Stabbing Slinkers: Tusk Evolution Among Artiodactyls

Abstract

Combat weaponry, including elaborate horns and antlers and complex dentition, evolved independently several times among mammals. While it is evident that tusk and tusk-like dentition have emerged primarily among males for intrasexual combat, it is unclear what ecological factors favor the retention or re-evolution of tusks. We investigated patterns of tusk evolution in artiodactyls while exploring specific ecological factors that might favor their use over other cranial weapons (e.g., antlers, horns). We show that among males, small (<15 kg), solitary species tend to retain well-developed canines, and more solitary species live in more closed habitats. These results suggest that tusks are a better weapon option for smaller, slinking artiodactyls in forested environments with low visibility, whereas larger taxa living in more open environment can bear the cost of elaborate headgear and are better served by communicating across distances an honest signal of fighting ability. Small species in dense habitats may also be more likely to be ambushed by predators and have a need to defend themselves; small, slicing daggers may be a better defensive weapon and allow more maneuverability and faster escape than cumbersome headgear in densely vegetated habitats.

Full text

ORIGINAL PAPER
Stabbing Slinkers: Tusk Evolution Among Artiodactyls
Doreen Cabrera
1
&Theodore Stankowich
1
#Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract
Combat weaponry, including elaborate horns and antlers and complex dentition, evolved independently several times among
mammals. While it is evident that tusk and tusk-like dentition have emerged primarily among males for intrasexual combat, it is
unclear what ecological factors favor the retention or re-evolution of tusks. We investigated patterns of tusk evolution in
artiodactyls while exploring specific ecological factors that might favor their use over other cranial weapons (e.g., antlers, horns).
We show that among males, small (<15 kg), solitary species tend to retain well-developed canines, and more solitary species live
in more closed habitats. These results suggest that tusks are a better weapon option for smaller, slinking artiodactyls in forested
environments with low visibility, whereas larger taxa living in more open environment can bear the cost of elaborate headgear
and are better served by communicating across distances an honest signal of fighting ability. Small species in dense habitats may
also be more likely to be ambushed by predators and have a need to defend themselves; small, slicing daggers may be a better
defensive weapon and allow more maneuverability and faster escape than cumbersome headgear in densely vegetated habitats.
Keywords Tusks .Canines .Weapon s .Defense .Deer
Introduction
Elaborate weaponry has evolved multiple times among mam-
mals (Emlen 2008; Stankowich 2012) in both extinct (giant
armadillo (Doedicurus)tailspikes,Smilodon saber teeth), and
extant species (elephant tusks, rhinoceros horns, bovid horns).
These structures are used for both prey capture (Smilodon
saber teeth) and sexual combat (deer antlers, beaked whale
tusks), but among extant species, sexually selected weaponry
is most prevalent among the artiodactyls in the form of cranial
appendages and elongated tusks. Most studies focus on un-
derstanding the ecological, social, and phylogenetic underpin-
nings of horns (Bovidae) and antlers (Cervidae) (Clutton-
Brock et al. 1980;Packer1983;Estes1991b;Lundrigan
1996;Caroetal.2003; Bro-Jørgensen 2007; Stankowich
and Caro 2009;Goss2012;), while the factors promoting the
evolution, retention, and elaboration of tusks have received
little attention (but see Geist 1971;Raiaetal.2015). Here,
we investigate the patterns of tusk evolution in artiodactyls
while exploring specific ecological factors that might favor
their use over cranial weapons (e.g., antlers, horns).
Animal weapons and ornaments typically help their bearers
gain greater access to reproductive partners, but each are un-
der distinct selective mechanisms (Rico-Guevara and Hurme
2018): weapons are a product of physical battle during male-
male competition while ornaments arise from indirect visual
competition between males seeking female preference. As
suggested by McCullough et al. (2016), structures that result
from male-male competition can be modelled using a weapon-
signal continuum where at one extreme, weapons like
Significance Statement While the function and evolution of tusks in
elephants, walruses, and even narwhals have received a great deal of
scientific and public attention, we know little about what drives the
evolution and maintenance of tusks in several groups of artiodactyls (e.g.,
pigs, muntjac, musk deer). Most male artiodactyls have some sort of
sexual weapon (e.g., antlers, horns, tusks), but we don’t know what
ecological factors promote the evolution of tusks in some and cranial
weapons in others. Using a comparative approach, we show that living a
slinking, solitary lifestyle in dense, closed habitats where long range
communication during sexual combat is not possible favors the evolution
of sharp, dagger-like tusks for combat during territorial disputes. We
discuss the sexual benefits of tusks over antlers considering species
ecology.
Electronic supplementary material The online version of this article
(https://doi.org/10.1007/s10914-018-9453-x) contains supplementary
material, which is available to authorized users.
*Theodore Stankowich
theodore.stankowich@csulb.edu
1
Department of Biological Sciences, California State University, 1250
Bellflower Blvd, Long Beach, CA 90840, USA
Journal of Mammalian Evolution
https://doi.org/10.1007/s10914-018-9453-x

artiodactyl tusks are used purely during combat and at the
other, signal structures are used solely to intimidate rival males
(e.g., stalk-eyed fly eye spans). At the center of the continuum,
however, are structures like deer antlers and bovid horns that
function to communicate strength before fights ensue.
Therefore, weapons may not only serve as honest signals of
fighting ability but also can inflict significant physical dam-
age. Here, we examine the ecological factors that favor the
transition between tusks as pure weapons and antlers/horns as
dual functioning structures.
The earliest most primitive artiodactyls, a group referred to
as the Dichobunoidea, consist of a rich record of extinct spe-
cies necessary for understanding the ancestral state and sub-
sequent evolution of Artiodactyla (Theodor et al. 2007).
Dental morphology of Diacodexis, the first documented artio-
dactyl that emerged in the early Eocene, and other
dichobunids suggest all had well-developed canine teeth, but
none possessed elongated tusks or headgear (Janis 1990). A
European radiation of artiodactyls lasting until the end of the
Eocene gave rise to several families, including
Cebochoeridae, a group of small- to medium-sized species
with caniniform first premolars and elongated canines useful
for grubbing, similar to extant pigs (Erfurt and Métais 2007).
Ruminants emerged during the middle Eocene as part of
the selenodont radiation in North America and Asia (Métais
and Vislobokova 2007). Early ruminants share several defin-
ing features, but unlike many extant artiodactyls, they lacked
cranial headgear (Gentry 1994). Modern forest-dwelling
tragulids (Tra gulus,Moschiola,Hyemoschus)morphological-
ly resemble these early taxa more than any other living rumi-
nants; however, due to a lack of fossil data, deep ruminant
phylogenetic relationships remain ambiguous. Evolutionary
history suggests that the earliest radiation of Ruminantia oc-
curred in conjunction to that of Tylopoda (camelids), while the
second radiation occurred in Central Asia’s early Oligocene
with the appearance of Pecora (bovids and cervids), tragulids
being the only primitive ruminant lineage to survive (Métais
and Vislobokova 2007). Extinct early ruminants are among
the smallest documented artiodactyls, and dental morphology
data indicate that primitive ruminants were herbivorous with
mostly browsing habits.
Climate change during the late Eocene promoted a shift in
biodiversity (Prothero 2017), and the sporadic availability of
high-quality foods in cold climates facilitated the radiation of
artiodactyl groups and the evolution of conspicuous structures
like large antlers, horns, and tusks (Geist 1966). Early
weapons were small and built to inflict maximum damage
(e.g., sharp horns and tusks), but structures would later in-
crease in size and complexity as they developed into status
symbols among males (Emlen 2008). Following the extinc-
tion of early ruminants, saber-toothed forms with duiker-like
bodies (Bslinkers^;Geist1998) arose in the Oligocene.
Slinkers were likely small (3–25 kg) and wove through dense
undergrowth with a body plan consisting of muscled haunches
and a long muscular back to aid in saltatory escape and loco-
motion. A number of extant species display the characteristic
Bduiker syndrome^that includes distinct weapons (tusks:
Tragulus,Hydropotes; short stabbing horns/antlers:
Cephalophus,Madoqua,Neotragus,Oreotragus,
Philantomba,Raphiceros,Sylvicapra,Tetracerus; or both:
Muntiacus) developed for stabbing combat, security, and for-
aging to support a solitary forest dwelling lifestyle (Barrette
1977;Geist1998). Dagger-like canines may provide slinkers
with the appropriate weapons to deliver quick pain inflicting
jabs as opposed to possessing elaborate cranial weapons that
increase their chances of entanglement. Barrette (1977)hy-
pothesized that the combination of small body size and the
need to live in closed habitats further promoted an inconspic-
uous slinker lifestyle dependent on quick cover when individ-
uals encounter threatening situations.
With the emergence of gregariousness and increased body
size (Brashares et al. 2000), weapons intended for maximizing
superficial damage would no longer be effective. Severe
wounds created by external protrusions might have attracted
predators to the group. Geist’s dispersal theory (1966,1971,
1974) suggested that as mammals evolved towards larger body
size, social ungulates could not afford to engage in life-
threatening altercations, creating selection against tusks. One
evolutionary solution would have been to reduce dagger-like
weapons and replace them with combat weaponry oriented to-
wards signaling fighting ability but also delivering less fatal
damage (Geist 1966). Specialized structures like horns and ant-
lers were a more effective fighting form because they visually
deterred competitors before engaging in costly fighting forms.
AsfamouslydiscussedbyJarman(1974), social behavior, as a
result, became increasingly prominent among grazers in open
landscapes, which have lost their combat teeth (non-tusked
cervids have either peg-like vestigial upper canines or lack them
entirely; bovids lack upper canines: Ungar 2010). Combat be-
tween males would have shifted from forced withdrawal by
quick blows to intimidation of the opponent into withdrawing
by wrestling with their newly developed headgear.
Parrying, the act of evading attacks with a counterblow,
likely initiated the development of deflecting structures in
the form of cranial bumps (Geist 1998). Usage of cranial
weaponry would have been significantly more successful at
delivering lasting pain to opponents and further promoted se-
lection for bone and dermal growths (horns and antlers).
Protruding dagger-like canines would have been no match
for head clubs and as a result were shortened to avoid detri-
mental blows. As slashing became less effective, intricate
headgear allowed parrying individuals to hold their opponents
head from striking with sharp canines. As used by extant
muntjacs (Barrette 1977), protoantlers in deer became the first
defensive cranial structures while canines remained the offen-
sive structure (Geist 1998).
J Mammal Evol

In this study, we examine the ecological factors that influ-
ence the evolution (retention, loss, and re-evolution) of tusks
in artiodactyls. We hypothesize that a solitary slinking lifestyle
promotes the elaboration and retention of tusks, while the
evolution of larger body sizes and movement into an open
habitat in larger groups favors the evolution of antlers and,
thus, the loss of tusks. We predict that species with smaller
body sizes, living in closed environments, and living a solitary
lifestyle would be more likely to have larger tusks.
Additionally, as body mass, habitat openness, and group size
increases, tusks should be lost. We use tusk measurements
from skulls of both sexes of a variety of artiodactyls and com-
parative phylogenetic analysis to test this hypothesis.
Methods
In our study, we considered all extant artiodactyl families and
included representatives (63 species) from each lineage bear-
ing tusks. Tusk measurements were taken from collections at
the Natural History Museum of Los Angeles County,
American Museum of Natural History, and the National
Museum of Natural History at the Smithsonian Institution
(Appendix). We measured ten specimens per species, five
female and five male skulls. Only adult skulls with intact
canines were used. We assumed that tusks on each side of
the jaw developed approximately symmetrically so we mea-
sured the most intact upper canine tooth and lower canine
tooth in each skull without regard for side. For all specimens,
we adopted six measurements used by Gittleman and Van
Valkenburgh (1997): upper canine height, upper canine
length, upper canine width, lower canine height, lower canine
length, and lower canine width. Canine height (H) was mea-
sured from the tip to the dentine-enamel junction. Uncurving
tusks were measured using digital calipers (mm). To compen-
sate for distortion caused by outwardly curving tusks (e.g.,
Suidae) when measuring height, we formed a flexible metal
wire from the tip to the dentine-enamel junction down the
midline of the lingual side of the tooth and then repeated this
measurement on the buccal side of the tooth; we then took the
average of these two measures as the total Bheight.^The
anteroposterior length (L) and the mediolateral breadth (W)
across the lingual sides of the upper and lower canines were
measured at the tooth base (dentine-enamel junction).
Measurements were averaged across specimens to obtain a
single representative value for each dimension (height, length,
width) for each sex for each species. We estimated tooth volume
using the formula for the volume of a four-sided pyramid:
Volume = (HLW)/3. We took measurements for both upper
and lower canines, for each species we selected the tooth (upper
vs. lower) with the largest average height and volume. For
groups with either small Bpeglike^canines, variably present
canines, or absent canines, we simply surveyed skulls in
museums and photographs of skulls online to assess their con-
dition, instead of actually measuring specific museum speci-
mens. Four species of cervids typically displayed Bpeglike^
upper canines in all individuals, but possessed incisiform lower
canines (Cervus albirostris, C. elaphus, C. nippon,
Hippocamelus antisensis). Upper measurements, in this case,
were scored as 1 mm in length, width, and height (volume =
0.33 mm
3
), but lower measurements were scored a volume of
0mm
3
. In five different cervid species, individuals varied in the
presence or absence of Bpeglike^upper canines (Mazama amer-
icana, M. gouazoubira, Ozotoceros bezoarticus, Rangifer
tarandus, Rusa unicolor); these species were scored as
0.5 mm in height, width, and length (volume = 0.04 mm
3
). In
species where tusks or tusk-like dentition was absent, all dimen-
sions were assigned a 0. Caniform canines were absent altogeth-
er for all species of Antilocapridae, Bovidae, and Giraffidae. All
measures of canine height and volume were transformed to
achieve normality assumptions using log
e
(X + 1) function.
For all species for which we had sufficient tooth measure-
ments, we collected a variety of natural history data from
published (Nowak 1999; Wilson and Mittermeier 2011)and
online (UMMZ 2015) sources. We collected body mass data
for each sex separately and log
e
transformed it to achieve
normality. If more than one source provided a mass, we used
the average of those values as our mean sex mass. Mating
strategies were coded as 0 if monogamous and 1 if polygy-
nous (Caro et al. 2003). Following Stankowich et al. (2014),
we used collected habitat use data from IUCN (2017; Version
3.1) and assigned an openness score (0–1) to each habitat
type: Temperate Forest (0.2), Tropical Forest (0.1), Savanna
(0.7), Temperate Shrubland (0.6), Tropical Shrubland (0.5),
Tundra (0.9), Temperate Grassland (0.8), Tropical Grassland
(0.8), Wetlands (0.3), Rocky (0.8), Desert (0.95), Marine
(0.8), Artificial Grassland (0.8), Urban (0.8), Artificial
Marine (0.8), and Caves/Subterranean (0.05). We edited the
IUCN habitat list to only include the primary two to three
habitats inhabited per species giving preference to those in
which species spend the most time (Stankowich and Caro
2009). Finally, we coded sociality as the typical group size
for species: (1) solitary or pairs, (2) 3–10, (3) 11–50, (4) >
50 (Caro et al. 2003). Blinded methods were not used in this
study as no behavioral data were recorded and/or analyzed;
other experimenters, however, collected different elements of
the dataset and compiled them together once completed. The
complete dataset can be found in the online Appendix.
To account for the effect of phylogenetic relatedness be-
tween species, we used phylogenetic general least squares
(PGLS) analyses to test for correlated evolution between traits
in R (R Core Team 2012). We used a composite Artiodactyla
tree downloaded from 10KTrees (Arnold et al. 2010) for all
analyses. To test for the effect of each natural history factor on
canine height and volume in each sex, we conducted PGLS
analyses using the ‘caper’package (Orme et al. 2012), where
JMammalEvol

lambda is computed using maximum likelihood methods.
Because group size was significantly correlated with each of
the other predictors, we analyzed its effects on tusk size sep-
arately; therefore, we tested each canine measure in two sep-
arate PGLS models: (1) with group size by itself and (2) with
habitat openness, body mass, and mating system combined in
a single model. Finally, we phylogenetically reconstructed the
evolutionary history of male canine height using the contMap
function in ‘phytools’(Revell 2012), which estimates the
maximum likelihood ancestral states of characters. All data
generated or analyzed during this study are included in the
supplementary information files.
Results
Results of PGLS statistical models of canine height and vol-
ume in males and females can be found in Table 1. Our anal-
yses showed that among artiodactyls, male canine height
(Fig. 1) and volume decreased as species increased in body
size (Fig. 2) and formed larger groups (Fig. 3), and smaller
solitary species have larger canines (Table 1). We found no
statistically significant effects, however, of body mass and
group size on canine size in females. We found no significant
effects of habitat openness or mating system type in either sex
for any measure of tusk size (Table 1).We did find, however,
that more solitary artiodactyls generally live in more closed
habitats (t=2.424,p= 0.018). This relationship suggests that
habitat openness is, indeed, related to tusk development.
Discussion
The deep ancestral relationships between the artiodactyl line-
ages are unclear; therefore our confidence in the accuracy of
tusk size reconstruction at deep nodes is limited, and
attempting to understand the loss of enlarged canines in ex-
change for cranial weapons (e.g., antlers, horns) remains a
difficult task. As predicted in our study, however, we found
that both body size and sociality strongly affected the
evolution of tusks in males; small solitary species tend to
develop larger canines. Though we did not find a direct rela-
tionship between canine size and habitat openness, additional
tests, however, show that solitary species tend to live in closed
environments, suggesting an ecological niche of being small,
solitary, living in closed dense forests, and using tusks to fight
over territory and/or mates. Similarly, one study found that
small-bodied slinkers primarily live in solitude amongst thick
tropical brush where large canines are favored for defending
small areas with low visibility over costly and cumbersome
headgear (Bro-Jørgensen 2008). Artiodactyls living in closed
habitats may not be able to detect or assess opponents from
afar; confrontations can quickly escalate and small slicing
weapons allow for better offense, maneuverability, and rapid
escape. Here, we discuss the general evolutionary transition
from stabbing tusks to displaying antlers and provide insight
into tusk evolution in each group that bears them.
Defenders of material resources need to be equipped with a
set of weapons that will allow them to quickly inflict maxi-
mum damage on competitors. When early forest-dwelling ar-
tiodactyls made the move to open arid zones, social behavior
allowed species to successfully exploit these new environ-
ments (Estes 1974). We found that tusk-bearing artiodactyls
are primarily solitary and live in closed habitats (temperate
and tropical forests; Table 2) where they are able to defend
small territories. However, movement into more open habitats
favored the evolution of larger body sizes that could support
elaborate forms of visual communication (e.g., large cranial
weapons, flashing of conspicuous rump patches; Estes 1991a;
Caro 2005;Raiaetal.2015). Though our results do not sup-
port a directrelationship between tusk loss and movement into
open habitats, our finding that artiodactyls increase in size and
become more social as they move into open habitats and sub-
sequently lose their tusks supports our hypothesis that moving
into open habitats favored the evolution of cranial appendages
that would serve both as signals to assess competitors and as
weapons in violent fights (i.e., they lie at the center of the
weapon-signal continuum, McCullough et al. 2016).
Our ancestral state reconstruction based solely on measure-
ments of male canine height (Fig. 1) from extant species
Table 1 Results of PGLS
statistical models of canine height
and volume in male and female
artiodactyls
Male Female
Height Volume Height Volume
tptptptp
Group size −2.958 0.004 −2.662 0.010 −1.656 0.103 −0.943 0.349
Body mass −0.211 0.015 −2.137 0.037 −1.865 0.067 −1.211 0.231
Habitat openness −0.155 0.878 −0.441 0.661 −0.408 0.685 −0.592 0.556
Mating system −0.203 0.840 −0.203 0.840 −0.335 0.739 −0.046 0.964
Results in bold signify factors that were statistically significant in explaining tusk retention in relation to canine
height and volume
J Mammal Evol

agrees with the fossil record and suggests that prominent but
not tusk-like male canines are the ancestral condition in artio-
dactyls. The entelodonts, one of the earliest pig-like artiodac-
tyls that emerged during the late middle Eocene, shared prim-
itive dental and skeletal features with other suoids including
canine tusks (Prothero 2017). Evidence from dental wear on
these early forms suggests that they were omnivores and scav-
engers (Prothero 2017). Suoids are separated into two families
that differ in canine morphology (Ungar 2010). In tayassuids,
canines grow vertically downwards, while suid canines often
emerge from the maxilla anterolaterally and curve upward
(MacKinnon 1981; Prothero 2017). True suids emerged and
diversified in the Miocene, which include the
kubanochoerines, a group of large pigs that developed a horn
between their eyes (Harris and Li-Ping 2007; Prothero 2017).
Interlocking canines of the tayassuids function in intraspecific
combat, while the Bout-turned^canines of suids function as
signals of status and rank prior to engaging in conspecific
fights (Herring 1972). The upper tusks of male babirusas are
the most extreme morphological example where they ascend
from the maxilla and curve onto themselves (MacKinnon
1981). This difference in tusk function between suids and
tayassuids supports our hypothesis in that suids (~88 kg
avg.) are larger in mass than tayassuids (~30 kg avg) (Smith
et al. 2003), although their openness of habitat score is similar.
The small hornless hypertragulids of the Eocene resembled
duikers but with highly arched bodies, slender dog-like front
legs, and blade-like canines used for territory defense, and
they hid amongst thick cover from predators (Métais and
Vislobokova 2007;Rössner2007). Tragulids (chevrotain or
Fig. 2 The relationship of log
e
(male canine height (mm) +1) and
log
e
(male body size (kg)) among male artiodactyls (N=63, t=−0.211,
p=0.015)

Fig. 1 Phylogenetic tree (Arnold et al. 2010) of artiodactyls showing
ancestral state reconstruction of log
e
(male canine height (mm) +1).
Branch Bwarmth^indicates larger tusk heights where red represents the
most prominent canines (tusks) and blue is the absence of enlarged ca-
nines. Intermediate colors (yellow and green) signify species that retained
tusks but are smaller relative to the babirusa (Babyrousa babyrussa)that
develops the most elaborate form of tusks
R
JMammalEvol

mouse-deer) as we know today radiated later during the
Oligocene retaining much of their ancestral body plans
(Geist 1998), making them the most primitive extant ruminant
(Rössner 2007). Tragulids are largely solitary but may be
found living in small groups (3–10 individuals) and their long,
slender legs allow them to run through dense undergrowth.
Male chevrotain will engage in primitive territorial fighting
where they stand antiparallel, slashing at each other’sneck
and sides often causing detrimental wounds (Dubost 1975,b;
Ralls et al. 1975). For this reason, tragulids develop a layer of
toughened skin on their backs to protect from intraspecific
attacks (Dubost and Terrade 1970; Dubost 1975b; Jarman
1989). Relative to body size, tragulid upper canines are short
and laterally oriented with a backwards curve much like the
muntjac, but the outward flare permits effective combat
(Aitchison 1946).
True antlers first emerged and diversified during the early
Miocene in cervids that bore thick upper canines (Geist 1998),
which were subsequently lost. This is supported by our ances-
tral state reconstruction, which suggests early ruminants did
possess small canine teeth (see light blue branches at the base
of the ruminant clade in Fig. 1). One of these forms,
Hoplitomeryx, strangely bore permanent pronged horns over
its eyes, a long horn on its nose, and long jutting canines
(Prothero 2017). Figure 1suggests that elongated tusks re-
evolved twice among cervids, first in muntjacs and tufted deer
(Muntiacus) and second in water deer (Hydropotes)asthese
groups reclaimed the slinker lifestyle. Muntiacus muntjak
and Hydropotes inermis are on average larger (24 kg
and 30 kg, respectively; Table 2) than tusk-bearing
tragulids and moschids but smaller than non-tusked ant-
lered cervids (average male mass = 114.64 kg). Muntjacs
are unique in that not only are the upper canines pre-
served but males also bear antlers. In this dual weapon
system, antlers are necessary for permitting the use of
widely divergent tusks (Janis and Scott 1987)during
combat (Barrette 1977), and indeed tusks are used as
secondary weapons (Aitchison 1946) during intraspecific
encounters (Yahner 1980). Further study of the evolu-
tionary transition from tusks to antlers in this group is
warranted.
While entirely devoid of cranial weaponry, Moschus and
Hydropotes bear large saber-like tusks with points directed
more downward (Janis and Scott 1987; Sánchez et al. 2010)
than the antlered muntjacs (Aitchison 1946). These long tusks
Fig. 3 log
e
(Canine height (mm) +1) as a function of increasing group size
among male artiodactyls (N = 63, t=−2.958, p= 0.004). The outliers
(filled circles) in the 11–50 column are all suoideans
Table 2 Results of measured
mean male and female canine
heights and natural history data
from literature searches on body
mass, sociality, and habitat
preference for species with
enlarged canines
Species Body Mass (kg) Canine Height (mm) Sociality Habitat openness
Male Female Male Female
Tragulus javanicus 1.9 1.9 14.81 4.36 1 0.40
Tragulus napu 4 4 15.26 2.92 1 0.45
Muntiacus muntjak 24 24 19.15 0 0 0.45
Muntiacus reevesi 14.5 12.5 26.82 8.08 0 0.53
Hydropotes inermis 30 30 25.09 9.12 0 0.37
Tayassu peccari 32.5 32.5 25.25 26.62 2 0.47
Pecari tajacu 21.5 21.5 30.84 27.87 2 0.58
Hylochoerus meinertzhageni 207.5 150 34.29 38.71 2 0.10
Moschus moschiferus 12 12 36.85 9.56 0 0.50
Sus scrofa 182 182 42.62 16.86 2 0.43
Babyrousa babyrousa 71.5 71.5 106.3 13.32 0 0.20
Sociality scores were assigned using the following: solitary or pairs= 1, 3–10 = 2, 11–50 = 3, > 50 = 4. Openness
scores were obtained from habitat use data from IUCN (Version 3.1) and editing it to include the primary 2–3
habitats in which species spend the most time. Larger openness scores correspond to the most opened habitats
MMM (mean male mass), MFM (mean female mass), MCH (male canine height), FCH (female canine height)
J Mammal Evol

are used as sexual weapons when engaged in intraspecific
combat to inflict wounds (Aitchison 1946). Musk deer fiercely
slash at their opponents often piercing the skin and even punc-
turing vital organs (Zhang et al. 1970;Sathyakumar,1992).
The upper canines of females, however, are subtle at 1 cm and
do not extend beyond the lower lip. Moschidae includes seven
extant species (Prothero 2017) with a rich fossil record and all
also bear large, saber-like canines. The earliest known
moschid, Dremotherium, emerged in the late Oligocene but
the group was completely extinct by the end of the early
Miocene except for Micromeryx, which survived for the re-
mainder of the Miocene (Prothero 2007). While the deep an-
cestral relationships of cervids, moschids, and bovids are un-
clear, and our ancestral state reconstruction is dependent pure-
ly on extant taxa, Fig. 1suggests that the common ancestor of
these groups possessed small- to medium-sized canines but
lacked tusk-like weapons and both Hydropotes and early
moschids evolved robust tusks independently.
Bovids are among the most diversified group of artiodac-
tyls ranging in body size and occupying a variety of habitats.
The earliest bovid, Eotragus, arose in the late early Miocene
weighing about 18 kg and bearing straight horns roughly 8 cm
long (Prothero 2017). Bovids from the Ice Age, however,
were the most impressive; Pelorovis antiques particularly
weighed up to 2000 kg with horns spanning 4 m in width
(Prothero 2017). Bovid horns are commonly used in combat,
especially among bovines, and horns in antelope in particular
are used for both signaling and fighting(Gosling 1986). While
all bovids lack upper canine teeth (Ungar 2010), a number of
extant bovids displaying the characteristics of a slinker have
evolved short stabbing horns (Cephalophus,Madoqua,
Neotragus,Oreotragus,Philantomba,Raphiceros,
Sylvicapra,Tetracerus). While antlers and horns are used to
signal status and fighting ability, the pointed horns of these
territorial forest-dwelling species are likely pure weapons,
possibly the result of selection favoring a tusk-like weapon
in an animal that lacked canines to elongate but already had
horns that could be repurposed instead. In our analysis, we
only sampled a few species belonging to this peculiar group of
bovids; therefore, future studies are needed that include a
more comprehensive representation of duiker-like species.
From our analysis, we show that body mass and sociality
are important in explaining the retention and subsequent loss
of sexual weaponry among artiodactyls. Small, solitary spe-
cies that rely on close-quarter combat for material resource
defense and rapid escape through dense forest to evade pred-
ators are likely to carry enlarged canines. Our results suggest
that emergence from cover facilitated the evolution of com-
plex social strategies; fewer close-range encounters meant
tusks were no longer optimal and elaborate cranial structures
allowed artiodactyls to advertise their fighting ability to con-
specifics from afar (Geist 1966;Janis1982). Special cases like
the muntjac and Chinese tufted deer that not only retain tusks
but also develop antlers, however, warrant further investiga-
tion. Because slinkers live inconspicuously among thick cover
and tend to be solitary, it is difficult to observe their behavior
in nature. Future studies are needed that target the specific
ecological and behavioral factors of these special cases in
order to understand the evolutionary changes that result in
the retention of multiple combative weapons.
Acknowledgments We thank James Dines and David Janiger at the Los
Angeles County Museum of Natural History, and curators at the
American Museum of Natural History and the National Museum of
Natural History at the Smithsonian Institution for access to their collec-
tions and support. We thank members of the Stankowich Lab at California
State University Long Beach and two anonymous reviewers for com-
ments on previous versions of this manuscript.
Compliance with Ethical Standards
This research was supported by funds from California State University
Long Beach, College of Natural Sciences and Mathematics. For this type
of study formal consent is not required. This article does not contain any
studies with human participants or animals performed by any of the au-
thors; all subjects were previously collected specimens deposited in mu-
seum collections.
Conflict of Interest The authors declare that they have no conflict of
interest.
References
Aitchison J (1946) Hinged teeth in mammals: a study of the tusks of
muntjacs (Muntiacus) and Chinese water deer (Hydropotes inermis).
Proc Zool Soc Lond 116:329–338
Arnold C, Matthews LJ, Nunn CL (2010) The 10kTrees website: a new
online resource for primate phylogeny. Evol Anthropol 19:114–118
Barrette C (1977) Fighting behavior of muntjac and the evolution of
antlers. Evolution 31:169–176
Brashares JS, Garland T, Arcese P (2000) Phylogenetic analysis of coad-
aptation in behavior, diet, and body size in the African antelope.
Behav Ecol 11(4): 452–63.
Bro-Jørgensen J (2007) The intensity of sexual selection predicts weapon
size in male bovids. Evolution 61:1316–1326
Bro-Jørgensen J (2008) Dense habitats selecting for small body size: a
comparative study on bovids. Oikos 117:729–737
Caro TM (2005) Antipredator Defenses in Birds and Mammals.
University of Chicago Press, Chicago
Caro TM, Graham CM, Stoner CJ, Flores MM (2003) Correlates of horn
and antler shape in bovids and cervids. Behav Ecol Sociobiol 55:32–41
Clutton-Brock TH, Albon SD, Harvey PH (1980) Antlers, body size and
breeding group size in the Cervidae. Nature 285:565–567
Dubost G (1975) Le comportement du Chevrotain africain, Hyemoschus
aquaticus Ogilby (Artiodactyla, Ruminantia). Z Tierpsychol 37(4):
403–448. https://doi.org/10.1111/j.1439-0310.1975.tb00889.x
Dubost G, Terrade R (1970) La transformation de la peau des Tragulidae
en bouclier protecteur. Mammalia 34:505–513
Emlen DJ (2008) The evolution of animal weapons. Annu Rev Ecol Evol
Syst 39:387–413
Erfurt J, Métais G (2007) Endemic European Paleogene artiodactyls:
Cebochoeridae, Choeropotamidae, Mixtotheriidae, Cainotheriidae,
Anoplotheriidae, Xiphodontidae, and Amphimerycidae. In:
JMammalEvol

Prothero DR, Foss SE (eds) The Evolution of Artiodactyls. Johns
Hopkins University Press, Baltimore, pp 59–94
Estes RD (1974) Social organization of the African Bovidae. In: Geist V,
Walther F (eds) The Behaviour of Ungulates and Its Relation to
Management. IUCN Morges, Switzerland, pp 166–205
Estes RD (1991a) The Behavior Guide to African Mammals Vol 64.
University of California Press, Berkeley
Estes RD (1991b) The significance of horns and other male secondary
sexual characters in female bovids. Appl Anim Behav Sci 29:403–451
Geist V (1966) The evolution of horn-like organs. Behaviour27:175–214
Geist V (1971) The relation of social evolution and dispersal in ungulates
during the Pleistocene, with emphasis on the Old World deerand the
genus Bison. Quaternary Res 1(3): 285–315
Geist V (1974) On the relationship of social evolution and ecology in
ungulates. Am Zool 14:205–220
Geist V (1998) Deer of the World: Their Evolution, Behavior, and
Ecology. Stackpole Books, Mechanicsburg
Gentry AW (1994) The Miocene differentiation of Old World Pecora
(Mammalia). Hist Biol 7:2 115–158
Gittleman JL, Van Valkenburgh B (1997) Sexual dimorphism in the ca-
nines and skulls of carnivores: effects of size, phylogency, and be-
havioural ecology. J Zool 242:97–117
Goss RJ (2012) Deer Antlers: Regeneration, Function and Evolution.
Academic Press, New York
Gosling LM (1986) The evolution of mating strategies in male antelopes. In:
Rubenstein DI, Wrangham RW (eds) Ecological Aspects of Social
Evolution. Princeton University Press, Princeton, pp 244–281
Harris JM, Li-Ping L (2007) Superfamily Suoidea. In: Prothero DR, Foss
SE (eds) The Evolution of Artiodactyls. Johns Hopkins University
Press, Baltimore, pp 130–150
Herring SW (1972) The role of canine morphology in the evolutionary
divergence of pigs and peccaries. J Mammal 53:500–512
IUCN (2017) The IUCN Red List of Threatened Species. IUCN Global
Species Programme Red List Unit. http://www.iucnredlist.org .
Accessed 6 April 2017
Janis CM (1982) Evolution of horns in ungulates: ecology and paleoecol-
ogy. Biol Rev 57:261–318
Janis CM (1990) The correlation between diet and dental wear in herbiv-
orous mammals, and its relationship to the determination of diets of
extinct species. In: Boucot AJ (ed) Evolutionary Paleobiology of
Behaviour and Coevolution. Elsevier, Amsterdam, Toronto, pp
241–259
Janis CM, Scott KM (1987) The interrelationships of higher ruminant
families with special emphasis on the members of the Cervoidea.
Am Mus Novitates 2893:1–85
Jarman PJ (1974) The social organization of antelope in relation to their
ecology. Behaviour 48: 215–67
Jarman PJ (1989) On being thick-skinned: dermal shields in large mam-
malian herbivores. Biol J Linn Soc 36(1–2):169–191. https://doi.
org/10.1111/j.1095-8312.1989.tb00489.x
Lundrigan B (1996) Morphology of horns and fighting behavior in the
family Bovidae. J Mammal 77:462–475
MacKinnon J (1981) The structure and function to the tusks of babirusa.
Mammal Rev 11:37–40
McCullough EL, Miller CW, Emlen DJ (2016) Why sexually selected
weapons are not ornaments. Trends Ecol Evol 31:742–751
Métais G, Vislobokova I (2007) Basal ruminants. In: Prothero DR, Foss
SE (eds) The Evolution of Artiodactyls. Johns Hopkins University
Press, Baltimore, pp 189–212
Nowak RM (1999) Walker’sMammals of the World. Johns Hopkins
University Press, Baltimore
Orme D, Freckleton R, Thomas G, Petzoldt T, Fritz S, Isaac N, Pearse W
(2012) caper: comparative analysis of phylogenetics and evolution
in R. R package version 0.5. http://CRAN.R-project.org/package=
caper
Packer C (1983) Sexual dimorphism: the horns of African antelopes.
Science 221:1191–1193
Prothero DR (2007) Family Moschidae. In: Prothero DR, Foss SE (eds)
The Evolution of Artiodactyls. Johns Hopkins University Press,
Baltimore, pp 221–226
Prothero DR (2017) The Princeton Field Guide to Prehistoric Mammals.
Princeton University Press, Princeton
R Core Team (2012) R: A language and environment for statistical com-
puting (Ver. 2.14.12). R Foundation for Statistical Computing,
Vienna.
Raia P, Passaro F, Carotenuto F, Maiorino L, Piras P, Teresi L, Meiri S,
Itescu Y, Novosolov M, Baiano MA, Martínez R, Fortelius M
(2015) Cope's rule and the universal scaling law of ornament com-
plexity. Am Nat 186:165–175
Ralls K, Barasch C, Minkowski K (1975) Behavior of captive mouse
deer, Tragulus napu. Z Tierpsychol 37:356–378
Revell LJ (2012) Phytools: an R package for phylogenetic comparative
biology (and other things). Methods Ecol Evol 3:217–223
Rico-Guevara A, Hurme KJ (2018) Intrasexually selected weapons. Biol
Rev. https://doi.org/10.1111/brv.12436
Rössner GE (2007) Family Tragulidae. In: Prothero DR, Foss SE (eds)
The Evolution of Artiodactyls. Johns Hopkins University Press,
Baltimore, pp 213–220
Sánchez IM, Domingo MS, Morales J (2010) The genus Hispanomeryx
(Mammalia, Ruminantia, Moschidae) and its bearing on musk deer
phylogeny and systematics. Palaeontology 53:1023–1047
Sathyakumar S (1992) The musk deer. Sanctuary Asia 12:52–57
Smith FA, Lyons SK, Ernest SKM, Jones KE, Kaufman DM, Dayan T,
Marquet PA, Brown JH, Haskell JP (2003) Body mass of late
Quaternary mammals. Ecology 84: 3403
Stankowich T (2012) Armed and dangerous: predicting the presence and
function of defensive weaponry in mammals. Adaptive Behavior 20:
32–43
Stankowich T, Caro T (2009) Evolution of weaponry in female bovids.
ProcRSocB276:4329–4334
Stankowich T, Haverkamp PJ, Caro T (2014) Ecological drivers of anti-
predator defense in carnivores. Evolution 68:1415–1425
UMMZ (2015) Animal Diversity Web. University of Michigan Museum
of Zoology. http://animaldiversity.ummz.umich.edu/site/index.html.
Accessed 6 April 2017
Theodor JM, Erfurt J, Métais G (2007) The earliest artiodactyls:
Diacodexeidae, Dichobunidae, Homacodontidae, Leptochoeridae,
and Raoellidae. In: Prothero DR, Foss SE (eds) The Evolution of
Artiodactyls. Johns Hopkins University Press, Baltimore, pp 32–58
Ungar PS (2010) Mammal Teeth: Origin, Evolution, and Diversity. Johns
Hopkins University Presss, Baltimore
Wilson DE, Mittermeier RA (2011) Handbook of the Mammals of the
World. Vol. 2. Hoofed Mammals. Lynx Edicions, Barcelona
Yahner RH (1980) Barking in a primitive ungulate, Muntiacus reevesi:
function and adaptiveness. Am Nat 114:157–177
Zhang BL, Dang FM, Li BS (1970) The Farming of Musk Deer.
Agricultural Publishing Company, Peking
J Mammal Evol