist. Soc. 36:101-138.
493
H. JINCHU.
1989. The feeding ecology of giant pandasand
the Asiatic black bears in the Tangjiahe Reserve, China.
Pages 212-241 in J. L. Gittleman,ed. Carnivorebehavior,
ecology, and evolution. Corell Univ. Press, Ithaca,N.Y.
SHIELDS,G.F., AND T.D. KOCHER. 1991.
Phylogenetic
relationshipsof NorthAmericanursidsbased on analysisof
mitochondrialDNA. Evolution 45:218-221.
L.M. 1995. Morphological correlates of dietary
SPENCER,
resourcepartitioningin the African Bovidae. J. Mammal.
76:448-471.
1990. Factors affecting the
STIRLING,
I., ANDA.E. DEROCHER.
evolution and behavioralecology of the moder bears. Int.
Conf. Bear Res. and Manage. 8:189-204.
R. DMI'EL, D. BAHARAV,
ANDA.
TAYLOR,C.R., A. SHKOLNIK,
The biomechanicalprinciplesrealized in the upperjaw of
long-snoutedprimates. Pages 249-264 in J.G.Else andP.C.
Lee, eds. Primateevolution.Vol. 1. CambridgeUniv. Press,
Cambridge,U.K.
L.B. 1981a. Evolution of skull shape in carnivores
RADINSKY,
2. Additionalmoder carnivores.Biol. J. Linn.Soc. 15:369388.
. 198 b. Evolution of skull shape in carnivores 1.
Representative modern carnivores. Biol. J. Linn. Soc.
15:369-388.
1985. Approachesin evolutionarymorphology: a
searchfor patterns. Ann. Rev. Ecol. Syst. 16:1-14.
BORUT.1974. Running in cheetahs, gazelles, and goats:
energycost andlimbconfiguration.Am. J. Physiol.227:848850.
THOMPSON,
R.S., ANDJ.I. MEAD. 1982. Late Quaternary
environments and biogeography of the Great Basin.
QuaternaryRes. 17:39-55.
VANVALKENBURGH,
B. 1985. Locomotordiversitywithin past
andpresentguildsof largepredatorymammals.Paleobiology
11:406-428.
1988. Trophicdiversitywithinpast andpresentguilds
of large predatorymammals. Paleobiology 14:155-173.
. 1989. Carnivoredentaladaptationsand diet: a study
of trophic diversity within guilds. Pages 410-436 in J.L.
Gittleman,ed. Carnivorebehavior,ecology, and evolution.
Comell Univ. Press, Ithaca,N.Y.
A.F. 1985. Empiricalrelationshipsbetween predator
VEZINA,
and prey size among terrestrial vertebrate predators.
Oecologia 67:555-565.
RAMSAY,M.A., AND I. STIRLING. 1988. Reproductive biology
VOORHIES,
M.R., ANDR.G. CORNER. 1986.
and ecology of female polar bears (Ursus maritimus). J.
Zool. (Lond.) 214:601-634.
R. 1951. Notes on the NuamiutEskimo andmammals
RAUSCH,
of the AnakturukPass region,BrooksRange,Alaska. Arctic
4:147-195.
1953. On the statusof some arcticmammals. Arctic
6:91-148.
M.L. 1968. The strategy of body size in
ROSENZWEIG,
mammaliancarnivores. Am. Midi. Nat. 80:299-315.
SCHALLER,G.B., H. JINCHU,P. WENSHI, AND Z. JING. 1985. The
giant pandasof Wolong. Univ. Chicago Press, Chicago, Ill.
298pp.
Arctodusas a potentialbreakerandflakerof late Pleistocene
megafaunalremains. CurrentRes. in the Pleistocene 3:4951.
D.E. 1976. Cranialvariationin polarbears. Int. Conf.
WILSON,
Bear Res. and Manage. 3:447-453.
PREUSCHOFT,
H., B. DEMES,M. MEYER, ANDH.F. BAR. 1986.
W. XIAOMING,
, T. QITAO,K.G. JOHNSON,
S. HEMING,
AND
The giant bear
W.C. 1989. Classification of recent Carnivora.
WOZENCRAFT,
Pages 569-593 in J.L. Gittleman,ed. Carnivorebehavior,
ecology, and evolution. CornellUniv. Press, Ithaca,N.Y.
YONEDA,M., ANDH. ABE. 1975.
Sexual dimorphism and
geographicvariationin the skullof theEzo brownbear(Ursus
arctosyesoensis). MemoirsFacultyAgric. HokkaidoUniv.
9:265-276. (In Japanesewith English summary.)
494
Ursus 10:1998
Appendix. 1. Untransformedvalues for morphometric variables used in principal components analysis of northern bear
skulls. Acronyms are explained in Fig. 1 and the text; sources are given in footnotes.
Species
Giant pandaa
Asiatic black bearb
Americanblack bearc
Brown bear-Europeand
Brown bear-continentale
Brown bear-coastalf
Cave bearg
Polar bearh
Short-facedbear1
Bodymass
(kg)
CBL
(mm)
CH
(mm)
FL
(mm)
ML
(mm)
MH
(mm)
ZW
(mm)
NW
(mm)
DENTL
(mm)
DENTA
(mm2)
125
86
101
141
148
277
248
288
335
249
212
257
307
324
377
430
331
380
114
80
90
107
123
132
168
112
160
80
59
78
108
110
130
153
103
125
174
155
170
221
233
271
301
222
274
102
68
80
92
110
132
146
96
137
197
136
149
193
207
252
292
199
289
42
30
42
46
49
57
68
52
88
75
57
57
67
75
82
97
60
86
1613
707
768
1145
1288
1420
2194
790
1891
a Davis (1964).
b
Ognev (1931), Pocock (1933), Bromlei (1965).
c Erdbrink
(1953a), Davis (1964), Ericksonet al. (1964), Bunnell and Tait (1981), Pelton (1982).
d Ognev (1931), Erdbrink(1953b), Kurt6n(1955, 1964), Krott(1962), Novikov et al. (1969).
e Ognev (1931), Rausch (1951, 1953), Davis (1964), Yoneda and Abe (1975).
f Ognev (1931), Rausch (1953), Davis (1964), Yoneda and Abe (1975).
g Kurt6n(1955, 1958, 1964, 1967, 1976).
h Ognev (1931), Kurt6n(1955, 1964), Ramsay and Stirling (1988).
1 Kurt6n(1967).
Appendix2. Sources of food habits studies used in analysis
of dietary differences among northern bears.
H. 1990. Blackbearsin theNorthForkof the Flathead
CARRILES,
AoI, T. 1985. Seasonal change in food habits of Ezo brown
bear (Ursus arctos yesoensis Lydekker) in northern
Hokkaido. Res. Bull. Coll. Exp. For.42:721-732, Hokkaido
Univ., Horonobe-cho,Japan.
ARNER,D.H. 1948. Fall food of the blackbearin Pennsylvania.
Penn. Game News:13.
1989. Final reportEast Front
AUNE,K., ANDW. KASWORM.
grizzly studies. Mont. Dep. Fish, Wildl., andParks,Helena.
332pp.
L.E., AND M.R. PELTON.1980. Seasonal foods and
BEEMAN,
feeding ecology of black bears in the Smoky Mountains.
Int. Conf. Bear Res. and Manage.4:141-147.
River Valley, Montana. M.S. Thesis, Univ. Montana,
Missoula. 164pp.
E.F. 1950. Bear-moose relationshipson the Kenai
CHATELAIN,
Peninsula. Trans.NorthAm. Wildl. Conf. 15:224-234.
CICNJAK,L., D. HUBER,H.U. ROTH,R.L. RUFF,ANDZ. VINOVRSKI.
1987. Food habitsof brownbearsin PlitviceLakesNational
Park,Yugoslavia. Int. Conf. BearRes. andManage.7:221226.
W.K. 1957. Seasonal food habits of the Kodiakbear.
CLARK,
Trans.North Am. Wildl. Conf. 22:145-151.
1939. Notes on
ANDT.E. CLARKE.
C., A.L. NELSON,
COTTAM,
in George
black
bear
of
the
winter
food
habits
early
WashingtonNationalForest. J. Mammal.20:310-314.
BENNETT,L.J., P.F. ENGLISH,AND R.L. WATTS. 1943. The food
habitsof the blackbearin Pennsylvania.J.Mammal.24:2531.
BERDUCOU,
C., L. FALIU,AND J. BARRAT. 1983. The food habits
of thebrownbearin the nationalparkof thewesternPyrenees
(France)as revealedby faeces analysis. Acta Zool. Fennica
174:153-156.
G.G. 1986. Availabilityanduse of foods by black
BERTAGNOLI,
bearsin Wisconsin. M.S. Thesis, Univ. Wisconsin, Stevens
Point. 42pp.
ANDJ. HUOT.1994. Food habitsof the
BOILEAU,
F., M. CRETE,
black bear, Ursus americanus,and habitatuse in Gasp6sie
Park,easternQuebec. Can. Field-Nat. 108:162-169.
F.G. 1965. Bears of the south far-easternUSSR.
BROMLEI,
Zdatel'stvo "Nauka",Moskva-Leningrad. 138pp. Indian
Natl. Sci. DocumentationCent., New Delhi. (Translated
from Russian.)
CRAIGHEAD,J.J., J.S. SUMNER,AND G.B. SCAGGS. 1982.
A
definitivesystemfor analysisof grizzlybearhabitatandother
wildernessresources. Wildl.-WildlandsInst. Monogr. No.
1, Missoula, Mont. 279pp.
EAGLE, T.C., AND M.R. PELTON. 1983.
Seasonal nutrition of
black bears in the GreatSmoky MountainsNational Park.
Int. Conf. Bear Res. and Manage.5:94-101.
ELGMORK,K., ANDJ. KAASA. 1992. Food habits and foraging of
the brown bear Ursus arctos in central south Norway.
Ecography 15:101-110.
N.P. 1986. Seasonal movements, habitatselection,
GARNER,
and food habits of black bears (Ursus americanus) in
ShenandoahNationalPark,Virginia. M.S. Thesis, Virginia
Poly. Inst. and State Univ., Blacksburg. 104pp.
D.M., ANDM. WHITE.1983. Black bearfood habitsin
GRABER,
Yosemite NationalPark. Int. Conf. Bear Res. andManage.
DIETANDMORPHOLOGY
OFNORTHERN
BEARS* Mattson
5:1-10.
GRENFELL,
W.E., JR.,ANDA.J. BRODY.Seasonal foods of black
bears in Tahoe National Forest,California. Calif. Fish and
Game 69:132-150.
1983. Ecological studies
EDITORS.
HAMER,D., ANDS. HERRERO,
of the grizzly bear in Banff National Park. Parks Canada
Contract WR 4-80. Univ. Calgary, Calgary, Alta, Can.
303pp.
, ANDK. BRADY.1991. Food and habitatused
by grizzly bears, Ursusarctos, along the continentaldivide
in WatertonLakes NationalPark,Alberta. Can. Field-Nat.
105:325-329.
D.F. 1972. Food habits of black bears in interior
HATLER,
Alaska. Can. Field-Nat. 86:17-31.
M.R.VAUGHAN.1988. Seasonalfood habits
HELLGREN,
E.C.,AND
of black bears in Great Dismal Swamp, Virginia-North
Carolina. Proc. Annu. Conf. Southeast. Assoc. Fish and
Wildl. Agencies 42:295-305.
HOLCROFT,
A.C., ANDS. HERRERO. 1991. Black bear, Ursus
americanus,food habitsin southwesternAlberta.Can.FieldNat. 105:335-345.
1985. Managingblack bear
IRWIN,
L.L., ANDF.M. HAMMOND.
habitats for food items in Wyoming. Wildl. Soc. Bull.
13:477-483.
JAMNICKY,J. 1988. The food of the brown bear (Ursus arctos
L.) in the Tatraregion. Folia Venatoria 18:197-213. (In
Czechoslovakianwith English summary.)
1991. Cabinet-Yaakecosystem
KASWORM,
W.F.,ANDT.J.THIER.
grizzly bear and black bear research 1990 progressreport.
U.S. Fish and Wildl. Serv., Missoula, Mont. 53pp.
K.C. 1986. Grizzly and black bear feeding ecology
KENDALL,
in Glacier National Park,Montana. U.S. Dep. Inter.,Natl.
ParkServ., West Glacier,Mont. 42pp.
LANDERS, J.L., R.J. HAMILTON, A.S. JOHNSON, AND R.L.
MARCHINTON.
1979. Foods and habitatof black bears in
southeasternNorth Carolina. J. Wildl. Manage. 43:143153.
ANDJ.R. WEGGE.1984. Black
LECOUNT,
A.L., R.H. SMITH,
bearhabitatrequirementsin centralArizona. ArizonaGame
and Fish Dep., Spec. Rep. No. 14, Tucson. 28pp.
A.G. 1989. Spring and summerfood habits of
MACHUTCHON,
black bears in the Pelly River Valley, Yukon. Northwest
Sci. 63:116-118.
MAEHR,
D.S., ANDJ.R. BRADY. 1984. Food habits of Florida
black bears. J. Wildl. Manage.48:230-235.
MATrSON,
D.J., B.M. BLANCHARD,ANDR.R. KNIGHT.1991. Food
habits of Yellowstone grizzly bears, 1977-1987. Can. J.
Zool. 69:1619-1629.
MCLELLAN,
B.N., ANDF.W. HOVEY.1995. The diet of grizzly
bearsin the FlatheadRiver drainageof southeasternBritish
Columbia. Can. J. Zool. 73:704-712.
AND D. TAIT. 1982. The grizzly
MILLER,S.J., N. BARICHELLO,
bears of the Mackenzie MountainsNorthwest Territories.
Northwest Territ. Wildl. Serv. Completion Rep. No. 3,
Yellowknife, Can. 118pp.
K.R.D. 1963. Ecology of the grizzlybear(Ursusarctos,
MUNDY,
L.) in GlacierNationalPark,BritishColumbia. M.S. Thesis,
495
Univ. Alberta,Edmonton,Can. 102pp.
A. 1981. The grizzlies of MountMcKinley. U.S. Dep.
MURIE,
Inter.,Natl. ParkServ. Sci. Monogr. Ser. No. 14. 251pp.
1978. Ecological studiesof the
NAGY,
J.A., ANDR.H. RUSSELL.
boreal forest grizzly bear (Ursus arctos L.): Annual report
for 1977. Can. Wildl. Serv., Ottawa,Can. 72pp.
ANDB.C.
M.C.S. KINGSLEY,
, A.M. PEARSON,
GOsKI.1983. Ecologicalstudiesof grizzlybearsin the Arctic
Mountains,northernYukon Territory,1972 to 1975. Can.
Wildl. Serv., Ottawa,Can. 104pp.
1983. A
, ANDC.B. LARSEN.
. ., .
study of the grizzly bears on the barren grounds of
Tuktoyaktuk Peninsula and Richards Island, Northwest
Territories,1974 to 1978. Can. Wildl. Serv., Ottawa,Can.
136pp.
G.A. 1956. Carnivorousmammalsof the fauna of
NOVIKOV,
the USSR. Izdatel'stvo Akademii Nauk SSSR, MoskvaLeningrad. Israel Prog. for Transl., Jerusalem. 238pp.
(Translatedfrom Russian.)
E.K. TIMOFEEVA,
, A.E. AIRAPET'YANTS,YU.B. PUKINSKY,
ANDI.M. FOKIN.1969. Nekotorye osobennostipopulyatsii
burykh medvedey Leningradskoioblasti. Zoologicheskii
Zhural 68:885-900. (In Russian.)
T. AOI,H. TORII,
T. ITO,ANDK. MAEDA.
NOZAKI,E., S. AZUMA,
1983. Food habits of Japaneseblack bear. Int. Conf. Bear
Res. and Manage. 5:106-109.
ODACHI, S., ANDT. AOI. 1987. Food habits of brown bears in
Hokkaido,Japan. Int. Conf. Bear Res. andManage.7:215220.
OSTI,F. 1979. Ulteriorecontributoallo studio delle abitudini
alimentari dell'Orso bruno delle Alpi (Ursus arctos L.).
Estrattoda NaturaAlpina, Ser. 2, 17:39-48. (In Italian.)
A.M. 1975. The northerninteriorgrizzly bear Ursus
PEARSON,
arctosL. Can. Wildl. Serv. Rep. Ser. No. 31. 86pp.
PELCHAT,
B.O., ANDR.L. RUFF. 1986. Habitat and spatial
relationshipsof black bears in boreal mixwood forest of
Alberta. Int. Conf. Bear Res. and Manage. 6:81-92.
PIEKIELEK,
W., ANDT.S. BURTON.1975. A blackbearpopulation
study in northernCalifornia. Calif. Fish and Game 61:425.
POELKER,
R.J., ANDH.D. HARTWELL.1973. Black bear of
Washington. WashingtonState Game Dep., Biol. Bull. No.
14, Olympia. 118pp.
E. 1986. Habitatselection in the brown bear in
PULLIAINEN,
easternFinland. Pages 113-115 in G.P. Contrerasand K.E.
Evans, compilers. Proceedings grizzly bear habitat
symposium. U.S. For. Serv. Gen. Tech. Rep. INT-207.
RAINE,
R.M., ANDJ.L. KANSAS.1990. Black bearseasonalfood
habitsand distributionby elevation in Banff NationalPark,
Alberta. Int. Conf. Bear Res. and Manage. 8:297-304.
RAMSAY,M.A., ANDK.A. HOBSON.1991. Polar bears make
little use of terrestrialfood webs: evidence from stablecarbonisotope analysis. Oecologia 86:598-600.
RAZUMOVSKY,B.I. 1966. Bears (Ursus arctos arctos L.) of the
Maly-Sheksnin interfluvium. Zoologicheskii Zhurnal
45:725-729. (In Russian with English summary.)
.1991. Ecology of the Asiatic black bear (Ursus
496
Ursus 10:1998
thibetanus)in Sichuan,China. Mammalia55:221-237.
K. LINDSAY,S. RUSSELL,
F.
RUSSELL,
D., B. VANDRIMMELEN,
ANDB. SMITH.
1974. Projectreportongrizzly bearSCHWAB,
mountain goat investigations in Knight Inlet, B.C. B.C.
Wildl. Branch,Victoria,Can. 90pp.
ANDG. ANDERSON.
N.A. WOODY,
RUSSELL,
R.H., J.W. NOLAN,
1979. A studyof the grizzly bear(UrsusarctosL.) in Jasper
National Park, 1975-1978 final report. Can. Wildl. Serv.,
Ottawa,Can. 102pp.
SCHALLER,G.B. 1968. Food habits of the Himalayan black
bear (Selenarctos thibetanus)in the Dachigam Sanctuary,
Kashmir. J. Bombay Nat. Hist. Soc. 66:156-159.
1994. Ungulatesin northwestTibet.
, ANDG. BINYUAN.
Natl. Geog. Res. and Exploration10:267-293.
bears and black bears of the Apgar Mountainsin Glacier
National Park, Montana. M.S. Thesis, Univ. Montana,
Missoula. 133pp.
1976. On the
SHARAFUTDINOV,
I.Yu., ANDA.M. KOROTKOV.
ecology of the brownbearin the southernUrals. Int. Conf.
Bear Res. and Manage. 3:309-311.
A.A. 1976. The European brown bear in the
SLOBODYAN,
Carpathians.Int. Conf. Bear Res. and Manage.3:313-319.
SPENCER,
H.E., JR. 1966. The blackbearandits statusin Maine.
State of Maine, Dep. Inland Fish. and Game, Game Div.
Bull. No. 4, Augusta. 55pp.
STELMOCK,
J.J.,ANDF.C. DEAN.1986. Brown bearactivity and
habitatuse, Denali National Park 1980. Int. Conf. Bear
Res. and Manage. 6:155-168.
Y. 1979. The food habit of the Japaneseblack bears
, L. HONG,TALIPU,L. HUA, R. JUNRANG,
Q. MINGJIANG, TAKADA,
in the Centralmountainarea,Naganoperfecture.J. Mammal.
ANDW. HAIBIN.1987. Status of large mammals in the
Soc. Japan8:40-53. (In Japanesewith English summary.)
TaxkorganReserve,Xinjiang,China. Biol. Conserv.42:5371.
TISCH,E.L. 1961. Seasonalfood habitsof the black bearin the
Whitefish Range of northwesternMontana. M.S. Thesis,
ANDZ. JING.1985. The giant
P. WENSHI,
,H. JINCHU,
MontanaState Univ., Bozeman. 108pp.
Ill.
of
Univ.
Chicago Press, Chicago,
pandas Wolong.
S.K. 1965. K biologii burogomedvedyaVostochnoi
USTINOV,
298pp.
AND
Sibiri. Pages 229-234 in Okhotnich'e-proryslovye zveri:
S. HEMING,
W. XIAOMING,
, T. QITAO,K.G. JOHNSON,
1989. The feeding ecology of giant pandasand
H. JINCHU.
biologiya i khozyaistvennoe, No. 1. Rossel'kheizdat,
Moskva. (In Russian.)
China.
in
the
black
bears
Asiatic
Pages
TangjiaheReserve,
P. 1976. A studyof the brownbear(Ursusarctos)
212-241 in J.L.Gittleman,ed. Carnivorebehavior,ecology,
VALKENBURG,
in the proposednortheasternadditionto Mount McKinley
and evolution. CornellUniv. Press, Ithaca,N.Y.
NationalPark. M.S. Thesis, Univ. Alaska,Fairbanks.83pp.
1991. Interrelationship
SCHWARTZ,
C.C.,ANDA.W. FRANZMANN.
WILLEY,C.H. 1978. The Vermontblack bear. VermontFish
of blackbearsto moose andforest successionin the northern
and Game Dep., Montpelier. 73pp.
coniferousforest. Wildl. Monogr. 113. 58pp.
1972. The status of the brown
the
in
The
brown
bear
1972.
0.
ZUNINO,
F., AND S. HERRERO.
SEMENOV-TIAN-SHANSKII,
NationalPark, Italy, 1971.
in
Abruzzo
bear
13:98-102.
Zool.
Ser.
U.S.S.R.
arctos)
(Ursus
Reserve,
Aquilo
Lapland
Biol. Conserv. 4:263-272.
SERVHEEN,C. 1983. Grizzly bearfood habits,movements,and
1978. Blackbearmovements
ZYTARUK,
B., ANDD. CARTWRIGHT.
habitat selection in the Mission Mountains, Montana. J.
and food habits in New Brunswick. East. Black Bear
Wildl. Manage. 47:1026-1035.
Workshop4:227-240.
SHAFFER,S.C. 1971. Some ecological relationshipsof grizzly