ArticlePDF Available

Taphonomy, Life History, and Human Exploitation of Rhinoceros sinensis at the Middle Pleistocene Site of Panxian Dadong, Guizhou, China


Abstract and Figures

Rhinoceros remains are commonly found in Chinese Pleistocene archaeological localities. This study examines the characteristics of the Rhinoceros sinensis sample from Panxian Dadong, a karst cave in the mountains of western Guizhou province, with a mammalian fauna in association with stone artefacts and human remains from the late Middle Pleistocene (MIS 6-8). The distribution of skeletal elements shows a predominance of foot (metapodial and phalanges) and lower limb (carpals and tarsals) bones, while the dental age-at-death profile, constructed using dental eruption and tooth wear data, is characterised by a high frequency of prime age adult teeth. There is little taphonomic evidence for the involvement of non-human carnivores or natural agencies in the formation of the faunal assemblage. Instead, it appears that human activities were responsible for the unexpected prevalence of prime age adults.
Content may be subject to copyright.
Taphonomy, Life History, and
Human Exploitation of Rhinoceros
sinensis at the Middle Pleistocene
Site of Panxian Dadong,
Guizhou, China
Department of Anthropology, Florida State University, PO Box 3067772, Tallahassee, FL
32306-7772, USA
Department of Anthropology/Geography, California State University, Stanislaus, One
University Circle, Turlock, CA 95382, USA
ABSTRACT Rhinoceros remains are commonly found in Chinese Pleistocene archaeological localities.
This study examines the characteristics of the Rhinoceros sinensis sample from Panxian
Dadong, a karst cave in the mountains of western Guizhou province, with a mammalian fauna
in association with stone artefacts and human remains from the late Middle Pleistocene (MIS 6-
8). The distribution of skeletal elements shows a predominance of foot (metapodial and
phalanges) and lower limb (carpals and tarsals) bones, while the dental age-at-death profile,
constructed using dental eruption and tooth wear data, is characterised by a high frequency of
prime age adult teeth. There is little taphonomic evidence for the involvement of non-human
carnivores or natural agencies in the formation of the faunal assemblage. Instead, it appears
that human activities were responsible for the unexpected prevalence of prime age adults.
Copyright ß2008 John Wiley & Sons, Ltd.
Key words: China; Middle Pleistocene; rhinoceros; age-at-death profile; taphonomy
A principal research objective for investigators of
Palaeolithic cave sites is to understand site
formation processes and the use of caves by
hominins and other animals. One key source of
information is the accumulated animal fossil
remains. Faunal analysis has proven to be an
essential component of site formation studies, and
it is particularly informative when bioturbation
and diagenesis obscure finer-level stratigraphic
distinctions within travertines, breccias or clays,
or when no site structures or features are discernible.
Rhinoceroses are found at approximately 80%
of all archaeological sites in China (Tong, 2000,
2001b), suggesting they were an important
prehistoric food source. There were northern
representatives of the Rhinocerotidae, including
Dicerorhinus and Coelodonta, that were two-horned
and adapted to colder climatic conditions. To the
south of the Qinling Mountains, the Ailuropoda-
Stegodon faunas of the Pleistocene include abun-
dant examples of Rhinoceros sinensis, a large species
with a single horn and the hypsodont (high-
crowned) molars that are associated with grazing
species. Its closest affinities are with the more
International Journal of Osteoarchaeology
Int. J. Osteoarchaeol. (2008)
Published online in Wiley InterScience
( DOI: 10.1002/oa.1025
* Correspondence to: Department of Anthropology, Florida State
University, P.O. Box 3067772, Tallahassee, FL, 32306-7772, USA.
Copyright #2008 John Wiley & Sons, Ltd. Received 11 March 2008
Revised 16 August 2008
Accepted 27 August 2008
tropical rhinoceroses from India and Java
(Colbert & Hooijer, 1953).
Although rhinoceros are commonly found in
the Pleistocene cave deposits of southern China,
the process by which they became elements of
faunal assemblages is rarely studied. Generally,
they are just identified and included in faunal lists
that describe paleontological localities and
archaeological sites. This paper describes the
detailed taphonomic and mortality analysis of late
Middle Pleistocene R. sinensis material excavated
from Panxian Dadong cave in Guizhou province,
southern China. The clear presence of humans at
Panxian Dadong is documented by five isolated
teeth and numerous stone tools made of lime-
stone, chert and basalt (Miller-Antonio et al.,
2004), tools made of R. sinensis teeth (Miller-
Antonio et al., 2000), and cutmarks, percussion
damage and burning on faunal material (Sche-
partz et al., 2003). The evidence for human
activity and rhinoceros fossils throughout the
stratigraphic sequence at Dadong makes it crucial
to evaluate the possible role played by humans in
the formation of the rhinoceros sample.
R. sinensis is well represented in the Dadong
fauna and comprises 24% of the total mammalian
collection identifiable to taxon. The diversity of
mammalian species represented at Dadong
indicates that a variety of habitats were available
in the vicinity of the cave (Schepartz et al., 2003;
Bekken et al., 2004). R. sinensis, like many other
rhinoceroses, apparently had the flexibility to
withstand extremes of climate and the ability to
exploit a wide range of plant foods (Osborn,
Our interest in rhinoceroses at Panxian
Dadong is part of a larger faunal analysis designed
to: (1) provide information on the palaeoenvir-
onment of upland southern China during the
Middle Pleistocene; (2) investigate the inter-
actions between prehistoric humans and mamma-
lian species; and (3) clarify the role of humans,
carnivores and natural geological processes in the
formation of the Panxian Dadong faunal assem-
blage. This paper builds upon our previous study
of the taphonomy and mortality profile for the
proboscidean Stegodon orientalis (Schepartz et al.,
2005) which identified a striking predominance
of young animals that were most likely brought
into the cave through hominin activities. Here we
test the hypothesis that the same pattern applies
to R. sinensis.
Regional setting and site characteristics
Little is known about the Middle Pleistocene
palaeoenvironment of upland southern China in
comparison with the well-studied loess sequences
of north China. While many of the karst caves of
Guizhou remain unexplored, Dadong was the
focus of geological, palaeontological, and archae-
ological investigations for over ten years (detailed
in Huang et al., 1995, and special volumes of Acta
Anthropologica Sinica 16(3), 1997, and Asian
Perspectives 43(2), 2004) because of its deep
stratigraphic sequence, the abundance of fossil
mammals, and associated archaeological materials.
It preserves late Middle Pleistocene deposits that
provide palaeoenvironmental context for the
periods when human activities are documented
(Karkanas et al., 2008).
Dadong is located on the western Guizhou
Plateau (2583703800N, 1048440E; Figure 1), a
region where continuous ranges of peaks contain
numerous caves. Dadong is the middle of three
interconnecting caverns stacked within a 230 m
high hill approximately 1630 m ASL. The
entrance is presently located 32.4 m above
the valley floor as a result of recent uplift of
the plateau; during the late Middle Pleistocene, it
would have been closer to the valley floor and
near the confluence of three small rivers that
drained into the porous limestone of the lower
cave. A later collapse produced a smaller pitfall
entrance at the back of the cave.
Dadong’s 8000 m
main chamber contains a
clastic sequence of flowstones, finegrained clays
and silt, consolidated gravels, and large limestone
blocks. Uranium-series (U-series) dates (Shen
et al., 1997) and electron spin resonance (ESR)
dates (Rink et al., 2003; Jones et al., 2004) of tooth
enamel samples suggest that most of the
excavated levels at Dadong were deposited
between 130–300 ka. The stratigraphic layers
(Figure 2A) document an environmental sequence
with substantial climatic fluctuations (Karkanas
et al., 2008) corresponding to marine isotope stages
(MIS) 6–8. The faunal assemblage at Dadong
shows the loss of more temperature-sensitive
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
L. A. Schepartz and S. Miller-Antonio
species such as non-human primates during colder
periods, while the hardier stegodonts and
rhinoceroses persist throughout the sequence.
The Dadong faunal sample consists of a variety
of species that are characteristic of the Ailuropoda-
Stegodon faunas of Middle Pleistocene southern
China (Zhang et al., 1997; Pan & Yuan, 1997;
Schepartz et al., 2003). There are many ungulates
(cervids, small and large bovids, pigs) and lesser
numbers of primates (macaques, colobines and
hominins) and carnivores (mustelids, foxes,
hyenas, tigers, leopards, pandas). Other notable
genera are large-bodied ungulates that would not
ordinarily frequent caves, such as Stegodon orientalis,
Rhinoceros sinensis, and the giant tapir Megatapirus
augustus. This range of species suggests that the
Pleistocene environment was mixed woodland. In
addition, the recovery of pandas, bamboo rats
and colobine monkeys (Pan & Yuan, 1997)
suggests the existence of some densely forested
areas with bamboo during warmer periods.
Materials and methods
The Rhinoceros sinensis component
The mammalian faunal sample from Dadong
(n¼7045) consists of 3.3% skull fragments,
12.5% isolated teeth and 84% postcranial
elements. The sample is primarily characterised
by fragmentary bone, and there are very few
examples of articulated elements. The most
common identified species is R. sinensis, which
comprises 24% of the total elements identifiable
to taxon, followed by Stegodon orientalis at 13%
(Schepartz et al., 2003).
Figure 1. Map of China showing locations of (1) Panxian Dadong and (2) Nanjing.
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
Rhinoceros sinensis at Panxian Dadong, China
From the total number of identified specimens
(NISP) of 285 rhinoceros elements, we analyse
224 specimens that are identifiable to tooth class
or skeletal element (the remaining 61 elements
are tooth fragments). The sample consists of teeth
(n¼121), skull (n¼3), and postcranial elements
(n¼100). Rhinoceros material is found throughout
the sequence of excavated strata, as indicated in a
vertical distribution of all the piece-plotted
specimens (greater than 2.5 mm) (Figure 2B). It
is important to stress that none of the skeletal
elements were found in articulation or could be
conclusively associated with each other.
Skull bone is surprisingly rare given the large
number of teeth present and the high likelihood
that dense portions such as the mandibular corpus
would be preserved relative to some of the more
fragile skeletal elements. In terms of counts of
elements, the postcrania are clearly dominated by
foot bones (metapodials and phalanges) and
lower leg bones (carpals and tarsals) (Figure 3).
Among these, phalanges are the most frequently
recovered elements.
Seven post-cranial specimens (i.e. 7%) are from
immature animals; two of the femoral portions
may represent the same individual. A minimum
Modern disturbance
521 520 519 518 517 516 515 514 513 512
E46 F46 G46 H46 I46 J46
Flowstone Cemented
East Profile
512 513 514 515 516 517 518 519 520
Figure 2. (A) Stratigraphic profile of Panxian Dadong excavations, drawn by P. Karkanas and adapted from Figure 5 in
Karkanas et al. (2008). (B) Plot of excavated rhinoceros dental and skeletal elements. North is to the right. The y-axis is
the depth in metres from the hypothetical surface of 100 m; the x-axis is a north–south range of the main excavation area
in metres. Lighter circles are teeth, darker circles are bone.
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
L. A. Schepartz and S. Miller-Antonio
number of individuals (MNI) based on the counts
of the non-dental elements is 4, although this
must be viewed as a low estimate of the number of
individuals present because it does not factor in
skeletal maturation criteria.
Rhinoceros tooth classes and tooth
eruption sequence
Over the course of rhinoceros evolution the
anterior dentition has gone from having special-
ised cutting and shearing functions to being more
rudimentary and highly variable. The living Asian
species retain a tusk-like lower incisor, while the
African species only have greatly reduced incisors
(Ballenger & Myers, 2001). By the Pleistocene
most rhinoceros dentitions became dominated by
a series of large grinding premolars and molars
(Osborn, 1903; Koulmann & Koulmann, 1970).
The permanent dentition of R. sinensis lacks
canines and has small upper and lower incisors
(I1 and I2) and a small first premolar (P1) that is
frequently not present. These teeth are rarely
preserved in the palaeontological record of China
(cf. Colbert & Hooijer, 1953; Tong, 2001a, b); it
follows that most analyses focus on the posterior
dentition which includes the remaining premolars
(P2–4) and three molars (M1–3). Maxillary
premolars and molars are similar in morphology:
both have large buccal ectolophs and lingual
meta- and protolophs separated by a deep valley
that widens as the tooth wears (Osborn, 1903).
The permanent M3 is the most distinctive of the
maxillary teeth, having a triangular rather than a
square or rectangular shape (Figure 4, Stage 3
wear). The mandibular premolars and molars
have two crescent-shaped or columnar lophs
(termed the metalophid and hypolophid) that
form posterior and anterior valleys (Osborn,
1903; Colbert & Hooijer, 1953) (Figure 5). The
mandibular tooth classes are very difficult to
distinguish, particularly when isolated worn teeth
are evaluated.
R. sinensis also has four deciduous cheek-teeth.
In the literature these are alternately referred to
as deciduous molars (cf. Colbert & Hooijer,
1953), deciduous premolars (Tong, 2001a) or
simply deciduous (D) 1–4 (Gue
´rin, 1980). We use
the latter terminology here. The deciduous teeth
erupt in the following sequence, as summarised in
Tong (2001a): D2, D3, D1, and finally D4.
Occasionally, D1 erupts after D4. The permanent
tooth eruption sequence is generally: M1, M2,
P2, P3, P4 and M3 (Groves, 1967; Borsuk-
Bialynicka, 1973; Gue
´rin, 1980, as cited in Tong,
2001a). The most variable teeth are D1, which
can be retained into young adulthood, and M2
which always erupts before P4. The most reliable
teeth for ageing on the basis of the eruption
sequence are M1, P4 and M3 (Tong, 2001a). In
terms of relative age assessment, living rhino-
ceroses are classified into neonates (including
animals from birth to less than 1.5 years),
juveniles (1.5–4 years), subadults (4–6 years)
and adults (over 6 years) (Foster, 1965). The cows
Skeletal Elements
Skull Scap Hum Rad/Ul Femur Tibia Carp/Tars Metapodial Phalanx
Figure 3. Skull and post-cranial representation of rhinoceros elements. This figure is available in colour online at
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
Rhinoceros sinensis at Panxian Dadong, China
reach sexual maturity by six years, the bulls are
fully mature at 7–10 years, and the lifespan may
encompass 45 years (Ballenger & Myers, 2001).
Table 1, derived from data on the extant African
species of Diceros bicornis (black rhinoceros) and
Ceratotherium simum (white rhinoceros), presents
the posterior tooth eruption sequence for these
different age cohorts. From the table, the
following major distinctions in dental develop-
ment can be summarised:
(1) neonates and animals less than 1.5 years old
have only deciduous teeth;
(2) early juveniles have worn deciduous teeth and
permanent molars with beginning stages of
Figure 4. Wear stages (0–5) and examples of maxillary tooth morphology. This figure is available in colour online at
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
L. A. Schepartz and S. Miller-Antonio
(3) late juveniles have highly worn D1s and D4s,
and P2s and P3s with little wear;
(4) subadults have P4s with little wear and
unworn M3s;
(5) adults are characterised by the greater stages
of premolar and molar wear.
This information can be used, in conjunction
with tooth wear data, to develop an age-at-death
profile for the Dadong rhinoceroses.
The total count of rhinoceros maxillary and
mandibular teeth at Dadong is 70 and 51,
respectively (Figures 6 and 7). Few of the teeth
are preserved in jaws, but one of the most
complete examples is a large mandible with the
right P2–P4 and left P3 (Figure 5). Among the
tooth classes, premolars are most frequent at
Dadong. There are only two maxillary D1s in the
sample. These are small teeth often retained into
adulthood and heavily worn. They are rarely
preserved as individual teeth. We will focus
primarily on the posterior maxillary dentition due
to the larger sample size and the greater precision
in recognising tooth class distinctions for
maxillary teeth. The minimum number of
individuals (MNI) based on the highest frequency
tooth category (14 maxillary P2s) would be 7, but
a higher MNI is suggested by the presence of 7
maxillary M3s and 4 maxillary D4s. As animals do
not retain M3s and D4s concurrently and none of
these teeth have good antimeres in the sample, a
dental MNI of 11 is a better estimate.
Figure 5. Specimen 99D F46:455 mandible with LP3,
RP2–P4. This figure is available in colour online at www.
Table 1. Tooth eruption and age cohorts (modified from Tong, 2001a)
Teeth present Key features Age cohort (yrs)
D1–D4 Deciduous teeth only Neonate (<1.5)
D1–D4, M1–M2 Deciduous teeth and permanent molars Early juvenile (app. 1.5–3)
D1?, D4, P2–P3, M1–M2 D4 not yet replaced Late juvenile (app. 3–4)
P2–P4, M1–M3 M3 erupted but not in use Subadult (4–6)
P2–P4, M1–M3 M3 in use Adult (>6)
Maxillary Dentition
D1 D2 D3 D4 P2 P 3 P 4 M1 M2 M 3
Tooth Class
Figure 6. Distribution of maxillary tooth classes (n¼70). This figure is available in colour online at www.interscience.
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
Rhinoceros sinensis at Panxian Dadong, China
It should be emphasised that these MNI
estimates are based upon analysing the Dadong
sample as a single entity without stratigraphic
divisions. Interestingly, if the dental sample is
divided into upper (n¼35) and lower (n¼39)
subsamples following what appears to be a
potential gap in the spatial distribution of
piece-plotted elements around the elevation of
95 m (Figure 2B) and corresponding roughly to
the division between layers V and IV (Figure 2A),
the frequencies of deciduous teeth, premolars and
molars are basically the same for each subsample
(upper: 37% deciduous, 29% premolars, 34%
molars; lower: 38% deciduous, 28% premolars,
33% molars), and the combined MNI determi-
nation remains the same as that for the combined
Table 2 includes descriptive statistics for the
Dadong maxillary sample. Basic comparison with
data published in Colbert & Hooijer (1953)
suggests that R. sinensis at Dadong is similar in size
to their study collection, although our measures
do not correspond exactly with their system of
measurement (they measured at the base of
crowns; we measured maximum BL and MD
measures across the crown and perpendicular to
each other). In addition to what is reported in the
table, we have one deciduous incisor that is
heavily worn. The crown height and mesiodistal
length are the measures most affected by age-
related changes as the tooth is used. The
variability in the buccolingual measures, as seen
in the relatively high standard deviations, is
probably best explained as representing size
variation due to sexual dimorphism.
Determination of dental age wear stages
Rhinoceros teeth develop several characteristic
features as they wear. The initial changes involve
blunting of crests and the beginning of dentine
Table 2. Descriptive statistics of maxillary dentition
Tooth nMean SD Range
D1 MD 2 26.2 1.414 25.2–27.2
BL 2 23.9 7.85 18.4–29.5
HT 2 21.4 0.354 21.1–21.6
D2 MD 1 30.5
BL 1 30.3
HT 1 23.2
D3 MD 2 39.4 3.182 37.1–41.6
BL 1 43.8
HT 1 40.5
D4 MD 4 46.9 10.315 34.2–55.4
BL 4 50.8 7.507 42.1–59.3
HT 3 51.8 24.239 24–68.7
P2 MD 15 30.7 2.24 27.4–35.2
BL 16 35.7 4.82 21.5–40.4
HT 16 25.5 8.562 12.8–47.9
P3 MD 12 38.2 3.799 30.8–43.1
BL 11 45.9 5.8 35.7–51.8
HT 11 31.8 17.646 9.7–61.5
P4 MD 9 40.7 4.514 33.3–48.3
BL 9 56.6 5.474 48.5–63.7
HT 9 21.2 5.979 11.3–34.1
M1 MD 6 49.1 3.695 47.2–55
BL 6 60.3 5.013 54.4–67.8
HT 6 29.8 5.091 21.3–35.1
M2 MD 3 51.3 1.301 50–52.6
BL 3 53.3 7.209 45–58
HT 3 30.7 8.056 21.6–37
M3 MD 6 43.9 6.25 34.2–49.7
BL 6 48 6.547 41.7–56.4
HT 5 32 4.932 25.5–37.5
MD, mesiodistal length; BL, buccolingual breadth; HT,
crown height.
Mand ibular Dentition
D2 D3 D4 P1 P2 P3 P4 M1 M2 M3
Tooth Class
Figure 7. Distribution of mandibular tooth classes (n¼51). This figure is available in colour online at www.interscience.
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
L. A. Schepartz and S. Miller-Antonio
exposure along the crest ridges. In contrast to the
relatively flat advanced wear of the mandibular
dentition, the maxillary teeth develop strongly
sloped or dished occlusal surfaces as wear
progresses on the meta- and protolophs. The
anterior premolars (P2s) tend to wear flat and to
eradicate the crest features more completely.
With age, the mesial-distal length of all teeth is
significantly reduced due to interproximal wear,
and the enamel is often missing from the mesial
and distal surfaces of heavily worn teeth. Based on
this wear information, we scored individual teeth
on a 0–5 scale, as listed below, to facilitate age
estimates. Higher scores reflect greater levels of
wear. Figure 4 provides examples of the wear
stages. Louguet (2006) recently published another
attritional ageing system for Rhinocerotid teeth.
While it is fundamentally the same as our
methodology in terms of the recognition of the
wear processes, the scoring is reversed and
therefore not directly comparable to our system:
0: no wear or an unerupted tooth;
1: edge of crests show very slight wear;
2: edge of crests show dentine and flattening;
3: major dentine exposure but crest morphology
still visible;
4: flat wear and major features of crest enamel are
5: wear to the root on the lingual side or strong
scooped or sloping wear.
According to tooth eruption patterns (Table 1)
and this wear scheme, the youngest rhinoceroses
(neonates and animals aged less than 1.5 years)
can be identified by the presence of unworn or
slightly worn (stages 0–1) deciduous teeth. The
use of the deciduous teeth for ageing is
complicated by their continued presence in the
mouth for several years; in some cases D1 and D4
may be in use throughout the juvenile phases
(Tong, 2001a). Thus the presence of deciduous
teeth alone is not evidence for very young ages
and their wear scores must also be factored into
an age-at-death profile. At the other end of the
age spectrum, the presence of an M3 with any
stage of wear indicates an individual in the adult
age cohort.
Deciduous teeth comprise 22% of the dental
sample identifiable to tooth class. They demon-
strate a range of wear scores that can be used for
assigning them to the neonate and juvenile age
cohorts. About 30% of the deciduous teeth are
from neonates or individuals under 1.5 years of
age, while 44% are heavily worn and represent
late juvenile stage individuals (Figure 8). The
permanent teeth also provide important infor-
mation for the construction of the age-at-death
profile. There are very few lightly used teeth in
the permanent dentition (Figures 9 and 10).
Maxillary teeth have predominantly higher
scores, indicating that the majority are from
adult animals. Conversely, the mandibular tooth
scores show a pattern of less wear, with no teeth
scored at Stage 5 (Figure 10). The difference
between the maxillary and mandibular wear may
be explained by taphonomic factors. It is unlikely
Figure 8. Wear stages for the deciduous teeth, maxillary and mandibular combined (n¼27). This figure is available in
colour online at
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
Rhinoceros sinensis at Panxian Dadong, China
that the Dadong dental sample lacks mandibular
teeth with advanced wear. However, due to their
morphology, those teeth are more likely to
fragment when heavily worn and the resulting
pieces cannot be assigned to tooth class. For this
reason, we suggest that the maxillary tooth scores
are more characteristic of the sample.
When both permanent and deciduous tooth
scores are then assigned to the different age
cohorts, an age-at-death profile for the sample
shows some clear patterning. There are 8.3% of
the teeth from neonates and animals aged less
than 1.5 years, 5.8% from early juveniles, 27.3%
from late juveniles, 9.1% from subadults and
49.6% from adults (Figure 11).
There are several factors complicating the
construction of a dental age profile for this
sample. Deciduous teeth are more difficult to
interpret than the permanent teeth. Since
deciduous teeth are shed, the presence of isolated
deciduous teeth does not necessarily indicate
death of the individual or young age (although
the sample does contain some unworn teeth
without root resorption that could only come
from carcasses of young animals). Figure 11
therefore represents the developmental age of
some individuals in combination with age-at-
death information from carcasses. Another
complication stems from the nature of the
sample. Few specimens at Dadong can be
associated with other teeth and must con-
sequently be considered to represent individual
elements. This bias, which is the largest source of
potential error for this analysis, cannot be
avoided when isolated teeth constitute the bulk
of the sample. In addition, as discussed above, we
Figure 9. Distribution of wear stages for each tooth class of the maxillary permanent dentition (n¼55). This figure is
available in colour online at
Figure 10. Distribution of wear stages for each tooth class of the mandibular permanent dentition (n¼30). This figure is
available in colour online at
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
L. A. Schepartz and S. Miller-Antonio
examined the sample as a whole without spatial or
temporal divisions to maximise the sample size.
This has important ramifications in the interpret-
ation of the age-at-death profile, but the use of an
unstratified sample probably does not introduce
as much bias to this study as might be expected,
because some degree of time-averaging is
characteristic of all faunal assemblages from
prehistoric sites that are the product of repeated
events (Haynes, 1991).
Mortality profiles
The two basic mortality profiles that archae-
ozoologists use to model population age structure
are termed ‘catastrophic’ and ‘attritional’. In a
catastrophic profile, successive age classes con-
tain fewer individuals because it represents a
population’s age structure at the time of death as
the result of a non-selective event (such as a flood
or fire). An attritional (normal or cemetery)
profile is bimodal or U-shaped, with most
individuals from the youngest and oldest age
classes (Klein, 1982; Lyman, 1994). Both profiles
assume that the populations do not exhibit
dramatic fluctuations in births or survival, and
are therefore stable and stationary. Haynes (1991:
Figure 6.4) refined this classification by specifying
four types of age profile in living and extinct
proboscidean assemblages. In addition to attri-
tional and bimodal distributions (Types A and B),
he described another pattern (Type C) that is
dominated by prime-age adults. This results from
selective mortality affecting males only (for
example, killings by hunters interested in ivory),
or from non-selective mortality affecting declin-
ing populations. His fourth distribution (Type D)
shows no patterning and results from various
causes, including insufficient sample size.
Applying Haynes’s (1991) age profile interpret-
ations to the Dadong rhinoceros data, several
observations are relevant. A Type A profile might
be expected in a time-averaged sample such as
ours, but the Dadong distribution does not show
this pattern. In contrast to a Type B bimodal
profile, where a mode with substantial numbers of
very young individuals is expected, we have only
one clear mode of adults (almost half of the total
sample). A Type C profile is the best fit. The
question is whether these are prime age animals
rather than older, more vulnerable individuals.
There are several of the later sequence premolars,
M1s and M2s, with advance wear scores, but
there is a notable lack of M3s in Stage 4 or 5 wear.
We interpret this as an absence of teeth from
advanced age adults (Figures 9 and 10) and a
predominance of teeth from prime individuals
(see also Borsuk-Bialynicka, 1973: Table 4, p.15).
There is also a sizeable representation of late
juveniles and subadults – those who are pre-
sumably inexperienced and vulnerable to preda-
tion as they are no longer travelling under their
mother’s protection (c.f. Foster, 1965; Goddard,
Taphonomic indicators of accumulation
and transport
Several activities could produce the rhinoceros
assemblage at Panxian Dadong. The material may
L. Juvenile
E. Juvenile
Age Cohorts
Figure 11. Distribution of age cohorts from individual teeth, deciduous and permanent combined (n¼121). This figure
is available in colour online at
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
Rhinoceros sinensis at Panxian Dadong, China
have accumulated through five factors: the natural
loss of deciduous teeth, the natural death of
animals, water flow into the cave, transport by
porcupines, or predation by large carnivores or
humans. It is therefore important to examine each
of these factors in light of what appears to be a
differential representation of teeth from adult
There is some question as to why rhinoceroses
would naturally appear in a cave fauna. With
abundant drainage systems on the Guizhou
Plateau and rivers in the Dadong valley, they
would probably not have had to venture into the
cave to obtain water, but it might have been a
source of mineral salts. Accidental deaths of
young rhinoceroses might occur naturally in karst
environments where natural fissure traps are
formed through limestone dissolution. There
may have been fissures in the Panxian region
during the Middle Pleistocene. Dadong itself
was a large, high-ceilinged cavern by that time,
based on the dating of a substantial columnar
speleothem (stalacto-stalagmite) that had formed
in the interior (Shen et al., 1997). Young animals
may also have become mired after wandering
into the cave, but this seems unlikely during
the times when humans were active there. Any
of these natural processes whereby rhinoceroses
could have been added to the Dadong assemblage
would have produced a different signature from
what is found – namely, more complete and
extensive representation of skeletal elements.
The Dadong rhinoceros specimens show a
variety of breakage and damage that suggests
they were subject to minimal destructive or
natural transport processes. A small number of
bones appear chalky or surface weathered, and
five elements display trampling damage. These
changes may reflect diagenic processes as well
as physical destruction. The low frequency of
polish and rounding that can be attributed to
water action (10 elements, or 3.5%) does not
provide support for the idea that water transport
was particularly important in Dadong. Overall,
most of the rhinoceros sample was not substan-
tially altered by natural processes, suggesting
that these activities were not important factors in
the formation of the assemblage. This is also true
for the total faunal sample (c.f. Schepartz et al.,
Furthermore, the rhinoceros sample shows
minimal post mortem damage from animal proces-
sing. Rodents collect faunal elements that they
gnaw to maintain their incisor functioning.
Porcupines and small rodents were active in
Dadong, especially during later time periods, but
their damage to bones and tooth roots was fairly
limited, affecting approximately 5% of the total
sample (Schepartz et al., 2003). Only 10% of the
rhinoceros sample exhibits rodent gnawing. This
proportion is actually quite low when compared
with faunal assemblages from other localities in
the general region, such as Lang Trang from
Vietnam, where almost all of the teeth exhibit
gnawing damage (Vu The Long et al., 1996).
Big and small cats such as Chinese wild cats and
tigers, hyenas, foxes and other small carnivores
were present in Dadong. These species are rare
(4% of the sample identifiable to taxon), and
there are few that would have preyed upon
animals the size of rhinoceroses. Scavenging
carnivores, such as hyenas, could have trans-
ported portions of younger and smaller animals
into the cave. This might have been particularly
true in the case of elements that are left at kill sites,
such as crania or lower limbs (c.f. Blumenshine,
1986; Lyman, 1994). Hence, scavenging might
account for the high proportion of teeth in the
assemblage, if scavengers were able to detach
crania or mandibles. The numbers of foot
elements might also be explained.
There is very limited evidence for carnivore
damage on the Dadong bones (although this
probably under-represents the possible extent of
carnivore activity in the generation of bone
assemblages; c.f. Lyman, 1994). One bone clearly
displays carnivore chewing damage. The possible
role of carnivores in the generation of the sample
can also be evaluated by examining the ratio of
carnivores to ungulates and herbivores. Carni-
vore-generated assemblages, such as den sites,
ordinarily contain high frequencies of carnivore
bone (Brain, 1981). The Dadong faunal sample
falls outside the range of values observed for
carnivore-generated den deposits (at 4.6%, or 55
carnivores to 1176 herbivores and ungulates, of
the elements identified to taxon). The scarcity of
carnivores at Dadong is notable, as it is a cave
with side chambers and crevices suitable for
denning. It could be that the current faunal
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
L. A. Schepartz and S. Miller-Antonio
sample represents an area of the cave that was not
extensively used by the carnivores, or that
carnivore presence in the cave alternated with
human activity phases (c.f. Stiner, 2002).
Evidence for hominin activity and selection
for prime age adults
Although the above factors may have contributed
to the formation of the Dadong faunal assem-
blage, we suggest that none of them were the
single or primary source of accumulation. As the
material is found in association with stone tools
and hominin fossils, humans probably played
some role in their accumulation. There is an
excellent example of human-produced modifi-
cation on a distal portion of a rhinoceros radius
that has a bashed concentric fracture in the centre
of the shaft (Figure 12). Bones from other species
also have stone tool cutmarks, impact fractures,
and burning that we attribute to hominin
activities. Burning is the most common alteration,
occurring on 3% of the bones. Although it is
impossible to rule out other explanations for the
impact fractures and burning, the co-occurrence
of these features with cutmarks (and the lack of
their co-occurrence with carnivore damage)
strengthens our interpretation of them as result-
ing from hominin behaviour.
Evaluating the human role is complicated, as
they could have acted as hunters or scavengers.
The large size of rhinoceroses means that any
processing of their carcasses might entail
behaviours that reduce the carrying load. For
example, data on recent elephant hunters verify
that they spend considerable time and energy
processing at the kill site. They consume on-site,
or smoke and dry meat to decrease its carrying
weight (Fisher, 1993). If the Dadong rhinoceroses
died or were killed away from the cave, which is
highly probable, the problem of transport could
result in similar signatures for human hunting and
human or carnivore scavenging.
The age structure of the rhinoceros dental
sample, with its predominance of prime age
animals’ teeth, sheds light on the possible role of
hominins in its formation. Numerous studies of
fossil faunal assemblages attribute profiles domi-
nated by prime age adults to human activity
(Gaudzinski & Roebroeks, 2000; Steele, 2003).
For several species, the fact that the more difficult
to procure animals are targeted has been
explained as the result of exploiting them for
special resources such as antler or tusk.
Tong (2001a) studied the fossil assemblage of
Dicerorhinus mercki from the Middle Pleistocene
Nanjing Man site in South China. He reported
that most of the rhinoceros remains were from
very young or juvenile animals (74% deciduous
teeth still preserved in maxilla or mandibular
bone). Tong argued that humans are the only
natural predators of rhinoceroses due to the large
body size of even the juvenile animals, which can
approach 500 kg (Bigalke et al., 1950, as cited by
Tong, 2001a). Thus he viewed the Nanjing
assemblages as the product of human hunting.
The age-at-death profile and taphonomy of the
rhinoceroses at Dadong contrast strongly with
that observed by Tong. We have isolated teeth
from adult animals and very few cranial fragments
Figure 12. Distal radius with concentric impact fracture.
This figure is available in colour online at www.interscience.
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
Rhinoceros sinensis at Panxian Dadong, China
with or without dentition. If human activities are
responsible for the rhinoceroses in the Dadong
assemblage, a very different selective strategy is
involved. Given the distribution of skeletal
elements in our rhinoceros sample, we cannot
make a compelling argument for active hunting
rather than scavenging by hominins. However,
we suggested elsewhere that the large molar teeth
may have been a raw material for tools (Miller-
Antonio et al., 2000). It follows that the overall
larger size and greater crown height of young
adult teeth would make them best suited for this
purpose. Sizeable mandibles are often smashed
for their marrow at kill or scavenging sites, and
that might explain their limited representation at
Dadong. This could also explain why there is a
scarcity of cranial remains but an abundance of
teeth in the assemblage (Schepartz et al., 2003).
The small number of teeth and bones from
neonates, young animals less than 1.5 years of
age, and young juvenile rhinoceroses at Dadong
may reflect a taphonomic bias against the preser-
vation of more fragile elements (c.f. discussion in
Steele, 2003). A comparison between the rhino-
ceros and stegodont representation at Dadong is
informative with regard to this issue. The
distribution of skeletal elements is extremely
similar for both of these large-bodied species, yet
their dental mortality profiles are very different as
few adult stegodonts are represented (Schepartz
et al., 2005). The preservation of so many young
stegodont remains indicates no selective bias
against individuals from the youngest age classes,
and supports our interpretation of deliberate
procurement of elements from adult rhinoceroses.
The analysis and interpretation of the Panxian
Dadong rhinoceros dental and skeletal material
illustrates that the sample consists largely of teeth
from prime age adults (almost 50%), although
individuals from all age cohorts (neonates and
animals under 1.5 years, early juveniles, late
juveniles, subadults and adults) are represented.
This pattern deviates from the attritional distri-
bution expected from natural mortality of a wild
population. It also does not fit the catastrophic
death assemblage pattern.
Based on the available archaeological, faunal
and taphonomic data, the rhinoceros faunal
assemblage at Dadong is predominantly a result
of hominin activities (either hunting or scaven-
ging). Almost certainly, other collection agents
(carnivore, rodent, or geological processes)
played some role in its formation. The rhinoceros
assemblage is distinctive in that it differs from the
observed pattern at the Nanjing Man site and also
the mortality profile for Stegodon orientalis at
Dadong, both of which are characterised by a
higher prevalence of younger animals. This
unexpected finding is supportive of our earlier
hypothesis that rhinoceros teeth were sought for
use as a raw material for tool-making. This study
supports the results of other taphonomic work
(Schepartz et al., 2003) that suggests that humans
played an important role in the formation of the
Dadong faunal assemblage.
This work was supported by the National Geo-
graphic Society #7786-05, the National Science
Foundation BCS-9727688, the L.S.B. Leakey
Foundation, the Wenner-Gren Foundation for
Anthropological Research, the Henry Luce
Foundation, the University of Cincinnati
Research Council and the Charles P. Taft Fund,
and California State University, Stanislaus. We
thank Tong Haowen for generously sharing his
expertise on rhinoceros dentition, Takis Karkanas
for his photographic expertise, and Huang Wei-
wen, Takis Karkanas, Kanani Paraso, Hou Yamei,
Wang Wei, Si Xinqiang and Liu Jun for their
suggestions and contributions to the fieldwork at
Dadong. We also thank an anonymous reviewer
for their very careful reading of the manuscript
and their insightful comments.
Ballenger L, Myers P. 2001. ‘‘Rhinocerotidae’’ (on-line),
Animal Diversity Web. Available at http://
information/Rhinocerotidae.html [accessed 27
December 2007].
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
L. A. Schepartz and S. Miller-Antonio
Bekken D, Schepartz LA, Miller-Antonio S. 2004.
Taxonomic abundance at Panxian Dadong, a
Middle Pleistocene cave in south China. Asia in
the Middle Pleistocene, Asian Perspectives 43(2): 333–359.
Bigalke R, Steyn T, de Vos D, de Waard K. 1950.
Observations on a juvenile female square-lipped or
white rhinoceros (Ceratotherium simum simum Burchell)
in the national Zoological Gardens of South Africa.
Proceedings of the Zoological Society of London 120: 519–
Blumenshine RJ. 1986. Carcass consumption sequences
and the archaeological distinction of scavenging and
hunting. Journal of Human Evolution 15: 639–659.
Borsuk-Bialynicka M. 1973. Studies on the Pleistocene
rhinoceros Coelodonta antiquitatis (Blumenbach).
Palaeontologica Polonica 29: 9–19, XXIII.
Brain CK. 1981. The Hunters or the Hunted?. University of
Chicago Press: Chicago.
Colbert EH, Hooijer DA. 1953. Pleistocene mammals
from the limestone fissures of Szechwan, China.
Bulletin of the American Museum of Natural History 101:
Fisher JW Jr. 1993. Foragers and farmers: material
expressions of interaction at elephant processing
sites in the Ituri forest, Zaire. In From Bones to Behavior:
Ethnoarchaeological and Experimental Contributions to the
Interpretation of Faunal Remains, Hudson J (ed.). Center
for Archaeological Investigations, Occasional Paper
No. 21: Carbondale, IL; 247–262.
Foster JB. 1965. Mortality and ageing of black rhino-
ceros in east Tsavo Park, Kenya. East African Wildlife
Journal 3: 118–119.
Gaudzinski S, Roebroeks W. 2000. Adults only. Rein-
deer hunting at the Middle Palaeolithic site Sal-
zgitter Lebenstedt, northern Germany. Journal of
Human Evolution 38(4): 497–521.
Goddard J. 1970. Age criteria and vital statistics of a
black rhinoceros population. East African Wildlife
Journal (African Journal of Ecology) 8: 105–121.
Groves CP. 1967. On the rhinoceroses of southeast
Asia. Sa
¨ugetierkundliche Mitteilungen, Mu
¨nchen 15: 221–
´rin C. 1980. Les rhinoce
´ros (Mammalia, Perisso-
dactyla) du Mioce
`ne terminal au Ple
`ne supe
ieur en Europe occidentale. Documentes Laboratoire
´ologie Lyon 79: 1–1185.
Haynes G. 1991. Mammoths, Mastodonts, and Elephants.
Cambridge University Press: Cambridge.
Huang WW, Si XQ, Hou YM, Miller-Antonio S,
Schepartz LA. 1995. Excavations at Panxian
Dadong, Guizhou province, southern China. Current
Anthropology 36: 844–846.
Jones HL, Rink W, Schepartz LA, Miller-Antonio S,
Huang WW, Hou Y, Wang W. 2004. Combined
electron spin resonance (ESR)/uranium-series dat-
ing of mammalian tooth enamel at Panxian Dadong,
Guizhou province, China. Journal of Archaeological
Science 31: 965–977.
Karkanas P, Schepartz LA, Miller-Antonio S, Wang
W, Huang WW. 2008. Late Middle Pleistocene
climate in southwestern China: inferences from the
stratigraphic record of Panxian Dadong Cave, Guiz-
hou. Quaternary Science Reviews 27: 1555–1570.
Klein RG. 1982. Age (mortality) profiles as a means of
distinguishing hunted species from scavenged ones
in Stone Age archaeological sites. Paleobiology 8:
Koulmann M, Koulmann J. 1970. Dentition et evol-
ution d’un mammal: le rhinoce
´ros. Revue de Stomato-
logie et de Chirurgie Maxillo-faciale 71: 133–138.
Louguet S. 2006. Determining the age of death of
Proboscids and Rhinocerotids from dental attrition.
In Recent Advances in Ageing and Sexing Animal Bones,
Ruscillo D (ed.). 9
International Council for Archaeo-
zoology Conference, Durham 2002: Oxford; 179–188.
Lyman RL. 1994. Vertebrate Taphonomy. Cambridge Man-
uals in Archaeology. Cambridge University Press:
Miller-Antonio S, Schepartz LA, Bakken D. 2000. Raw
material selection and evidence for rhinoceros tooth
tools at Dadong Cave, southern China. Antiquity 74:
Miller-Antonio S, Schepartz LA, Karkanas P, Hou Y,
Huang W, Bekken D. 2004. Lithic raw material use
at Dadong. Asia in the Middle Pleistocene, Asian Perspect-
ives 43: 314–332.
Osborn HF. 1903. The extinct rhinoceroses. Memoirs of
the American Museum of Natural History 1(3): 75–164.
Pan YR, Yuan CW. 1997. Pleistocene primates from
Panxian Dadong, Guizhou province. Acta Anthropo-
logica Sinica 16: 201–208. (in Chinese with English
Rink WJ, Schepartz LA, Miller-Antonio S, Huang W,
Hou Y, Bakken D, Richter D, Jones HL. 2003.
Electron spin resonance (ESR) dating of tooth
enamel at Panxian Dadong cave, Guizhou, China.
In Current Research in Chinese Pleistocene Archaeology,
Shen C, Keates SG (eds). BAR Monograph Series:
Oxford; 95–106.
Schepartz LA, Bakken D, Miller-Antonio S, Paraso K,
Karkanas P. 2003. Faunal approaches to site for-
mation processes at Panxian Dadong. In Current
Research in Chinese Pleistocene Archaeology, Shen C,
Keates SG (eds). BAR Monograph Series: Oxford;
Schepartz LA, Stoutamire S, Bekken DA. 2005. Stego-
don orientalis from Panxian Dadong, a Middle Pleis-
tocene archaeological site in Guizhou, south China:
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
Rhinoceros sinensis at Panxian Dadong, China
taphonomy, population structure and evidence for
human interactions. Quaternary International 126–
128: 271–282.
Shen GJ, Liu J, Jin LH. 1997. Preliminary results on U-
series dating of Panxian Dadong in S-W China’s
Guizhou province. Acta Anthropologica Sinica 16(3):
1–8 (in Chinese with English abstract).
Steele TE. 2003. Using mortality profiles to infer
behavior in the fossil record. Journal of Mammalogy
84: 418–430.
Stiner MC. 2002. Carnivory, coevolution, and the
geographic spread of the genus Homo.Journal of
Archaeological Research 10: 1–63.
Tong H. 2000. Les Rhinoce
´ros des sites a
humains de Chine. L’Anthropologie 104: 523–
Tong H. 2001a. Age profiles of rhino fauna from the
Middle Pleistocene Nanjing Man site, South China
– explained by the rhino specimens of living
species. International Journal of Osteoarchaeology 11:
Tong H. 2001b. Rhinocerotids in China – systematics
and material analysis. Geobios 34: 585–591.
Vu The Long. de Vos J, Ciochon RL, 1996. The fossil
mammal fauna of the Lang Trang Caves, Vietnam,
compared with southeast Asian fossil and recent
mammal faunas: the geographical implications.
IPPA Bulletin 14: 101–109.
Zhang ZH, Liu J, Zhang HG, Yuan CW. 1997.
A Pleistocene mammalian fauna from Panxian
Dadong, Guizhou Province. Acta Anthropologica Sinica
16: 207–220 (in Chinese with English abstract).
Copyright #2008 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2008)
DOI: 10.1002/oa
L. A. Schepartz and S. Miller-Antonio
Full-text available
Caune de l’Arago is a Middle Pleistocene site in Southern France, where Acheulean artefacts and hominin fossils were excavated. Rhinoceros (Stephanorhinus hemitoechus) remains from Level F (MIS 12) were studied from a zooarchaeological and taphonomic perspective to investigate the potential human exploitation of this large taxon. The well‐represented butchering marks, as well as scarce carnivore marks, indicate primary access to rhinoceros carcasses by hominins. The juvenile‐dominated mortality profile further suggests aggressive scavenging or occasional opportunistic hunting. Furthermore, differential skeletal representation shows that humans selectively transported the nutrient‐rich body sections. In addition, thorough processing of the carcasses for consumption inside the cave is highlighted by the frequent cut marks, intensive fragmentation and regular spatial distribution. The analyses of the Arago Level F rhinoceros confirm the rhinoceros exploitation by humans during the Lower Paleolithic; however, the rhinoceros mortality profile differs from that of large mammals such as equids, argali and all other game species of this archaeological layer. Caune de l’Arago is the earliest hominin site where the systematic exploitation of rhinoceros has been documented.
Full-text available
Some ungulate species at Upper Pleistocene and Holocene archeological sites in South Africa exhibit catastrophic mortality profiles, while others exhibit attritional ones. The awareness by Stone Age people that some species are especially amenable to driving or snaring probably accounts for the catastrophic profiles. Natural catastrophic death immediately followed by human scavenging is a much less likely explanation because the species samples comprise material lumped from deposits that accumulated more or less continuously over hundreds or even thousands of years during which period there is no reason to suppose the repeated occurrence of natural catastrophes nearby. The inability of Stone Age people to obtain prime-age adults in species that are not particularly amenable to driving or snaring presumably accounts for the attritional mortality profiles. Although the species that display attritional profiles conceivably were scavenged, the high proportion of very young individuals in the profiles suggests active hunting. Very young individuals are much less abundant in attritional profiles from local non-archeological sites, probably because their carcasses were removed from the record before burial, primarily by carnivore or scavenger feeding. Scavenging would account for the abundance of very young individuals in the archeological sites only in the unlikely event that people could regularly locate carcasses before other predators did. In general, geomorphic/sedimentologic context is probably the best criterion for determining whether a species characterized by a catastrophic profile in an archeological site was hunted or scavenged. At the majority of known sites, active hunting is suggested. In the case of a species characterized by an attritional profile in an archeological site, the proportion of very young individuals in the sample probably provides the best criterion for distinguishing hunting from scavenging. A relatively high proportion of very young individuals suggests active hunting.
The diminishing population of African and Asian elephants can be compared to the extinction of other elephant-like species, such as mammoths and mastodonts, which occurred more than ten thousand years ago. The purpose of this book is to use the ecology and behavior of modern elephants to create models for reconstructing the life and death of extinct mammoths and mastodonts. The source of the models is a long-term and continuing study of elephants in Zimbabwe, Africa. These models are clearly described with respect to the anatomical, behavioral, and ecological similarities between past and present proboscideans. The implications of these similarities on the life and death of mammoths and mastodonts is explored in detail. The importance of this book is primarily its unifying perspective on living and extinct proboscideans: the fossil record is closely examined and compared to the natural history of surviving elephants. Dr. Haynes's studies of the places where African elephants die (so-called elephant burial grounds) are unique.