TREES:INVESTIGATING
PATTERNSOF
HANGINGBEARS FROMPHYLOGENETIC
MACROEVOLUTION
JOHNL.GITTLEMAN,'
TN37996, USA
Biology,Universityof Tennessee, Knoxville,
Departmentof Ecologyand Evolutionary
Abstract: Phylogeneticinformationof the family Ursidaeis well resolvedandreadilyavailablefor investigatingmacroevolutionaryquestions. Using
completephylogeniesof the ursidsandrelatedterrestrialcarnivores,I investigatewhetherpatternsof body size andlife historyevolutionin bearsdiffer
fromothercarnivoreswith respectto cladogenesis,speciesrichness,andoverallphyletictrends.Largebody size in bearsis not relatedto theirphylogenetic history,in contrastto most othercarnivoretaxa;this may relateto bears'relativelyrecentevolutionaryhistoryor theirlargebody size, which is
flexible for utilizinglow-qualityfoods, thusbufferingenvironmentalchange. Also, ratesof body size evolutionin bearsareaverageor perhapsslightly
slowerthanothercarnivores.Certainlife historytraits(birthweight, age eyes open, inter-birthinterval,longevity)arevery differentin bearsrelativeto
othercarnivores,even afteraccountingfor body size and phylogeny. In general,largebody size, flexibility in phyletic change in size, and slow life
historiesof ursidsmaybe an effectiveevolutionarystrategyfor dealingwithrecentenvironmentalstresses.
Ursus 11:29-40
Key words: bears,body size, life history,macroevolution,Ursidae
Bearsareevolutionarilyrelatedto terrestrialcarnivores
including canids, mustelids, hyaenas, and felids. Systematically, ursids are separatedfrom these other taxa
by: (1) unspecialized incisors and molars with broad,
flat, tuberculatedcrowns; (2) the shearing function of
carassials in ancestraltaxa largely replacedby a crushing function; (3) large and powerful legs, with plantigradeposture;(4) an omnivorousdiet, with most species
leaning towardfrugivoryandfolivory; and, (5) most distinctive of all, a gigantic body. Clearly, these features
are not only useful for systematics,they also elicit interesting ecological andevolutionaryquestions. Does large
body size in bearscontributeto higherextinction(orlower
speciation)rates relative to other carnivores? Are bears
hampered reproductively with respect to life histories
becauseof the lengthof time it takesto grow large? These
questions are comparativein nature,specifically requiring detailed informationabout relative differences with
other carnivore species and knowledge of evolutionary
history.
Fortunately,we now have considerable information
frombasic naturalhistoryfield studiesand,morerecently,
fromphylogeneticsystematics. Armedwithphylogenetic
trees, we can examine the timing and patternof macroevolutionary changes among bears and closely related
lineages. In this paper, I present findings from recent
studies comparingtrendsin bearsrelative to othercarnivores with respectto: (1) phylogenetictrees, (2) macroevolutionarytrendsof speciationand body size changes
using these trees, and (3) variationin life historypaterns.
My aim is to show how bears differ from other carnivores as well as to illustrateby example how mod ana-
lytical techniquescan be used to test macroevolutionary
hypotheses using phylogenies.
Two caveatsareneeded. Manyof the patternsdetected
here are preliminary;validity will rest with more and
betterdata. Also, emphasisis placed on patternsof macroevolutionarydifferencesbetween bears and other carnivores ratherthan explanations for these differences.
Futurework, especially at the populationlevel, will sort
out why bears are unique.
URSIDSYSTEMATICS
A BriefHistoryof Taxonomy
Reviews of ursid systematics are found elsewhere
(Simpson 1945; Ewer 1973; Stains 1984; Flynn et al.
1988; Wayne et al. 1989; Wozencraft 1989a,b; Flynn
1996). Excluding the giant (Ailuropodamelanoleuca)
andred (Ailurusfulgens)pandas,the familyUrsidaecomprises the following species:
Helarctos malayanus,Malayan sun bear
Melursus ursinus, sloth bear
Tremarctosornatus, Andean (spectacled) bear
Ursus americanus,Americanblack bear
Ursus arctos, brown bear
Ursus maritimus,polar bear
Ursus thibetanus,Asiatic black bear
Relative to other carnivorefamilies, the taxonomy of
bears is uncontroversial. Ursidae is one of the most recent taxonomicunits for familial designationin the class
Mammalia (Simpson 1945), the last family of arctoids
to appearin the fossil record (Stains 1984), and one of
1Presentaddress:
Departmentof Biology, GilmerHall, Universityof Virginia,Charlottesville,VA22903-2477, USA, e-mail:
JLGittleman
@Virginia.edu
30
Ursus 11:1999
the most homogeneousCarnivoranfamilies, as reflected
by minimal subfamilial (i.e., Tremarctinae)levels (see
Wozencraft1989b, 1993). Populationvariancein many
ursids is significant and continues to be assessed, particularly in the brown bear and polar bear (see Wayne
and Koepfli 1996). Indeed, Kurten and Anderson
(1980:184) exclaimed in referenceto brown and grizzly
bear systematics: "Some 232 Recent and 39 fossil "species" and"subspecies"(list in Erdbrink,1953) have been
proposedforthistaxon-a wasteof systematiceffortwhich,
as far as we know, is unparalled.".Fortunately,with the
exceptionof some genericnames (see above species listing; also, Zhang and Ryder 1993), species identification
is stable;therefore,the presentanalysis is not hampered
by species uncertainty. Of course, there is debate about
taxonomicplacementof the pandas,especially the giant
panda,which sharesmany features(herbivorous,plantigrade and gigantic) with ursids (O'Brien et al. 1985,
1991; Schaller et al. 1985; Mayr 1986.) Both familial
and ordinalanalyses (e.g., Wyss and Flynn 1993, Zhang
and Ryder 1993, Vranaet al. 1994, Bininda-Emondset
al. 1999) stronglyindicatethatthe giant pandais closely
relatedto the bears. The red pandais more of an enigma,
as it arguably can be placed in its own family, in the
Procyonidae or in the Ursidae (Roberts and Gittleman
1984, Bininda-Emondset al. 1999). The comparative
analyses presentedhere will consider the red pandaas a
procyonidbecauseof the manyfunctionalcharacterssuggestingthis placementandto not obscureobservedtrends
between the ursids and other carnivores.
Phylogenies:Molecularand
Morphological
Due to the relatively small numberof taxa, accessibility of samplematerial,the extendedtime thatbearshave
evolved, and a fairly complete fossil record,ursidshave
receiveda greatdeal of phylogeneticstudy(Martin1989,
Wayne et al. 1989). At least 28 phylogenetic trees are
availablefor ursids (Table 1), which is a relatively high
numbercomparedto other carnivorefamilies, considering the numberof species. Further,many differenttypes
of characters,including morphological,molecular, and
behavioral and ecological elements, have been used in
phylogenetic studies (Wayne et al. 1989, Wozencraft
1989a, Bininda-Emondset al. 1999). As for all organisms (Page and Holmes 1998), molecular phylogenies
recently have become plentiful for ursids; for example,
new phylogenies for the entirefamily are availablefrom
cytochromeb gene (Zhang and Ryder 1993), combined
genes (Talbot and
cytochrome b/tRNAProand tRNAThr
Shields 1996), and partialsequence informationfrom 6
regions of mtDNA (Waitset al. 1999). In general, con-
Table 1. Indices relating to the distribution of taxonomic
coverage for and the resolution of a composite tree of the
Carnivora (from Bininda-Emonds et al. 1999). The
parenthetical value of percent resolution for the herpestids
refersto when safe taxonomic reduction(Wilkinson1995)was
used to improvethe resolution of this family
Number
of
Numberof
Taxon
Percent Elements/
sourcetree/
source elements resolution taxon
taxon
trees
Highergroups
Mustelidae
62
202
100
16.8
0.27
30
155
72.7
3.4
0.11
Otariidae
15
46
69.2
3.3
0.22
118
94.4
6.2
0.30
Phocidae
21
Ursidae
28
50
85.7
6.2
0.22
Canidae
36
180
69.7
5.3
0.15
Felidae
40
282
97.1
7.8
0.20
Hyaenidae
6
8
66.7
2.0
0.33
Herpestidae
9
53
1.4
0.16
Viverridae
9
90
27.8
(58.3)
97
2.6
0.29
gruence among phylogenies places the giant panda and
Andean(spectacled)bear as initial lineages in ursidevolution, the sloth bear as an ancient monotypic lineage,
and the black and polar bears as recent taxa. Also,
throughoutthe family,thereis completeresolutionamong
nodes, a major improvementfrom one of the first molecular phylogenies (see Wayneet al. 1989). Of course,
with the many phylogenies available for the ursids and
no clear methodsfor how to weight some characters(or
phylogenies) against others, it is important to assess
agreementamong all the availablephylogenetichypotheses.
A CompletePhylogeny
Bininda-Emondset al. (1999) assembled a complete
phylogeny for all 271 species of Carnivora,derivedfrom
166 phylogenies using matrixrepresentationwith parsimony analysis (Baum 1992, Ragan 1992). Essentially,
matrix representationpermits combinationof phylogenetic treesderivedfromvariousdatasources,even though
the datamay not be congruent;differentphylogenies and
elements are given equal weighting.
Most source trees are availablefor "higher"carnivore
taxa; ursids have the fourthgreatestnumberat the family level behind felids, canids, and mustelids (Table 1).
The groupsthat have been studiedmost intensively also
tend to have the most elements used for study,although
the ursids lag behind in this respect. A possible explanation for this trend is that postcranialmaterialis less
accessible for such large-bodiedanimalsas bears,due to
the difficulties in transportingand housing them in museum collections. Ursids do, however, compare favor-
INVITED
PAPER* MACROEVOLUTION
* Gittleman 31
Ursus arctos, Brown
Ursus maritimus, Polar
Ursus thibetanus, Asiatic black
Helarctos malayanus, Sun
Melursus ursinus, Sloth
Ursus americanus, American black
Tremarctos ornatus, Andean
Ailuropoda melanoleuca, Giant panda
Fig. 1. A composite tree of the Ursidae (from Bininda-Emondset al. 1999). Numbers referto primarynodes supported when
mergingtrees.
ably with respect to the numberof elements studiedper
taxon and the number of elements per source tree per
taxon.
Whatdoes a composite bearphylogeny look like, how
resolved is it, and how robust? Generally, the tree is
fairly well resolved (Fig. 1), showing 86.7% resolution,
and is reasonablyrobustbased on Bremersupportvalues
for the nodes (Bininda-Emondset al. 1999). Details of
the tree indicate:
1. The giant panda and the Andean (spectacled)bear
are the oldest lineages, supportingrecent molecular results (Zhang and Ryder 1993, Talbot and
Shields 1996).
2. The position of the American black bear remains
unclear and merits further work, especially with
the monophyly of Ursus uncertain,as pointed out
by others (Goldmanet al. 1989, Zhang and Ryder
1994). This is an especially vexing result given
that non-monophylyis rare across the entire order
(e.g., Phoca, Vulpes,Lutra, Mustela, Leopardus,
Onciphelis).
3. Therearesistertaxonlinks betweenthe brownbearpolar bear and sun bear-sloth bear, also suggested
in other studies (Cronin et al. 1991, Talbot and
Shields 1996).
Clearly,the majorweakness in the composite phylogeny is the ancestral clade leading to the ursines; also,
elevation of the polar bear to its own genus does not appear appropriate.
MACROEVOLUTIONARY
TRENDS
Completespecies-level phylogeniespermitmacroevolutionarystudyhithertonot possible (Purvis 1996). Phylogenies arenecessaryfor any test involving evolutionary
history,but this informationis often unavailableor misleading from taxonomies. A complete phylogeny makes
tests of macroevolutionarypatternmorerobust,less likely
to producemistakessimply due to biasedtaxonomicsam-
pling, and accessible to questions about speciation and
rates of evolution. The following results are based on
the complete carnivorephylogeny of Bininda-Emondset
al. (1999; see Fig. 2) unless otherwise noted.
Speciation Rates
Regardingthe entire phylogeny of the carnivores,do
some lineages contain significantly more species than
others? The null expectationcomes from a model where
all extanttaxahave the same possibility to diversify (Nee
et al. 1995). The comparativetest identifies ancestral
lineages that have given rise to a disproportionatenumber of extant species in comparisonto their contemporaries (Purvis et al. 1995). The results show that 8
lineages containsignificantlymorespecies thanexpected
(Fig. 2). (It should be cautionedthat the radiationsare
not independent;a radiationis more likely to give rise to
more taxa.) The bears are not one of these, despite the
fact that all of the significant radiations lie in the
caniforms. It is interestingto speculatewhethersome of
the generalcharacteristicsunitingthis group(e.g., primitive bullar construction,basicranialarterialcirculation,
postcranial skeleton) has encouraged species richness
relative to the feliforms.
Body Size
The most common explanationfor the unequaldistribution of species among lineages rests with body size
(Gittlemanand Purvis 1998): small body size is associated with high diversity because of niche partitioning,
metabolic rate, reproductiverate, or brain size. Carnivores are a good groupto test this hypothesisin because
the range of body sizes is greaterthan any other mammalian order, spanning over 4 orders of magnitude
(Gittleman 1985). Morover, as shown, we know that
there are significant differences in species-richness
among lineages of the same age.
Using the complete species-level phylogenies of the
carnivores and the ursids, we can test whether larger-
..
32
..
.
Ursus 11:1999
-,,,
I
60
I
A, ,,'
2
I
20
40
Millionsof yearsbeforepresent
"' "
'"
2II
0
et al. 1999), including estimated times of
Fig. 2. A composite tree for all 271 species of Carnivora(from Bininda-Emonds
than expected.
extant
descendents
more
to
rise
have
which
significantly
as
given
lineages
divergence as well
INVITED
PAPER* MACROEVOLUTION
* Gittleman
33
'
4A
a
cr3
2-
31 agOoD
X
Mu
CA
0
0
u
0-2 I PI,C
0?i5a
N
Url
a
B
3
;b
a
?b
CA1
Carnivore contrasts
1 Andeanvs other 'truebears'
-2 r-u
2 Asiatic black bearvs brownandpolar
03O
-4
aE
3 Giantpandavs rest of bears
-6
0
1
2
3
4
Body size contrasts
5
4 Bears vs sistertaxa (mustelids,
procyoids,pinnipeds,andred panda)
Fig. 3. Clade size and body size contrasts in carnivores (Gittlemanand Purvis 1998). The comparison on the extreme right is
between Ailurusand the mustelid/procyonidclade.
bodied clades (i.e., bears) contain more species than
smaller-bodiedclades. Once body masses are calculated
for species withinclades,differencesbetweensisterclades
are used to search for changes in size (Gittleman and
Purvis 1998). Five sets of (clade) contrasts are useful
here: (1) contrastsacross all carnivores;(2) ursids versus sister taxa (mustelids, procyonids, pinnipeds, red
panda);(3) contrastsbetween Tremarctosand the other
'true' bears;(4) contrastsbetween Ursus thibetanusand
U. arctos-U. maritimus; and (5) contrasts between
Ailuropodaandall the otherbears. Overall,the caniform
carnivores (canids, procyonids, pinnipeds, ursids and
mustelids) do show a tendency for smaller bodied lineages to have more species (Gittlemanand Purvis 1998).
However,on closer inspectionthe contrastsinvolving the
bears do not reveal a consistent pattern(Fig. 3): clade
contrastsof 3 and 4, for example, both have low clade
size contrastsbut very differentsize contrasts. Although
body size is often implicated as a correlateof speciesrichness in mammals,much of the variationin diversity
cannotbe attributedto size differences.
PhylogeneticPattern
The above phylogenetic tests fail to show that cladogenesis in bears is unusual relative to other carnivore
taxa or that large body size is, as predicted,correlated
with species richness. An obvious explanation is that
body size in bears does not reflect phylogenetic history,
either within the ursids or relative to other carnivores.
Again, using a phylogeny,we can investigatethis lack of
pattern. For such a test, though, it is necessary to use
phylogenetic informationthat is totally independentof
the characterevolution we are tracing (i.e., body size),
otherwisethe test may not be independent(Gittlemanet
al. 1996). Taxonomic ranks (Wozencraft 1989b) are
therefore employed along with a statistic (Moran's I;
Gittlemanand Kot 1989) to assess whetherphylogenetic
change of body size in bears is unusualrelative to other
carnivores (data on carnivore body weight are from
Gittleman and Purvis 1998 and available from the author). The simple answeris yes. A 'correlogram'shows
that all carnivore families other than the ursids,
herpestids, and hyaenids show significant correlation
between taxonomicrankand body size (Fig. 4). In other
words, close phylogenetic relationshipin the 3 families
does not necessarily predictvariationin size.
If a taxonomy reflects phylogenetic relationshipsand
closely related taxa are phenotypicallymore alike, then
body size shouldcorrelatewith phylogeny(or in this case,
taxonomic ranks). Ursids are clearly very different in
this respect: along with hyenids and herpestids,ursids
show no phylogenetic patternin body size. An obvious
explanationfor this is small sample size. This may explain some of the lack of correlation,but not all, as the
procyonidsarea smallfamilyandshow significancewhile
the herpestids are a relatively large group and do not
show significant correlation.
34
Ursus 11:1999
Rates of Evolution
One mechanismfor the decouplingof body size evolution from phylogenetic distance lies in differentialrates
of evolution. If body size in ursidseitherchanges slowly
or increases and decreases at unequal rates, as Kurten
(1968) once proposed,then statisticaltests of phylogeny
andbody size will reveal differentpatternsthanthatproducedby a typicalgradualmodel of evolutionarychange.
One method for exploring this idea is to calculate rates
of evolutionary change in body size through phylogenetic time. We (Gittlemanet al. 1996) calculatedrates
of evolutionin Darwin's- quantitativechangein a character from its initial dimension to its final dimension
divided by the amountof time elapsed (Gingerich1993)
-for body size and othertraitsamong variousmammal
taxa and found that, in general, morphological traits
evolved at slower rates than ecological or life history
traits. Might therebe differencesin ratesof evolution of
body size change within carnivores? The general patternis a negative relationshipbetween ratesof evolution
and time (Fig. 5), as expected, given that measures of
ratechangedeclineover longerperiods(Gingerich1993).
Against this overall trendwe can see that the ursids appear to change slightly slower than the other carnivore
taxa. This suggests that perhaps the bears are indeed
adoptingan evolutionarystrategyof "lifein the slow lane"
(Macdonald1992), a classic K-selectionstrategyfor dealing with stable but stressful environments. Although it
has clearly been shown that the r-K selection theory is
not useful at populationlevel (Stears 1992), phylogenetic analyses are now showing that the theory may apply to significant divergences in size, life histories and
abundanceacross higher taxonomic levels (McKinney
and Gittleman 1995, McKinney 1997).
In summary,the gigantic size of bears is an obvious
distinguishing characteristiccompared to other terrestrialcarnivores. The above phylogenetictests show that,
despite unequal speciation rates across carnivore taxa,
the large body size of bears is unrelatedto species richness. Further,a simple test of correlationbetween phy-
10
z
0
-10
g
sbf
f
Fig. 4. Correlogram showing observed patterns of z scores (from Moran's /) for body weight across carnivore families.
Phylogenetic patternsreflectcorrelationof body weight (fromleftto right)among species withgenera, genera withinsubfamilies,
and subfamilies withinfamilies.
* Gittleman
PAPER* MACROEVOLUTION
INVITED
01
35
,
+ Other-Other
O Bear-Other
- Bear-Bear
100
+
+
+
:++
c102
-?10-
++++
-_
+
-
+
c +1 -
+
E10
+'-
oc
+
-
,~~~+t
10
10'o
1
1
'
'
'
' '
10
100
Time (MY)
Fig. 5. Log of evolutionary rate (darwins) versus log time interval (million years ago) observed across designated carnivores for
body weight.
logenetic distance and body size is not significant in the
bears, which is unusual for carnivores and mammals in
general. Therearemanypotentialreasonsfor these negative results, including the possibility that the phylogenetic information used here is wrong. One biological
explanationis that rates of evolutionarychange in body
size aredifferentin the bears;here, some supportis given
to this view. Also, the dramatic size increase in bear
body size may have surpassedexpectedchangegiven their
relativelyrecentevolutionaryhistory. Anothermore ecological explanation (Stirling and Derocher 1990) is that
the large size in bears confers an adaptivestrategy,lending flexibility to environmentalchange, which separates
bears from expected phylogenetic pattern. Clearly, the
massive within-species variability in body size, for example the 2-3-fold increase in total body mass of polar
bears during the spring foraging season (Cattet et al.
1997), is indication that body size evolution is unusual
in bears.
COMPARISONSWITHOTHER
CARNIVORES
Tinkeringwith Size, Life Histories, and
Phylogeny
Life history patterns in carnivores are extremely diverse (Gittleman1986, 1993). For example, simple comparison of the largest and smallest species is profound:
By the time a female polar bear reaches sexual maturity
at around5-6 yearsof age, a least weasel (Mustelanivalis)
would alreadyhave 56 descendants! Such a fundamental difference has significant effects on population dynamics, ecological adaptations,and evolutionarychange.
Comparisonsof mammalianlife history patternsinvolve
many traits such as gestation length, birth weight, interbirth interval, and longevity. By comparing single traits
across the carnivores,it is possible to isolate which variables are significantly differentin the bears, thus revealing relative macroevolutionaryshifts in the ursids and
36
Ursus 11:1999
a_0
.Giant
a Blackbear
^
5
panda
A
-
4
aPolar bear
./
6
At
-
bear
0 Brown
y
.
(a)
0: Canidae
'S
3
~
~
~
21^
*o^
E
S
2.
/^y'^r~~~~2:
.*^'C~~
C5
0:
Ursidae
A:
Ailurus
0:
Ailuropoda
4
Procyonidas
Mustelldae
Viverridae
:
/
/X
1t
/l:
Hyaenidae
-1
-2
-3
0
-1
Lifehistorytrait
Familycomparisons
(samplesize)
Gestationlength
Birthweight
2.02 (6,90)
Weaningage
1.11 (6,64)
Longevity
Age at sexual maturity
2.50 (6,48)*
Inter-birthinterval
2.64 (6,50)*
Age eyes open
Growthrate
7.01 (6,63)**
3.12 (6,61)**
0.80 (6,58)
A.: Felidae
1
4
Table 2. Comparative relations of life history patterns among
carnivore families after statistically removing effects of size
and taxonomy. *P < 0.05; **P < 0.001. All other tests are not
significant.
1
2
3
4
0.81 (3,36)
6
5
Logarithm of female body weight (kg)
/7"K-selected"
~375.~~
3.75,
y = .17x + 2.143, r2 .461
/
(b)
3.5Brown bear
A
(It
- 3.25-5
,-
0
Black bear
X
O
AA
//
sizesize
Polarbear
^
/,,'
?;..-::.
,
..
? -
-o
..,
. ?'*cladea
.-'cladea
b . .._
dcade
.
as
aGiant
*
panda
temporal traits
a 2.75.
._
/
X
lc2
' 2.5-
m
AXO
) 2.25-
-.":_~~~
E
/*?V~
2
O
/
O: Canidae
t0: Ursidae
A: Ailurus
: Ailuropoda
~ ~
~-
Y
t 1.75-
J~
*
~1.75
*/~+.:*
0 ODA A A
Procyonidas
X: Mustelidae
?: Viverridae
D
Hyaenidae
1J.5:
1
-4
5
0
3e
2
b
3
4
Logarithmof female body weight (kg)
5
6
littersize,
fecundity
etc.
Fig. 7. A model of clade diffusion with respect to size and
size-related traits (e.g., life histories).
Species in more
closely related clades such as the bears tend to be more
similar and reflect how increased size and life histories may
decouple when "stress-selected".
Gittleman (1995).
y
4-
X
GiantpandaG
lack bear C
B
i
l
Pollar bear
Bro\wn bear
x[
21
0?
C 2.5,
0D
O*A
A0
0
I
AA
O
0
a'
05
A
A
A
Asiatic black bear
0 Canidae
Ursidae
A Ailurus
O Ailuropoda
+ Procyonidae
X Mustelldae
* Viverrtdae
* Hyaenidae
A Felidae
0
.5
A
.
E
cm
o
.5
0
-4
0
-3
-2
-
1
After McKinney and
.132x + 2.945, r = .122
X 3SW2
3.5,
0.
'stress-selected"
age
E 3-
(c)
(C)
-
0
- 2- 3- 4- 5- 6-
1
Logarithm of female body weight (kg)
Fig. 6. Plots of residual values - following allometric and
phylogenetic autoregressive conversions - for carnivore life
history traits and female body weight. Included are (a) birth
weight, (b)inter-birth interval, and (c) age eyes open.
potential life history features that are critical for management and conservation decisions.
To perform systematic comparisons of life histories
across carnivores, it is necessary to account for 2 variables. First, body size clearly is involved: a least weasel
cannot give birth to young the size of a polar bear, nor
can a polar bear reproduceas often as a weasel. Thus,
comparisonsof life histories must initially include plotting the allometry of life history traits against body size
(typically female body weight, as we are interested in
reproductive traits), then using residuals for comparative tests (see Gittleman 1986, 1993). Second, as with
most traits (Harvey and Pagel 1991), life histories are
correlatedwith phylogeny, thus comparative tests may
be weakened by not using independentdata points. Indeed, in a study of 13 life history traitsacross carnivores,
each trait was significantly related to phylogeny
* Gittleman
INVITEDPAPER* MACROEVOLUTION
(Gittleman 1993). There are many ways to solve this
statistical problem (Purvis et al. 1994, Martins and
Hansen 1996). Here, I use an autoregressive method
because it is appropriatewhen using taxonomicrankinformationand sample sizes commensuratewith the carnivores. In sum, the following analyses are based on 2
sets of residuals(allometricandautoregressive),thusproviding conservative tests for which variables the bears
differ (or do not) from other carnivores. The carnivore
life history data, definitions of variables, and details of
methodology are in Gittleman(1986, 1993).
I examined 8 life historytraitsfor differencesbetween
bears and the remainingcarnivorefamilies (due to availability of data,most analyses only include brown,black,
and polar bears, and giant panda). Four traits showed
significant differences: birth weight, longevity, interbirthinterval,and age eyes first open (Table2). In each
of these the ursids lie at the extreme of bivariatedistributions: bears have comparativelysmall neonates (Fig.
6a), long inter-birthintervals (Fig. 6b), and late developmental ages for opening eyes (Fig. 6c). These findings generallyagree with previous work (e.g., Eisenberg
1981; Gittleman 1986, 1993), although greateremphasis now is placed on just how differentthe bearsare with
respect to some life histories, as a stringenttest here accounts for size and phylogeny.
At least 6 general factors are consideredto influence
such life historydifferences(forreviews see Boyce 1988,
Harvey et al. 1989, Gittleman 1994): body size, brain
size, metabolic rate, phylogeny, ecology, and mortality.
All of these may,to some extent,drivelife historiesamong
ursids. Along with the flexibility of body size evolution
in bears,as shown above, some life historiesappearto be
relatively decoupled from allometricexpectation. If we
consider both large body size and slow life histories a
responseto environmentalchange or stress (Fig. 7), then
a modelfor how these factorsaresimultaneouslyresponding may be applicable (McKinney and Gittleman 1995,
McKinney 1997). Some life history traits(birthweight,
age eyes open, inter-birth interval) are clearly much
slower in bears thanin othercarnivores. Environmental
stressessuch as fluctuatingfood supplyor declininghabitat availabilty may be sufficiently handled by reduced
reproductiverates duringthese periods.
IMPLICATIONS
FOR LONG-TERM
CONSERVATION
AND EVOLUTION
Extinction,the end of a phyletic line without replacement, has occurredthroughoutthe history of life and is
the ultimate destiny of every species. Many mammals
have gone throughrapidand dramaticextinctionevents.
37
This is particularlyevident in large predatoryspecies,
such as saber-toothedcats (Smilodon),which have had
repeated extinction events (Van Valkenburgh 1991,
1999). In opting for a large body size, the ursids seemingly have employed a safer strategy: herbivory,flexibility in diet and size, and protracted life histories.
Relative to other carnivores,this strategymay be working, at least over macroevolutionarytime scales. The
frequencyof extinction of ursids duringthe Pleistocene
appearsrelatively low comparedto other carnivoretaxa
(Kurten and Anderson 1980) and, at present, the only
bear species to receive endangeredclassification is the
giant panda. This is not a message for complacency.
Rather,macroevolutionarytrendssuggest that the ursids
have adopteda size andlife history strategyfor the longhaul. Bears, like all other animals,obviously have their
limits; many large, charasmatic mammals like mastodonts (Mammut),Irish elk (Megaloceros), and sabertooths are dramatic examples. We now need critical
informationto find out what precise environmentalfactors influence changes in size and life histories, the exact traitswhich makebearsdistinctfromothercarnivores
as well as vulnerableto extinction.
ACKNOWLEDGMENTS
I thankM. Pelton and G. Burghardtfor inviting me to
present this work at the meeting, T. Fuller and M.
Vaughanfor comments on the manuscript,and the Departmentof Ecology and EvolutionaryBiology, University of Tennessee,for support.
LITERATURE
CITED
BAUM,B.R.
1992. Combiningtreesas a way of combining
data sets for phylogeneticinference, and the desirabilityof
combininggenetrees. Taxon42:637-640.
ANDA. PURVIS.
BININDA-EMONDS,
1999.
O.R.P.,J.L. GITTLEMAN,
information:
Buildinglargetreesbycombining
phylogenetic
A completephylogenyof theextantCarnivora
(Mammalia).
BiologicalReviews74:143-175.
BOYCE,M.S., EDITOR. 1988. Evolution of life histories of
mammals.YaleUniversityPress,NewHaven,Connecticut,
USA.
CATTET,M.R.L., S.N. ATKINSON,S.C. POLISCHUK,ANDM.A.
RAMSAY.1997. Predictingbody mass in polar bears: is
morphometricsuseful? Journalof Wildlife Management
61:1083-1090.
G.W. GARNER,
CRONIN,
M.A., S.C. AMSTRUP,
ANDE.R. VYSE.
1991. Interspecificand intraspecificmitochrondrialDNA
variationin NorthAmericanblack bears(Ursus). Canadian
Journalof Zoology 69:2985-2992.
EISENBERG,J.F. 1981. The mammalianradiations. University
of Chicago Press, Chicago, Illinois, USA.
38
Ursus 11:1999
D.P. 1953. A review of fossil and recent bears of
ERDBRINK,
the Old World with remarkson their phylogeny based on
their dentition. Jan de Lange, Deventer,Netherlands.
R.F. 1973. The carnivores. Cornell University Press,
EWER,
Ithaca,New York,USA.
FLYNN,J.J. 1996. Camivoranphylogenyandratesof evolution:
morphological,taxic, andmolecular. Pages 542-581 in J.L.
Gittleman, editor. Carnivore behavior, ecology, and
evolution.Volume2. CornellUniversityPress,Ithaca,New
York,USA.
, N. NEFF, ANDR. TEDFORD.1988.
_
Phylogeny of the
Carnivora. Pages 73-116 in M.J. Benton, editor. The
phylogeny and classification of tetrapods. Volume 2.
ClarendonPress, Oxford, United Kingdom.
P.D. 1993. Quantification and comparison of
GINGERICH,
evolutionaryrates. AmericanJournalof Science 293A:453478.
GITTLEMAN,J.L. 1985. Carnivorebody size: ecological and
taxonomic correlates. Oecologia 67:540-554.
_ . 1986. Carnivore life history patterns: allometric,
phylogenetic, and ecological associations. American
Naturalist 127:744-771.
. 1993. Carnivore life histories: a reanalysis in the
light of new models. Pages 65-86 in N. Dunstone and M
Gorman,editors.Mammalsas predators.OxfordUniversity
Press, Oxford, United Kingdom.
. 1994. Are the pandas successful specialists or
evolutionaryfailures? BioScience 44:456-464.
Pages 536-568 in J.L. Gittleman, editor. Carnivore
behavior,ecology, andevolution. CornellUniversityPress,
Ithaca,New York,USA.
MARTINS,
E.P.,ANDT.F.HANSEN.1996. The statisticalanalysis
of interspecificdata: a reviewandevaluationof phylogenetic
comparativemethods. Pages 22-75 in E.P.Martins,editors.
Phylogeniesandthe comparativemethodin animalbehavior.
OxfordUniversity Press, New York,New York,USA.
MAYR,E. 1986. Uncertaintyin science: is the giant pandaa
bear or a raccoon? Nature323:769-771.
M.L. 1997. Extinctionvulnerabilityandselectivity:
McKINNEY,
combining ecological and paleontological views. Annual
Review of Ecology and Systematics28:495-516.
_
tinkeringwith covariationin life history, morphologyand
behaviour. Pages 21-47 in K.J. McNamara, editor.
Evolutionarychangeandheterochrony.JohnWiley & Sons,
Chichester,United Kingdom.
NEE, S., E.C. HOLMES, A. RAMBAUT, AND P.H. HARVEY.1995.
Inferringpopulation history from molecular phylogenies.
PhilosophicalTransactionsof the Royal Society of London
B 344:25-31.
W.G. NASH,J.S. MARTENSON,
O'BRIEN,S.J., R.E. BENVENISTE,
D.E. WILDT,M. BUSH,R.K. WAYNE,AND
M.A. EICHELBERGER,
D. GOLDMAN.1991. Molecular biology and evolutionary
theory: the giant panda'sclosest relatives. Pages 225-250
in L. Warrenand H. Koprowski,editors.New perspectives
on evolution. Wiley-Liss, New York,New York,USA.
1985.
, W.G. NASH,D.E. WILDT,ANDR.E. BENVENISTE.
, C.G. ANDERSON,M. KOT, AND H.K. LUH. 1996.
Comparativetests of evolutionarylability and rates using
molecular phylogenies. Pages 289-307 in P.H. Harvey,
A.J.L. Brown, S. Nee, and J. MaynardSmith, editors. New
uses for new phylogenies. OxfordUniversityPress,Oxford,
United Kingdom.
_
, ANDM. KOT. 1989. Adaptation: statistics and a null
model for estimating phylogenetic effects. Systematic
Zoology39:227-241.
1998. Body size and species-richness
_ , ANDA. PURVIS.
in carnivoresandprimates.Proceedingsof theRoyal Society
of LondonB 265:113-119.
D., P.R. GIRI,ANDS.J. O'BRIEN.1989. Molecular
GOLDMAN,
genetic-distanceestimates among Ursidae as indicatedby
one- andtwo-dimensionalproteinelectrophoresis.Evolution
43:282-295.
P.H., ANDM.D. PAGEL.1991. The comparative method
HARVEY,
in evolutionarybiology. Oxford University Press, Oxford,
United Kingdom.
1989. Life history
, A.F. READ,ANDD.E.L. PROMISLOW.
variation in placental mammals: unifying the data with
theory. OxfordSurveys in EvolutionaryBiology 6:13-31.
KURTEN, B. 1968. Pleistocene mammals of Europe.
Weidenfieldand Nicolsen, London, United Kingdom.
_
1980. Pleistocene mammals of North
ANDE. ANDERSON.
America. Columbia Univ. Press, New York, New York,
USA.
D. 1992. The velvet claw: a naturalhistory of
MACDONALD,
the carnivores. BBC Books, London, United Kingdom.
L.D. 1989. Fossil historyof the terrestrialCarnivora.
MARTIN,
1995. Ontogeny and phylogeny:
, ANDJ.L. GITTLEMAN.
A molecular solution to the riddle of the giant panda's
phylogeny. Nature317:140-144.
PAGE,R.D.M., ANDE.C. HOLMES.1998. Molecular evolution.
Blackwell, Oxford, United Kingdom.
A. 1996. Using interspecies phylogenies to test
PURVIS,
macroevolutionary hypotheses. Pages 153-168 in P.H.
Harvey, A.J.L. Brown, S. Nee, and J. Maynard Smith,
editors. New uses for new phylogenies. OxfordUniversity
Press, Oxford, United Kingdom.
?____,
AND H.-K. LUH. 1994.
J.L. GITTLEMAN,
Truth or
consequences: effects of phylogenetic accuracy on two
Theoretical
Journal of
comparative methods.
Biology167:293-300.
_ , S. NEE, ANDP.H. HARVEY. 1995. Macroevolutionary
inferencesfromprimatephylogeny.Proceedingsof theRoyal
Society of LondonB 260:329-333.
RAGAN,M.A.
1992. Phylogenetic inference based on matrix
representation of trees. Molecular Phylogenetics and
Evolution 1:53-58.
M.S., ANDJ.L. GITTLEMAN.1984. Ailurus fulgens.
ROBERTS,
MammalianSpecies Account 222.
P. WENSHI,ANDZ. JING. 1985. The
G.B., H. JINCHU,
SCHALLER,
giant pandas of Wolong. University of Chicago Press,
Chicago, Illinois, USA.
G.G.
SIMPSON,
1945. The principles of classification and a
classification of mammals. Bulletin of the American
Museum of NaturalHistory 85:1-350.
H.J. 1984. Carnivores.Pages 491-521 in S. Anderson
STAINS,
and J.K. Jones, Jr., editors. Ordersand families of recent
* Gittleman
PAPER* MACROEVOLUTION
INVITED
mammalsof the world. Wiley, New York,New York,USA.
S.C. 1992. The evolution of life histories. Oxford
STEARNS,
University Press, Oxford, United Kingdom.
1990. Factors affecting the
STIRLING,
I., ANDA.E. DEROCHER.
evolution and behavioral ecology of the modern bears.
InternationalConferenceof BearResearchandManagement
8:189-204.
TALBOT,S.L., ANDG.F. SHIELDS. 1996.
A phylogeny of the
bears (Ursidae) inferredfrom complete sequences of three
mitochrondrial genes. Molecular Phylogenetics and
Evolution 5:567-575.
VAN VALKENBURGH,B.
1991.
Iterative
evolution
. 1999.
Ithaca,New York,USA.
, R.E. BENVENISTE,D.N. JANCZEWSKI,AND S.J. O'BRIEN.
1989. Molecular and biochemical evolution of the
Carnivora. Pages 465-494 in J.L. Gittleman, editor.
Carnivore behavior, ecology, and evolution. Cornell
University Press, Ithaca,New York,USA.
M. 1995. Coping with abundantmissing entries
WILKINSON,
in phylogenetic inference using parsimony. Systematic
Biology 44:501-514.
WOZENCRAFT, W.C.
Major patterns in the history of carnivorous
mammals.Annual Review of Earthand PlanetarySciences
27:463-493.
ANDW.C.
J.R. POWELL,
VRANA,P.B., M.C. MILINKOVITCH,
WHEELER.
1994. Higher level relationshipsof the arctoid
Carnivorabased on sequence data and "total evidence".
MolecularPhylogenetics and Evolution 3:47-58.
S.J. O'BRIEN,ANDR.H. WARD. 1999.
WAITS,L.P., J. SULLIVAN,
Rapid radiationevents in the family Ursidae indicated by
likelihoodphylogeneticestimationfrom multiplefragments
of mtDNA. MolecularPhylogeneticsandEvolution. 13:8292.
WAYNE,R.K., ANDK.-P. KOEPFLI. 1996. Demographic and
historicaleffects on genetic variationof carnivores. Pages
453-484 in J.L. Gittleman, editor. Carnivore behavior,
ecology, andevolution.Volume2. Corell UniversityPress,
1989a.
The phylogeny
of the recent
Carnivora. Pages 495-535 in J.L. Gittleman, editor.
Carnivore behavior, ecology, and evolution. Cornell
University Press, Ithaca,New York,USA.
of
hypercarnivory in canids (Mammalia: Carnivora):
evolutionary interactions among sympatric predators.
Paleobiology 17:340-362.
39
___
. 1989b. Classification of the recent Carnivora. Pages
569-593 in J.L. Gittleman, editor. Carnivore behavior,
ecology, and evolution. Cornell University Press, Ithaca,
New York,USA.
. 1993. Order Carnivora. Pages 279-348 in D.E. Wilson
and D.M. Reeder,editors. Mammalsspecies of the world.
SmithsonianInstitutionPress, Washington,D.C., USA.
WYSS,A.R., ANDJ.J. FLYNN. 1993. A phylogenetic analysis
anddefinitionof the Carnivora.Pages 32-52 in F.S. Szalay,
M.J. Novacek and M.C. McKenna, editors. Mammal
Phylogeny. Springer-Verlag,New York,New York,USA.
ZHANG, Y.P., AND O.A. RYDER.
1993.
Mitochondrial DNA
sequence evolution in the Arctoidea. Proceedings of the
National Academy of Sciences 90:9557-9561.
, AND
.
1994. Phylogenetic relationships of bears
(Ursidae) inferred from mitochondrial DNA sequences.
MolecularPhylogenetics and Evolution 3:351-359.