Functional Characterization of Ungulate Molars Using the Abrasion-Attrition Wear Gradient: A New Method for Reconstructing Paleodiets
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 1
Allen Press • DTPro System File # 01cc
P U B L I S H E D B Y T H E A M E R I C A N M U S E U M O F N AT U R A L H I S T O R Y
CENTRAL PARK WEST AT 79TH STREET, NEW YORK, NY 10024
Number 3301, 36 pp., 26 figures, 4 tables October 16, 2000
Functional Characterization of Ungulate Molars
Using the Abrasion-Attrition
Wear Gradient: A New Method for
Reconstructing Paleodiets
MIKAEL FORTELIUS1 AND NIKOS SOLOUNIAS2
ABSTRACT
The analysis of fossil ungulate cheek teeth has long been one of the main sources of
information about the terrestrial environments of the Cenozoic, but the methods used to extract
this information have been either imprecise or prohibitively laborious. Here we present a
method based on relative facet development that is quantitative, robust, and rapid. This method,
which we term mesowear analysis, is based on the physical properties of ungulate foods as
reflected in the relative amounts of attritive and abrasive wear that they cause on the dental
enamel of the occlusal surfaces. Mesowear was recorded by examining the buccal apices of
molar tooth cusps. Apices were characterized as sharp, rounded, or blunt, and the valleys
between them either high or low. The method has been developed only for selenodont and
trilophodont molars, but the principle is readily extendable to other crown types. Mesowear
analysis is insensitive to wear stage as long as the very early and very late stages are excluded.
Cluster analysis of the mesowear variables produces clusters reflecting four main groups
from abrasion-dominated to attrition-dominated: grazers, graze-dominated mixed feeders,
browse-dominated mixed feeders, and browsers. Most of the relatively few apparent anomalies
are explained by more detailed dietary information. Mesowear analysis provides resolution
within the main dietary classes and the clustering is virtually identical with and without the
index of hypsodonty. Discriminant analysis using all mesowear variables and hypsodonty
1 Professor of Ecological Paleontology, Department of Geology, Division of Geology and Paleontology, PO Box
11 FIN-00014 University of Helsinki, Finland, mikael.fortelius@helsinki.fi
2 Research Associate, Department of Paleontology, American Museum of Natural History and Associate Professor,
Department of Anatomy, New York College of Osteopathic Medicine, New York Institute of Technology, Old West-
bury, NY 11568 8000 USA, nsolouni@iris.nyit.edu
Copyright American Museum of Natural History 2000 ISSN 0003-0082 / Price $4.50AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 2
Allen Press • DTPro System
File # 01cc
2 AMERICAN MUSEUM NOVITATES NO. 3301
showed an overall correct classification of 76% of 64 species of living ungulates into the
conventional dietary categories of browser, grazer, and mixed feeder, while a smaller set of
27 ‘‘typical’’ species was correctly classified at 96%. Alternative ‘‘conservative’’ and ‘‘radi-
cal’’ dietary classifications that were employed to accommodate cases where dietary infor-
mation was controversial or unclear produced only marginally different results. Mesowear
analysis successfully resolved a test case using the Serengeti grazing succession and appears
to be superior to microwear analysis in two cases where the diet of fossil ungulates has been
previously studied by microwear and other conventional methods.
INTRODUCTION Perhaps a more satisfactory grouping could
be based on the time scale involved: the un-
Because of their ubiquity and close rela- worn (preformed) morphology of the tooth
tionship to understandable physical and bio- reflects evolution in deep geological time, the
logical relationships, fossil herbivore teeth wear pattern visible to the naked eye (or
are commonly used to reconstruct the diet preferably low magnification) reflects a sub-
and environment of extinct species. The main stantial portion of the individual’s life in eco-
limitation so far has been the fact that de- logical time, and the microscopically visible
tailed and reliable dietary interpretation from details (microwear) reflect a time that is short
mammalian herbivore teeth has involved ex- even in comparison with the individual’s life
pensive, laborious, and time-consuming
span. None of the levels is inherently more
methods, which have made it difficult to ap-
important than the others, but each answers
ply them to more than a few species from a
somewhat different questions; considering
limited number of localities (e.g., Rensber-
each level separately will improve functional
ger, 1973; Fortelius, 1982; Rensberger et al.,
interpretation.
1984; Teaford and Walker, 1984; Fortelius,
1985; Janis and Fortelius, 1988; Janis, 1990; The unworn morphology, expressed for
Solounias and Moelleken, 1992a, 1992b; So- example in classic terms such as hypsodont
lounias and Hayek, 1993; Hunter and For- or lophodont, reflects long-term adaptation
telius, 1994; Solounias et al., 1994, 1995). and sets the main mechanical constraints on
We present a greatly simplified procedure for what foods an individual can hope to suc-
rapidly characterizing the major (selenodont cessfully utilize. Masticatory morphology,
and trilophodont) morphological types of hypsodonty, tooth structure and enamel
molar teeth found among Cenozoic large structure belong to the preformed adaptations
mammalian herbivores in relationship to a in deep time. The information provided by
functional and ecological framework. The ul- preformed morphology is too general, how-
timate purpose is to develop a reliable and ever, to resolve morphologically homoge-
inexpensive method to rapidly analyze large neous groups such as the selenodont artio-
numbers of extinct taxa in museums so that dactyls (e.g., Jernvall et al., 1996). At the
entire ungulate paleocommunities may be other extreme, microwear provides direct in-
studied effectively. formation about the nature of the last few to
The dietary interpretation of mammalian several meals of an individual (Solounias et
teeth has traditionally involved either direct al., 1994; Teaford and Oyen, 1989).
(actualistic) comparison with living animals, In many ways it is the intermediate level,
the application of general functional princi- which we term mesowear, that seems to offer
ples, or—increasingly during the latest few the best hope for answering questions relat-
decades—the study of the wear patterns left ing to the average diet of a particular species
on teeth by food. The distinction between from a particular location in space and time
these approaches is somewhat blurred, how- (fig. 1). Mesowear patterns are best described
ever, for most actualistic comparisons since in terms of facet development (Butler, 1952;
Cuvier, have in fact contained a functional Janis, 1990), and best understood as the com-
element (Van Valkenburgh et al., 1990), and bined result of both attrition and abrasion,
wear patterns are almost unavoidably includ- i.e., the relative contributions of tooth-on-
ed in any morphological comparison of teeth. tooth and food-on-tooth wear (Butler, 1972;AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 3
Allen Press • DTPro System
File # 01cc
2000 FORTELIUS AND SOLOUNIAS: UNGULATE MOLARS 3
the attrition signal is partly or completely
masked by abrasion. Mesowear is best ob-
served with the naked eye or at low magni-
fication, for example with a hand lens. It af-
fects all occlusal surfaces of teeth but in the
present study we restrict our observations to
the buccal edges of the paracones and meta-
cones of upper molars as two variables ob-
servable from the buccal side: cusp relief and
cusp shape.
Cusp relief is to the relative difference in
height between cusp tips and inter cusp val-
leys as seen in buccal projection, i.e., how
high the cusps appear in lateral view. We will
show that there is strong empirical evidence
that the combination of absolutely high attri-
tion and abrasion, as seen in extreme grazers,
results in low occlusal relief. The theoretical
explanation for this relationship remains
somewhat obscure but the high relief typi-
cally seen in ‘‘fresh grass’’ grazers, like Ko-
bus ellipsiprymnus or Redunca redunca, sug-
gests that lower occlusal stress permits a
higher occlusal relief to develop.
Fig. 1. Mesowear features used in this study Cusp shape refers to the apex of the cusp
and a sample of hypsodont, mesodont, and (here paracone or metacone) described as
brachydont teeth. A. Capra hircus, goat, Basel sharp, rounded, or blunt, in decreasing order
Museum 3379 (tooth relief high and cusps sharp). of facet development. The theoretical basis is
B. Cervus duvaucelli AMNH 54498 (mesodont, easily understood in terms of the relative
tooth relief high and cusps rounded). C. Odoco- contribution of attrition and abrasion to the
ileus virginianus, white tailed deer, Canadian
Royal Museum Toronto 2091 (brachydont, tooth
total wear: sharp cusps mean that attrition
relief high and cusps sharp). D. Equus caballus, predominates strongly, whereas the attrition
domestic horse personal collection (hyperhypso- signal is almost completely masked by abra-
dont, tooth relief low and cups blunt; actually sion in blunt cusps.
concave and thus blunter than other grazers do). The ordinary dental wear visible to the na-
E. Kobus ellipsiprymnus (NS, personal collection) ked eye or at low magnification has not been
(tooth relief high and cups rounded). F. Alcela- a favorite target of research among mam-
phus buselaphus (NS, personal collection) (tooth malian paleontologists. Every (1970) and
relief low and cusps blunt). Height of relief is Every and Kühne (1971) attributed attritional
shown in relation to the actual length of cusps. wear to ‘‘thegosis,’’ or the active sharpening
of teeth not associated with chewing food.
Fortelius, 1985). The fibrous, nonbrittle plant There seems to be no evidence for such be-
foods that make up the diet of most selen- havior, however, nor any need to invoke it to
odont and lophodont ungulates can only be explain dental wear patterns (Osborn and
cut when the cutting edges push clean Lumsden, 1978). More relevant to the pre-
through the food, so that the occluding dental sent purpose, Guthrie (1990) reported differ-
facets come into direct contact (Rensberger, ences in the teeth of bison between sedge-
1973; Lucas, 1979; Walker, 1984). This eating and grass-eating individuals, and Janis
means that direct tooth-on-tooth wear (attri- (1990) used mapping of ‘‘gross dental wear’’
tion) will always occur in these forms, a fact to deduce diet in fossil mammals, based on
that explains why precise occlusion between comparison with Recent species. The method
upper and lower teeth is maintained until the we describe here uses the same principle,
very latest wear stages even in species where stripped to its barest essentials.AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 4
Allen Press • DTPro System
File # 01cc
4 AMERICAN MUSEUM NOVITATES NO. 3301AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 5
Allen Press • DTPro System
File # 01cc
2000 FORTELIUS AND SOLOUNIAS: UNGULATE MOLARS 5AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 6
Allen Press • DTPro System
File # 01cc
6 AMERICAN MUSEUM NOVITATES NO. 3301
MATERIALS former problem is addressed by the present
study as a whole, but the latter requires spe-
We used extant species to develop the new
cial attention. It was surprisingly difficult to
mesowear analysis and we tested the method
find large populations of individuals of well-
with both extant and extinct species. A sam-
known age to investigate the effect of wear
ple of extant ungulate species was used to
stage. We were able, however, to study a
examine the stability of mesowear during on-
population of 36 reindeer from Canada (a
togeny (mesowear stability test) for a test
personal collection) for which age is known
case using the Serengeti grazing succession
from annuli in the lower first molar roots (see
and for two pilot applications on fossil
Solounias et al., 1994, for more details) (ta-
equids and bovids.
ble 2). Individuals from six additional spe-
cies were selected (table 2): the plains zebra
DATABASE
aged by the degree of incisor wear (following
A comparative database was developed for methods of Bone, 1964; Klingel, 1965; Slade
2200 individuals representing upper molar and Godfrey, 1982) omson’s gazelle, bohor
teeth from 64 extant ungulate species (table reedbuck, bushbuck, and black-fronted dui-
1). The material was studied in the mam- ker by number of keratinous rings on horns,
malogy collections of several museums (see size of horns, and second molar height; and
Acknowledgments). The variables included African buffalo by size and curvature of
in table 1 are explained below and under horns (following a study for the springbuck,
Methods. The raw data are available on re- Rautenbach 1971).
quest from either author.
UNGULATE DIETS—DERIVING TABLE 1
SAMPLES FOR MESOWEAR STABILITY TEST
Several factors can complicate the rela-
Mesowear analysis is applicable only if tionship of diets to morphology in extant
factors related to diet have a significantly species. Although investigators who study
stronger influence on wear patterns than fac- the evolution and adaptation of ungulates
tors related to structure or wear stage. The have related morphology to reported dietsAMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 7
Allen Press • DTPro System
File # 01cc
2000 FORTELIUS AND SOLOUNIAS: UNGULATE MOLARS 7
(e.g., Webb, 1983; Fortelius, 1985; Janis, cases are treated as intermediate (mixed feed-
1988; Solounias et al., 1994), it is clear that er) and a ‘‘radical’’ classification (radi)
diets reported in the literature are often based where such cases are treated as extreme
on small samples from single populations, (browser or grazer). Comparison of results
and that diet would be best studied by exten- from the two classifications enables us to
sive field observation or fecal collections. gauge the effect of such subjective choices
Many ungulates are also opportunistic, and and to evaluate their impact on the results.
their diets vary by place and by season, so The variable Jad1 gives the diet as reported
that what can be observed for a few years by Janis (1988); Jad2 gives the correspond-
and in certain places may not hold for the ing value translated to the simple browser-
thousands to millions of years that a partic- grazer-mixed feeder classification. Our con-
ular species existed. We acknowledge that di- servative dietary classification differs from
etary information for Recent species is often that of Janis (1988) for three species: Antil-
less than perfect but cannot suggest any im- ocapra americana, Capreolus capreolus, and
mediate remedy. Furthermore, for the meth- Procavia capensis, but agrees with the pre-
od presented here, even a rough classification sent opinion of Dr. Janis (personal commun.,
based on physical properties of plant foods July 1999). Table 1 also features the ad hoc
would be preferable to one based on system- classifying variable Class, with four values:
atics or general appearance. fo (fossil), no (no particular class), mb (‘‘ma-
General dietary classes (browser, grazer, bra,’’ minute abraded brachydont, identified
mixed feeder) are problematic for two rea- under Results, Cluster Analysis), and ty (typ-
sons. They lump a diversity of foods and life- ical of its dietary class).
styles under three broad categories, and sci-
entists have used different definitions for TEST CASE: A KNOWN DIETARY SUCCESSION
them, not often explicitly stated. However, IN THE SERENGETI
these dietary classes are so firmly entrenched
We tested this method on a grazing suc-
in the literature that any new approach must
cession from the Serengeti (Bell, 1971), in-
somehow relate back to them. Although our
volving plains zebra, topi, wildebeest, and
study emphasizes a continuum—a spectrum
sometimes the hartebeest and Thomson’s ga-
of diets based on their mechanical (wear)
zelle. Bell observed that the zebra feeds on
properties in terms of abrasion and attri-
the rough higher grasses as it migrates. The
tion—we have chosen this general classifi-
removal of these grasses by the zebra enables
cation as the base against which to compare
the topi, the wildebeest, and the hartebeest to
the results of this study. We use the conven-
follow and feed on lower grasses. Finally, the
tional 90% cutoff points to define the classes:
gazelle follows, feeding on the smallest and
browsers take ⬍10% grass, grazers ⬍10%
softest vegetation. The precise sequence of
browse. The huge spectrum between these
this succession is essential for these species
extremes is lumped into the obviously het-
to find their required vegetation. We predict
erogeneous mixed feeder category.
that the mesowear will replicate this succes-
Table 1 lists the basic dietary classifica-
sion such that the zebra and the gazelle will
tions compiled from Janis (1988) and several
bracket the other three species.
other sources (in particular: Tener, 1965;
Schaller, 1967; Hofmann, 1973; Labâo-Tello
FOSSIL SAMPLES
and Van Gelder, 1975; Schaller, 1977; Sin-
clair, 1977; Gauthier-Pilters and Dagg, 1981; FOSSIL BOVIDS: Pachytragus laticeps and
McDonald, 1981; Chapman and Feldham- Pachytragus crassicornis are two species of
mer, 1982; Kingdon, 1982a; 1982b; Nowak early Caprini (Bovidae) from the Miocene of
and Paradiso, 1983; Hofmann, 1985; Schall- Samos collected by Brown in 1924 (Gentry,
er et al., 1986; Hofmann, 1989; Nowak, 1971; Solounias, 1981). P. laticeps has lon-
1991). In recognition of some uncertainty ger and more backwardly curved horn cores
and ambiguity we have used two parallel than P. crassicornis, which is also smaller
classifications of our 64 species: a ‘‘conser- and has a more posteriorly placed toothrow
vative’’ classification (cons) where doubtful (Gentry, 1971). Based on general consider-AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 8
Allen Press • DTPro System
File # 01cc
8 AMERICAN MUSEUM NOVITATES NO. 3301
ation of the cranial morphology and denti- in order to emphasize the shape of the tooth
tion, Gentry (1971) suggested that P. cras- outlines. Parallax due to photography was ig-
sicornis was more advanced and adapted to nored as we only scored qualitatively three
a harsher environment than P. laticeps. In general categories. We only used specimens
contrast, Solounias and Moelleken (1992b) in which the last molar was in occlusion and
determined from a tooth microwear analysis the first molar retained an occlusal shape
that P. laticeps was a grazer and P. crassi- similar to the second molar. Consequently,
cornis a mixed feeder. Based on masseter at- the effect of age was minimized, as discussed
tachment morphology and using the mor- further under Results. After experimenting
phology of the masseter muscle attachments, with various choices and standards, we set-
Solounias et al. (1995) concluded that both tled on the sharper buccal cusp of the second
species were mixed feeders and suggested upper molar; that is, either the paracone or
that P. laticeps was further removed from the metacone. This was done for simplicity, and
browsing mode than P. crassicornis. We we would like to stress that our experience
have collected mesowear data for these spe- strongly indicates that the choice of molar
cies to see whether this problem can be re- buccal cusp is not critical. (For example, the
solved. paracone and metacone are usually identical
FOSSIL EQUIDS: Hayek et al. (1992) used in mesowear of a single individual.) The se-
microwear to study the paleodiet of the fol- lection of the sharpest cusp will drive the
lowing fossil equids: Mesohippus insignis data slightly toward sharpness. This is a con-
from the Olcott Formation of Nebraska, Cor- servative decision, since after the very first
mohipparion goorisi from Trinity River Pit, wear stage, sharpness is never an artifact of
Flemming Formation, Cold Springs Texas, wear stage, whereas blunting may be so.
Cormohipparion quinni (formerly C. sphen- Occlusal relief was classified as high or
odus), from Valentine Formation, Cornell low, depending on how high the cusps rise
Dam Member Nebraska, and Cremohippa- above the valley between them (fig. 1). After
rion proboscideum Quarry 1 of Samos, some practice, simple scoring is sufficient,
Greece (Sondaar, 1971). We have collected but in borderline cases a quantitative index
mesowear data from the same populations in can be constructed as follows. The buccal
order to compare the results obtained by the profile of the tooth is projected onto a plane.
two techniques. The vertical distance between a line con-
All fossil samples were from the Division necting two adjacent cusp tips and two ad-
of Paleontology of the American Museum of jacent valley bottoms is measured, and divid-
Natural History (New York). ed by the length of the whole tooth (fig. 1).
For selenodont forms and plagiolophodont
METHODS equids, the limit between high and low was
arbitrarily set at 0.1, for hyracoids at 0.05,
SCORING ROUTINES FOR MESOWEAR
and for rhinoceroses at 0.03. These values
Although additional mesowear features were calibrated by the relief observed in the
can be defined (a larger and more detailed species included in the study, to separate the
study is in progress), the present study bases subjectively ‘‘low’’ from the subjectively
mesowear on two variables: occlusal relief ‘‘high.’’ Negative relief (cusp tip lower than
and cusp shape (fig. 1). Ungulate teeth were sides) was sometimes seen in hypsodont
inspected at close range, using a hand lens equids, and was treated as low. The progres-
when appropriate. The first several hundred sive blunting of a cusp (see below) will in-
specimens were photographed and traced on evitably reduce occlusal relief. That is to say,
paper, but once the standards had been set to cusp shape and relief are not entirely inde-
our satisfaction the rest of the material was pendent, but converge at the low and blunt
recorded by direct scoring. We provide sev- (‘‘grazer’’) end of the spectrum. We included
eral figures, simple outlines of molars, which occlusal relief, anticipating its usefulness for
show the buccal outlines of a selection un- the study of fossil forms, particularly non-
gulate teeth illustrating the various morphol- grazing hypsodont plagiolophodonts, which
ogies. Artistic shadow has not been included include many hipparions. Occlusal relief isAMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 9
Allen Press • DTPro System
File # 01cc
2000 FORTELIUS AND SOLOUNIAS: UNGULATE MOLARS 9
used in the analyses as a percentage and is for comparison. The logic here is simply that
given in table 1 as percent of high relief (per- the relationship between attrition and abra-
high). sion is studied separately for different total
Cusp shape was scored as sharp, rounded, wear regimes. However, for the cluster and
or blunt (fig. 1, table 1) according to the de- discriminant analyses, all crown heights were
gree of facet development. A sharp cusp has considered together.
(practically) no rounded area between the Hypsodonty indices (table 1, hypind) were
mesial and distal phase I facets, a rounded taken from Janis (1988): width divided by
cusp has a distinctly rounded tip without pla- the length of third lower molar (cement has
nar facet wear but retains facets on the lower been excluded in these measurements). For
slopes, while a blunt cusp lacks distinct fac- convenience, the species were partitioned
ets altogether. Cusp shape is also used as a into the conventional categories: brachydont
percentage and is given in table 1 as three (b), mesodont (m) or hypsodont (h). With
variables: persharp, perround, and perblunt. such a tripartite subdivision, the hypsodonty
Dental structure and phylogenetic history of most taxa can be readily determined by
obviously influence both occlusal relief and observation. It is primarily the mesodont spe-
cusp shape. The blunt cusps of true Bovini cies and those on the borderlines that need
and the low occlusal relief of present-day to be measured or carefully examined for
horses are at least partly the result of obvious correct assignment. Mesodont artiodactyls
structural modifications of the teeth. Apart are those with indices between 2.6 and 3.4
from calibrating the cut-off points for high and all perissodactyls and hyraxes with in-
and low relief for individual groups as ex- dices between 2.0 and 3.0. Brachydont artio-
plained above, we have not attempted to cor- dactyls have indices of 2.5 or lower, and hyp-
rect for these influences here, although this sodonts have indices of 3.5 or lower (table
may later become necessary. It is our con- 1, hyp). The camel was classified as hypso-
tention that these differences are due to adap- dont despite the unexpectedly low index re-
tive evolution and largely follow the pattern ported by Janis (1988), but the index report-
established by mesowear. They appear to am- ed was used in the analyses. Hog deer,
plify rather than disguise the mesowear sig- Grant’s gazelle, nyala, and the bushbuck are
nal, and are therefore not a problem unless all just below our limit for mesodonts. The
the resolution desired is very high. A brief takin, Thomson’s gazelle, and the common
‘‘how to do it’’ description of mesowear waterbuck are all just above our limit for
analysis is given within Methods. hypsodonts. Figure 1 shows examples of
brachydont, mesodont, and hypsodont teeth.
HYPSODONTY
STATISTICAL ANALYSES
The goal of our study is to develop a
method for the dietary interpretation of ex- The significance of differences observed
tinct species, and the currently best estab- in simple comparisons was tested using the
lished gross morphological predictor of diet Kruskal-Wallis test and the chi square test, as
in ungulates is hypsodonty (Janis, 1988), the appropriate. Hierarchical cluster analysis was
main aspect of functional durability in the performed for several sets of species, with
face of wear (Janis and Fortelius, 1988). Euclidean distance and complete linkage (to
Hypsodonty must obviously be included in enhance the distinctness of clusters), using
our study, but is perhaps best conceived of three mesowear variables with and without
as a proxy for overall wear rate (Solounias the index of hypsodonty. Three of the four
et al., 1994), rather than as a morphological mesowear variables are percentages of cusp
character as such, just as the mesowear var- sharpness and add up to 100; therefore a
iables are conceived as proxies for kinds of maximum of two of those were included for
wear. any analysis). Discriminant analysis was per-
In deciphering extinct species, the degree formed for single variables and for all com-
of hypsodonty can usefully be employed to binations of the mesowear variables plus the
select an appropriate subset of extant species index of hypsodonty, using two dietary clas-AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 10
Allen Press • DTPro System
File # 01cc
10 AMERICAN MUSEUM NOVITATES NO. 3301
sifications (conservative and radical) alter- couraged, for reasons exemplified by the
nately as grouping variable. We report the ‘‘mabra-syndrome’’ identified in this paper.
percentage of correct classifications from the An important part of more detailed meso-
jackknifed classification matrix (table 3). All wear analyses than those reported here will
statistical tests were performed on Systat 7.0 be the selection of the appropriate compara-
in a PC environment, using the default set- tive sample, much as one would do for es-
tings except where noted. timating body mass by regression techniques.
It would clearly not be wise to analyze a
PRACTICALITIES OF MESOWEAR DATA small and brachydont species in relation to a
COLLECTION AND ANALYSIS sample of large and hypsodont ones, for ex-
ample. The figures of teeth in this study are
Since this paper is intended to introduce offered as an aide to make a first rough as-
an easy and generally applicable method of sessment of the general mesowear regime of
paleodiet reconstruction we agree with one a sample.
of our reviewers (Dr. R. L. Bernor) that a
‘‘cookbook’’ section is warranted. RESULTS
The first step in mesowear analysis is ob-
STABILITY OFRELIEF AND CUSP SHAPE
viously to obtain the material. It is important
DURING WEAR
to select the specimens such that teeth in
very early and very late wear are excluded, REINDEER SPECIMENS: Ten young speci-
and to avoid scoring the shape of cusps that mens (49–51 months) showed a high consis-
are either damaged or modified by structural tency with 80% sharp cusps. The 16 older
elements (like the paracone of a rhinoceros). individuals (61–77 months) had 87% sharp
Pilot studies have revealed that lower teeth cusps and 12% rounded. The oldest 10 in-
consistently score more rounded than do up- dividuals, ranging 85–97 months, had 70%
pers, and we suggest that the two should not sharp and 30% rounded cusps. All specimens
be mixed. Using lower teeth requires build- were high in relief. Chi square tests of the
ing up a comparative sample of lower teeth. three groups gave low probabilities that the
We used only uppers in this paper. A sample three age classes are different from each oth-
of less than ten specimens should be treated er (p values ranged from 0. 552 to 0.187).
with caution, while more than 30 is probably We conclude that, although there is some in-
excessive. crease in roundedness with wear, the meso-
The scoring procedure itself is described wear signature is nevertheless reasonably
in the method section of this paper and is not maintained throughout an individual’s life
repeated here. There is no doubt that it can (table 2).
be refined, but care should be taken not to Six additional species were also selected
loose the generality of the method, since re- for the study of ontogenetic effects on the
stricting it to a single, morphologically uni- occlusal relief and cusp shape. All species
form group will severely limit the choice of showed that these parameters remain rela-
recent species available for comparison. Sev- tively stable from early middle wear until ad-
eral of our figures are intended as an aide for vanced age. In 24 Thomson’s gazelles with
scoring. Once the mesowear data exist as a 17 to 23 horn rings 83% of cusps were sharp.
file they can be analyzed by appropriate sta- Seven younger individuals had 71% sharp
tistical means (we are sure that both our cusps, while in 7 older individuals 29% of
choice and our application of methods can cusps were sharp. In the black-fronted duiker,
be improved upon!). The raw data allow the proportions of cusp shape (in this case,
comparisons between pairs or groups of spe- sharp and rounded) are reasonably stable
cies using (for example) the chi square test, throughout the life span of the individuals
while the discriminant and cluster analyses studied (table 2). Low relief and blunt cusps
require summary data in the form of (for ex- were found in 11 individuals of plains zebra
ample) percentages of occlusal relief and spanning 2–13 years (aged by incisor wear
cusp shape values. stages), showing that the bluntness is genu-
Blind reliance on statistics is to be dis- inely present in early as well as late wear.AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 11
Allen Press • DTPro System
File # 01cc
2000 FORTELIUS AND SOLOUNIAS: UNGULATE MOLARS 11
The crown height decreased from 13.0 to 8.0 sistent cusp shape morphology during wear
mm in the bushbuck and we found 75% and was commonly encountered during the re-
80% sharp cusps. High and rounded mor- cording of the large data set used in this
phology was found in four plains reedbucks study.
from horn stages 1–4 with 100% high and
rounded cusps. In three African buffaloes, CLUSTER ANALYSIS
horn stages 1–3, the crown height decreased
from 22.9 to 12.8 mm and the cusps were all Cluster analysis, using the index of hyp-
high and rounded (table 2). Similarly, con- sodonty and all mesowear variables exceptAMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 12
Allen Press • DTPro System
File # 01cc
12 AMERICAN MUSEUM NOVITATES NO. 3301
Fig. 2. Hierarchical cluster diagrams of all Recent species included in this study. A. Cluster based
on index of hypsodonty, percent high occlusal relief, percent sharp cusps and percent blunt cusps. B.
Cluster based on the same variables as (A) except index of hypsodonty. C. Cluster based on percent
sharp cusps and percent blunt cusps only. D. Cluster based on the same variables as (C) plus index of
hypsodonty. Symbols as in table 1. UPPER CASE ⫽ BROWSER, lower case ⫽ grazer, Mixed Case ⫽
Mixed Feeder.
percent rounded cusps, polarizes the full set based on percent sharp cusps and percent
of 64 Recent species into a pattern that blunt cusps only whereas D is based on the
groups grazers and browsers at the extremes, same variables as (C) plus index of hypso-
with mixed feeders in between (fig. 2). The donty. Figure 1 shows that the grazers group
figure shows hierarchical cluster diagrams of more clearly than browsers, which, although
all Recent species included in this study. A clumped at one end, also intersperse with
is a cluster based on index of hypsodonty, mixed feeders throughout the mixed feeder
percent high occlusal relief, percent sharp range. Mixed feeders also intersperse with
cusps and percent blunt cusps. B is a cluster browsers. There are two main clusters, cor-
based on the same variables as (A) except responding to attrition-dominated and abra-
for the index of hypsodonty. C is a cluster sion-dominated wear, respectively. The attri-AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 13
Allen Press • DTPro System
File # 01cc
2000 FORTELIUS AND SOLOUNIAS: UNGULATE MOLARS 13
tion-dominated cluster is divided into two that the cusps of the greater kudu are consid-
subclusters, one containing mostly browsers, erably less rounded is not picked up by our
the other mostly browse-dominated mixed crude analytical procedure.
feeders. The abrasion-dominated cluster also Close inspection shows that dispersed
shows two subclusters, one containing ‘‘ex- browsers are principally small species with
treme’’ grazers (bison bb, white rhino cs, topi significantly rounded cusps: the water chev-
dl, and the zebras eb, eg), the other contain- rotain HY; the duikers DR NA NI NG SL;
ing the rest of the grazers and the graze-dom- the hyraxes DD, HB, and two selective, long-
inated mixed feeders (the abbreviations after necked browsers; the gerenuk LW, and the
the names are the labels used in table 1 and dibatag EI. The water chevrotain and the dui-
the clusters). This second subcluster is also kers are well known to be highly frugivo-
divided into (1) all the remaining grazers, a rous, and are thus not typical browsers (Gau-
few graze-dominated mixed feeders, and one tier-Hion et al., 1980; Kingdon, 1982a;
browser (the greater kudu TT) and (2) mainly Lumpkin and Krantz, 1984; Feer, 1989; No-
graze-dominated mixed feeders and a set of wak, 1991). The rounding of their cusps is
small species (mainly duikers and hyraxes) probably due to the ‘‘tip crushing wear’’ typ-
that are abrasion-dominated for reasons not ically associated with frugivory (Janis,
related to grazing, as discussed below. Some- 1990). The hyraxes are opportunistic feeders
what unexpectedly, the effect of hypsodonty with varied diets including (in Heterohyrax
on this pattern is negligible: virtually the and Dendrohyrax) a high proportion of in-
same tree is obtained whether hypsodonty is sects (Kingdon, 1974). Why the grass dom-
included (fig. 2A) or excluded (fig. 2B). This inated mixed feeder Procavia capensis Pc
is universally observed in other sets analyzed shows such strong attrition-dominated wear
but is not shown further here. It is fortunate is unclear, but it seems reasonable to treat the
for the practical reason that the index of hyp- water chevrotain, the duikers and hyraxes as
sodonty is far more difficult to obtain for a a special case (‘‘mabra,’’ for ‘‘minute abrad-
species than are the mesowear variables. The ed brachydont’’) until more information be-
omission of occlusal relief (fig. 2C) has a comes available or a more adequate dietary
much more noticeable effect, most evident in classification is devised. There is no good
the disintegration of the grazing cluster seen reason to exclude the dibatag and the gere-
in the previous analyses. Adding hypsodonty nuk, although we hypothesize that their high-
recaptures some of the structure (fig. 2D), but ly selective feeding results in too little attri-
does not produce the clear cluster of grazers tion to mask even the small amount of abra-
seen in figures 2A, B. sion involved. As with the greater kudu, this
In figure 2A and B, two Indian deer (bar- may also be a calibration problem involving
asingha Cd, chital Ax), usually regarded as the degree of rounding of the cusps.
mixed feeders, cluster with the second rank Excluding the water chevrotain, the dui-
of grazers (hartebeest ab, wildebeest ct), and kers, and the hyraxes produces a more dis-
we suspect that they may be best interpreted tinct clustering (fig. 3), again with only mi-
as grazers, as indeed they are in our ‘‘radi- nor differences due to inclusion (fig. 3A) or
cal’’ classification in table 1. The mixed exclusion (fig. 3B) of hypsodonty. The mix-
feeders associated with the third rank of ing of grazers and grass-dominated mixed
grazers are also all strongly grass-oriented feeders reported above is still seen, and the
and in some cases (African buffalo Cs, greater kudu TT still clusters with these
mountain reedbuck Rf) may be equally well forms, but the remaining browsers are much
classified as grazers. The main anomaly is less dispersed. The mixed feeders, Indian rhi-
the greater kudu TT, a browser that persis- noceros Ru and springbuck Ma, fall among
tently clusters with these graze-oriented the browsers and the browser bongo BE falls
mixed feeders and grazers. The reason for among the attrition-dominated mixed feed-
this anomaly is essentially a simple calibra- ers. The springbuck is a selective mixed
tion problem: the greater kudu shows 100% feeder (Bigalke, 1978) that could well have
rounded cusps, like many typical grazers a mechanically browserlike diet and wear
such as the common waterbuck (ke). The fact pattern. The Indian rhinoceros is representedAMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 14
Allen Press • DTPro System
File # 01cc
14 AMERICAN MUSEUM NOVITATES NO. 3301
Fig. 3. Hierarchical cluster diagrams of all Recent species included in this study except the ‘‘ma-
bra’’—group. A. Cluster based on index of hypsodonty, percentage high occlusal relief, percentage sharp
cusps, and percentage blunt cusps. B. Cluster based on the same variables as (A) except index of
hypsodonty. Symbols as in table 1. UPPER CASE ⫽ BROWSER, lower case ⫽ grazer, Mixed Case ⫽
Mixed Feeder.
by a very small sample (5 specimens), and of species for which good dietary data are
in any case its placement in the less extreme available and where the interpretation seems
of the two browser clusters is no great anom- to be uncontroversial. We have selected 27
aly. The bongo is undoubtedly a browser but species to form a set of such ‘‘typical’’ spe-
is said to dig up and ingest roots (Nowak, cies. Their clustering pattern is essentially
1991), a habit that could well account for the free of anomalies (fig. 4); there are three
extra abrasion detected. main clusters, one for true grazers, one for
It may be of interest to explore the subset less extreme grazers and mixed feeders, and
one for browsers. The grazer-mixed feeder
cluster is cleanly divided into two subclus-
ters, one for grazers and one for mixed feed-
ers. The browser cluster is also divided into
two, with the slightly more abrasion-domi-
nated Sumatran rhinoceros DS, giraffe GC,
and mule deer OH forming their own sub-
cluster. Although such a set of ‘‘typical’’ spe-
cies may not say much about the dietary di-
versity of living species we feel that it forms
a good basis for comparison with fossil
forms, as shown below.
The ‘‘typical’’ set can also illustrate the
relative contribution of individual variables
to the resolution of the clusters. For a more
comprehensive and quantitative treatment we
refer to the discriminant analysis reported be-
Fig. 4. Hierarchical cluster diagram of a set low, especially table 3. Figure 5 we show the
of ‘‘typical’’ Recent species. Cluster based on per-
centage high occlusal relief, percentage sharp two most powerful variables and the weakest
cusps and percentage blunt cusps. Symbols as in variable obtained by discriminant analysis
table 1. UPPER CASE ⫽ BROWSER, lower case for the typical set: percent sharp cusps (fig.
⫽ grazer, Mixed Case ⫽ Mixed Feeder. 5A), hypsodonty (fig. 5B) and percent highAMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 15
Allen Press • DTPro System
File # 01cc
2000 FORTELIUS AND SOLOUNIAS: UNGULATE MOLARS 15
Fig. 5. Hierarchical cluster diagrams of a set of ‘‘typical’’ Recent species. A. Cluster based on
percentage sharp cusps only. B. Cluster based on index of hypsodonty only. C. Cluster based on per-
centage high occlusal relief only. Symbols as in table 1. UPPER CASE ⫽ BROWSER, lower case ⫽
grazer, Mixed Case ⫽ Mixed Feeder.
relief (fig. 5C). It is clear that both percent other mesowear variables. Percent high relief
sharp cusps and the index of hypsodonty seems to be particularly critical for the rec-
alone are capable of polarizing the species ognition of grazers and it alone does, in fact,
along a dietary axis, although neither works recognize all of the extreme grazers as such
as well as the full set of mesowear variables. (fig. 5C).
It is also clear that hypsodonty produces a
decidedly more mixed result in terms of diet DISCRIMINANT ANALYSIS
than does percent sharp cusps. The result
based on the highly skewed variable percent Discriminant analysis was performed to
high relief, is clearly much less satisfactory quantify the resolution of mesowear analysis
than the others, but as shown above, it does with respect to the three conventional dietary
significantly increase resolution within the classes of browser, grazer, and mixed feeder.
dietary classes when used together with the As already shown in the cluster analyses,AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 16
Allen Press • DTPro System
File # 01cc
16 AMERICAN MUSEUM NOVITATES NO. 3301
mesowear analysis resolves the structure and radical. Other combinations classified
within these classes in a biologically mean- less successfully, with a mean value of 61%
ingful way, which is completely missed by for conservative and 63% for radical.
an analysis based on the main classes only. For the minute abraded brachydonts-free
For example, cusp shape without occlusal re- set, the pattern was essentially the same but
lief gives higher scorings but a poorer clus- percentages correctly classified averaged 7–
tering pattern than does cusp shape with oc- 8% higher. For the typical set six out of the
clusal relief (figs. 2B and C, table 3). These ten combinations correctly classified 96% or
analyses should thus not be seen as a test of more of the species, percent high relief with
‘‘how well’’ mesowear analysis works, only either persharp or perround both giving
of how well it classifies diet at the most gen- 100% correct.
eral and conventional level in the different THREE-VARIABLE ANALYSES: As the num-
sets of species, and especially of how the ber of variables increase the percentage of
‘‘conservative’’ and ‘‘radical’’ dietary clas- species correctly classified mounts and the
sifications affect the pattern. difference between the conservative and rad-
SINGLE-VARIABLE ANALYSES: For the full ical classifications is reversed, so that where-
data set of 64 living species, the single var- as the radical classification gets the highest
iable that classified the species best was the percentages of species correctly classified for
index of hypsodonty (hypind), with an over- one and two variables, from three variables
all correct classification (jackknifed matrix onward, the conservative classification gets
reported throughout) of 65% for the conser- the highest scores. For three variables and
vative (cons) and 59% for the radical (radi) the full set of species, the combinations of
classifications (table 3). The second-best sin- hypsodonty index with any two cusp shape
gle variable was percent high relief (per- variables performs best (80% correct for con-
high), which correctly classified 55% (con- servative, 72% for radical), followed by
servative) and 56% (radical) overall. Percent combinations of hypsodonty index, percent
blunt cusps (perblunt) alone correctly clas- high relief, and any cusp shape variable
sified 58% of radical but only 48% of con- (76% for conservative, 70–72% for radical).
servative. The mean of all single-variable The mesowear variables without hypsodonty
analyses of the full set were 50% for con- give significantly lower values and the pat-
servative and 56% for radical. tern is reversed, with higher percentages cor-
For the set without the ‘‘minute abraded rectly classified for the radical classification.
brachydonts’’ (mabra) the pattern was similar The mean percentage correctly classified are
and the percentages correctly classified over- 72% for conservative and 68% for radical.
all were about 5% higher on average, with For the minute abraded brachydonts-free
hypsodonty index producing 63% for con- group, the pattern is the same but the per-
servative and 57% for radical. Percent sharp centages average 6% higher for conservative
cusps (persharp) performed distinctly better and 4% higher for radical. For the typical set,
for this set, about as well as hypsodonty in- all combinations except one correctly clas-
dex, with 61% for conservative and 69% for sify 96% or more of the species. The excep-
radical. For the typical set, where conserva- tion is hypsodonty index, percent high relief,
tive and radical coincide, percent sharp cusps and percent blunt, at 81% correct. This is a
(persharp) correctly classified 96% of the recurring pattern: percent high relief and per-
species, followed by percent rounded cusps cent blunt together seem to perform relative-
(perround) and hypsodonty index at 81% and ly poorly in most combinations of two or
a mean single-variable value of 78% (table three variables. Both variables are highly
3). skewed since most taxa have high relief and
TWO-VARIABLE ANALYSES: The combina- few blunt cusps.
tion of hypsodonty index and percent blunt FOUR-VARIABLE ANALYSES: All combina-
cusps (perblunt) performed best, at 80% cor- tions of hypsodonty with occlusal relief and
rect for conservative and 72% for radical. two of the three cusp shape variables give
The index of hypsodonty with persharp also the same result. For the full set of species,
classified well, at 76% for both conservative 76% (conservative) or 72% (radical) are cor-AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 17
Allen Press • DTPro System
File # 01cc
2000 FORTELIUS AND SOLOUNIAS: UNGULATE MOLARS 17
Fig. 6. Bivariate plots for the three dietary classes (conservative classification) of percentage high
occlusal relief (PERHIGH) against the index of hypsodonty (HYPIND). Data from table 1.
rectly classified. For the minute abraded (Kruskal-Wallis test, P ⬍ 0.001 for both the
brachydonts-free set these values are 80% conservative and the radical classifications).
(conservative) and 73% (radical). The typical The only species that show predominantly
set is classified at 96% correct. low relief are those of the zebras and the
white rhinoceros, which are all plagiolopho-
MESOWEAR AND DIET—A CLOSER LOOK dont grazers. However, plagiolophodont
This section gives a more detailed break- forms may also show high relief (for exam-
down of the general patterns found in cluster ple the hipparions included in this study),
and discriminant analyses of hypsodonty, oc- and feral horses that have lived on browse
clusal relief, and cusp shape in relation to also show high occlusal relief (unpublished
diet. data). Therefore, low relief appears to be a
HYPSODONTY: The index of hypsodonty is strongly associated with grazing (or at least
significantly different among conservative with highly abrasive food).
and radical dietary classes (Kruskal-Wallis CUSP SHAPE: Cusp shape appears to be
test, P ⬍ 0.001) except that the difference largely independent of hypsodonty, as all
between mixed feeders and grazers is only three cusp shapes occur in all crown height
significant for the conservative classification classes (fig. 7). Looking separately at the per-
(P ⫽ 0.003). centages of sharp, rounded, and blunt cusps
OCCLUSAL RELIEF: High occlusal relief is allows two further distinctions: no extant
the common state except in grazers, which grazer has more than 40% sharp cusps, and
cover the full range (fig. 6). No browser has no browser or mixed feeder has more than
less than 80% high relief, and only two 10% blunt cusps. Despite nearly complete
mixed feeders (the brachydont lesser kudu overlap of ranges, all three dietary groups in
and the hypsodont saiga) have lower values. all combinations are significantly different
The only hypsodont browser, the pronghorn, from each other in cusp shape (persharp and
has 100% high relief, but very high values perblunt; Kruskal-Wallis test; P ⱕ 0.001
are also found for most hypsodont mixed for both conservative and radical), except
feeders. There is no evidence that occlusal that mixed feeders are not significantly dif-
relief would systematically change with hyp- ferent from browsers. In most cases cusp
sodonty in browsers or mixed feeders, but in shape does not seem to change with hypso-
grazers the most hypsodont forms have low donty, but the only hypsodont browser (the
relief. These are the ‘‘extreme grazers’’ iden- pronghorn) has a very high percentage of
tified by the cluster analyses, primarily the sharp cusps, indicating a high attrition level
species of Equus and certain Alcelaphini. and arguing against any role for abrasive
Grazers as a group have significantly lower food or dust. In contrast, percent rounded
relief than either browsers or mixed feeders cusps seems to decrease with increasing hyp-AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 18
Allen Press • DTPro System
File # 01cc
18 AMERICAN MUSEUM NOVITATES NO. 3301
Fig. 7. Bivariate plots for the three dietary classes (conservative classification) of cusp shape. PER-
SHARP; PERROUND; PERBLUNT) against the index of hypsodonty (HYPIND). Data from table 1.
sodonty in grazers, suggesting that increased although the relief is clearly higher in artio-
abrasion may not be the main reason for the dactyls. Dots next to each tooth are 5 mm
increased wear in these forms. The plots apart; all views are buccal of adult upper mo-
shown in figures 6 and 7 change only in de- lar teeth (mostly M2). Figures 11–15 show
tails if the radical dietary classification is teeth of high relief and rounded apices.
used. Hierarchical cluster diagrams for fossil Again the perissodactyl overall has the lower
and Recent species are shown in figure 8. relief (fig. 15I). We find that the condition of
FIGURES OF SELECTED TEETH: A sample of sharp and rounded remain the same despite
sharp and high relief teeth is shown in figures the degree of relief. For example, figure 14D
9 and 10. Both perissodactyl and artiodaclyl and 14I have similarly rounded apices but the
teeth show that the apices are similarly sharp degree of relief differs. Figure 16 shows highAMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 19
Allen Press • DTPro System
File # 01cc
2000 FORTELIUS AND SOLOUNIAS: UNGULATE MOLARS 19
Fig. 8. Hierarchical cluster diagrams, fossil and Recent species. A. Fossil species and all Recent
species included in this study except the ‘‘mabra’’-group. Cluster based on percentage high occlusal
relief, percentage sharp cusps and percentage blunt cusps. B. Fossil and ‘‘typical’’ Recent species.
Cluster based on the same variables as (A). Symbols as in table 1. UPPER CASE ⫽ BROWSER, lower
case ⫽ grazer, Mixed case (capital first) ⫽ Mixed feeder, mixed case (lower-case first) ⫽ fossil Species.
and low cusps which are all rounded whereas creasing percentage of sharp cusps down to
figure 17 shows high rounded (17H, K) and Tragelaphus strepsiceros, which has 100%
low blunt cusps (17F, I). Figures 18 and 19 rounded cusps. Note that the hypsodont
show low and blunt cusps of perissodactyls. browser Antilocapra has a high proportion of
In some of the Equus individuals cusp are sharp cusps, as expected from its diet. Hyae-
more than blunt they are convex (fig. 18F moschus and most duikers (Cephalophus)
and 19D). So far we have classified such have a small percentage of blunt cusps, prob-
convex cusps as blunt. Figures 20 and 21 ably because of ‘‘tip-crushing’’ wear due to
show cusps of mixed modality. For example, frugivory as noted above. Litocranius also
figure 20A shows sharp and rounded cusps shows strong rounding, possibly because of
on the same tooth. Figure 20I shows a blunt the very low overall rate of wear (attrition
cusp and a rounded cusp on the same tooth. not enough to mask even very low abrasion).
Figure 21A shows strongly blunt cusps and Dendrohyrax dorsalis is the only browser in
21D and F mixed rounded and sharp cusps. our data set that features some blunt cusps,
We find that the mixed patterns within the but as discussed above hyraxes are too om-
same dentition are not common. nivorous to be qualified as typical browsers
COMPARISON OF CUSP-SHAPE HISTOGRAMS: and their dental wear is not well understood.
The cusp mesowear histograms for a selec- Figure 23 shows the percentages of cusp
tion of extant species from table 1 are given tip shapes for selected hypsodont mixed
in figures 22–26. Histograms are based on feeders. Ovibos, Saiga, and Capra are attri-
the raw data and are the same with the sum- tion-dominated and all species show at least
mary values given in table 1. 25% sharp cusps. Blunt cusps are only seen
Figure 22 shows the percentages of cusp in Lama and Capra, less than 10% in both
tip shapes for selected browsers; the data cases.
show a large array of differences between Figure 24 shows the percentages of cusp
species. Alces alces is the only species with tip shapes for selected ‘‘traditional’’ mixed
100% high and sharp cusps. The remaining feeders, arranged from the attrition-dominat-
browsers can be arranged according to a de- ed Antidorcas to the abrasion-dominated Ae-AMNH NOVITATES
Friday Jul 07 2000 10:13 AM 2000 novi 00137 Mp 20
Allen Press • DTPro System
File # 01cc
20 AMERICAN MUSEUM NOVITATES NO. 3301
Fig. 10. Relief low and high, sharp cusps. A:
Fig. 9. Relief high, sharp cusps. A: Alces al- Diceros bicornis AMNH 139694, left M2; B:
ces AMNH 6408, left M2-M3; B: 207705, left 139692, right M1; C: 113776, left M2; D:
M1; C: 19799, right M2; D: 98162, right M2-M3; 167693, right M2; E: TE 7990, left M2; F:
E: 173563, left M1-M2; F: Capra ibex AMNH AMNH 13778, left M2; G: 13776, left M2; H:
82264, left M3; G: 11571, left M2; H: 69428, TE 7987, right M2; I: Cervus canadensis AMNH
right M3; I: 117575, left M2; J: 69428, right M1; 40005, left M1; J: Equus burchelli AMNH
K: 57318, right M3; L: Litocranius walleri 119669, left M2; K: 82312, left M2; L: 204106,
AMNH 179215 left M3; M: 161173 right M3; N: right M2.
179217 right M1.
tip shapes for grazers. The data show a di-
pyceros. Most species have about 50% sharp versity that promises potential for future pa-
and 50% rounded cusps, while blunt cusps laeodiet analysis. The Reduncini (Redunca
are absent except for a few in Gazella thom- and Kobus) have a very high percentage of
soni, the more grass-oriented of the two ga- rounded cusps, a signal apparently related to
zelles. marginal or ‘‘fresh grass’’ grazing. All other
Figure 25 shows the percentages of cusp grazers have a significant proportion of blunt
tip shapes for some mixed feeders of the In- cusps. The Alcelaphini (Connochaetes, Al-
dian monsoon forest. All are highly abrasion- celaphus, and Damaliscus) are dominated by
dominated, like African fresh grass grazers rounded cusps but also show significant pro-
(and the browsing greater kudu). The cervids portions of sharp and blunt cusps. The zebras
(Axis axis and Cervus duvauceli) particularly (Equus) stand out with almost subequal
show a grazerlike profile with over 20% amounts of cusps of each type, while the
blunt cusps, and both are indeed treated as white rhinoceros (Ceratotherium) has over
grazers in our ‘‘radical’’ dietary classification 30% blunt cusps and no sharp ones at all.
(table 1). Bison is unique in our data set in having
Figure 26 shows the percentages of cusp mostly blunt cusps (over 70%).You can also read
