AMQ29(2) 6 Altamura 77-90.pub Available online http://amq.aiqua.it ISSN (print): 2279-7327, ISSN (online): 2279-7335 Alpine and Mediterranean Quaternary, 29 (1), 2016, 77 - 89 DIETARY BEHAVIOUR OF THE PLEISTOCENE CRETAN DWARF DEER: PRELIMINARY CLUES FROM MESOWEAR ANALYSIS OF CANDIACERVUS EX GR. CANDIACERVUS ROPALOPHORUS FROM BATE CAVE Sara Altamura Dipartimento di Scienze della Terra, Università Sapienza, Rome, Italy Corresponding author: S. Altamura ABSTRACT: During the Middle and Late Pleistocene Crete was inhabited by endemic dwarf deer. The smallest, Candiacervus ex gr. Candiacervus ropalophorus (sensu Palombo et al., 2008) have been believed to have inhabited rocky environments characterised by typi- cal Mediterranean vegetation. This research aims to investigate dietary behaviour in the sample of the smallest deer found in the Bate Cave late Middle - early Late Pleistocene deposits (152,000±20% to 105,000±20% years). The molar teeth of Candiacervus ex gr. C. ropalophorus from Bate Cave mainly show well-defined and high cusps, sometimes also percep- tible in teeth in an advanced wear stage. Therefore, it seems rational to suppose that deer had an eating behaviour approaching browsers. In order to substantiate this hypothesis, it was applied the mesowear method proposed by Fortelius & Solounias (2000) to infer the dietary behaviour of herbivores. The method was originally based on facet development on the occlusal surface of the second upper molar tooth, then extended to other molars, although results obtained show some inconsistencies. The number of M2 specimens of Candiacervus ex gr. C. ropalophorus from Bate Cave is too low to be statistically valid. Therefore, the analysis was extended to the last two lower and upper molars, divided in group according to their wear stage, with the double aim to check the consistence of results obtained by analysing upper and lower molars and verify to which extent wear may influence the cusp shape. The results obtained by extension of the method are en- couraging, in fact all the teeth reveal the same trend suggesting a prevalent mixed-feeder habit for Candiacervus ex gr. C. ropalophorus from Bate Cave although in some individuals a more marked browsing attitude is observed. A possible diet included leaves, shrubs and other tender and nutritious vegetation. Key words: Candiacervus ropalophorus, Crete, Mesowear, Dietary behaviour, Pleistocene. 1. INTRODUCTION The study of feeding behaviour in herbivorous mammals (ungulates) is of particular interest to infer habitat preference and behaviour of fossil herbivores and provides some clues about the en- vironment they inhabited. The food preference, indeed, reflects the vegetation cover and availabil- ity and partition of resources. According to their feeding behaviour, herbivores are generally classi- fied into three main ecological groups: browser, grazer and mixed feeder. Cervids are a highly heterogeneous group, most of species are ecologically flexible, some inhabit different environments, and some are opportunistic / mixed feeder, changing seasonality or occasionally their diet depending on resource availability and competition with other herbivores (see inter alios Geist, 1999; Palombo, 2005; Yamada, 2012; De Marinis & Toso, 2015). The varied dietary habits of herbivores may be in- ferred by analysing the morphology of the tooth itself. In particular, the method of mesowear analysis, as origi- nally proposed by Fortelius & Solounias (2000), pro- vides information about the prevalent feeding behaviour of a species basing on the observation of occlusal facet of the second upper molar (M2). The development of this facet is related to the relative ratio of attrition (tooth-to- tooth contact) and abrasion (food-to-tooth contact) that occurs during mastication activities. According to For- telius & Solounias (2000) the method provides stable trophic classification when at least 20 teeth are investi- gated. However, in the study of fossil species, some- times the sample is not large enough to provide firm results. Afterwards, the method has been extended to other teeth including the last two upper molars (two- tooth model - M2-M3) (Franz-Odendaal & Kaiser; 2003; Valli & Palombo, 2008), all molars (M1÷M3, Kaiser & Croitor, 2004; M1÷M3 and M1÷M3 Louys et al., 2011) or all the chewing teeth (four-tooth model - P4÷M3; P4÷M3) (Kaiser & Solounias, 2003; Kaiser, 2003; Kaiser & For- telius, 2003). The method has been applied in several studies dealing with the dietary strategy in fossil and extant her- bivores, such as Equidae (Kaiser & Solounias, 2003; Kaiser, 2003; Kaiser & Fortelius, 2003; Mihlbachler et al., 2011; Loffredo & DeSantis, 2014), Cervidae (Kaiser & Croitor, 2004; Palombo, 2005; Valli & Palombo 2008; Yamada, 2012) and others (Franz-Odendaal & Kaiser, 2003; Kaiser & Kahlke, 2005; Clauss et al., 2007; Rivals et al., 2011; Louys et al., 2011; Solounias et al., 2014). Furthermore, in order to obtain data comparable with other possible variables (e.g. Hypsodonty, %grass, habi- tat) (Rivals & Semprebon, 2006; Kaiser et al., 2009; Valli & Palombo, 2008; Loffredo & DeSantis, 2014) the Fig. 1 - Skulls (a: female specimen 6; b: male specimen 3), jaws (c: sn1-sx; d: 8-dx) and mounted skeletons (e) of Candiacervus ex gr. Candiacervus ropalophorus from Bate Cave (Paleontological Museum of Sapienza University, Rome). 78 Altamura S. mesowear qualitative variables (OR and cusps shape) were combined to convert them into mesowear score: high relief and sharp cusps = ‘‘0’’, high relief and rounded cusp = ‘‘1’’, low relief and sharp cusp = ‘‘2’’, low relief and round cusp = ‘‘3’’, low relief and blunt cusp = ‘‘4’’. Indeed, a score of 0 represents the most attrition- dominated mesowear signals, while a score of 4 would represent the most abrasion-dominated signals. (Rivals & Semprebon, 2006; Kaiser et al., 2009; Mihlbachler et al., 2011). The research aims to infer the dietary behaviour of the Cretan endemic deer, Candiacervus ex gr. Candi- acervus ropalophorus (sensu Palombo et al., 2008) (Fig. 1), found in the Bate Cave (Rethymnon) and dated be- tween 150,000±20% and 105,000±20% years (AAR method -Reese et al., 1996). During the Middle and Late Pleistocene, the Crete island was inhabited by at least five endemic deer which originated from a radiative evolutionary process that led to a niche partition among deer species differing in size, habitat and feeding behaviour (De Vos, 1996, 2000). The Candiacervus ropalophorus group includes the smallest among the Cretan endemic deer (Candiacervus ropalophorus and Candiacervus sp. II in De Vos, 1979, 1984, 1996), C. ex gr C. ropalophorus is believed to have inhabited rocky environments with a Mediterranean vegetation and to have had a mixed diet mainly consisting of bushes, shrubs, leaves and grass, whose availability and quality were subject to seasonal variations (Caloi & Palombo, 1996; van der Geer et al., 2006; van der Geer et al., 2010; van der Geer et al., 2014; Mujica, 2014; Palombo & Zedda, 2015). To vali- date such a hypothesis, I applied the mesowear method to the last two upper and lower molars also with the aim of investigating whether results obtained by analysing these chewing teeth are consistent with each other and, therefore, may provide hints to infer dietary behaviour when the M2 sample is not statistically valid, and the wear variation is particularly high. 2. MATERIALS AND METHODS The material object of this study is kept at the Pale- ontological Museum of the La Sapienza University of Rome and consists of 102 molars: 25 M2, 32 M3, 24 M 2 and 21 M3. I chose to analyse the upper and lower pe- nultimate and ultimate molars because results provided by the two-tooth model (M2-M3) are similar to those ob- tained by analysing only M2 (Franz-Odendaal & Kaiser, 2003), while the analysis of upper and lower M1 and P4 has been demonstrated to be less reliable (as least as ruminants is concerned, conversely to horses) (Kaiser & Solounias, 2003; Kaiser, 2003; Kaiser & Fortelius, 2003) because of the extent of the wear period (see Louys et al., 2011). Only permanent teeth showing different wear stages have been considered. The sample does not include teeth in very advanced stage of wear. The shape of cusps of molars of deer from Bate Cave ranges from high and sharp in unworn teeth to low and rounded/ sharp in the most worn one. Therefore, prior to perform- ing the mesowear analysis, C. ex gr. C. ropalophorus molars were grouped according to their degree of wear. The wear classes were evaluated according the IDAS system (individual dental age stages) based on tooth eruption and tooth wear (Anders et al., 2011) and the progressive decrease of crown height have been used as a guideline to define three wear stages: slightly worn teeth (1° wear stage, W1) (IDAS 2, permanent dentition is fully erupted), average worn teeth (2° wear stage, W2) (IDAS 3, the second molar starts to lose the inner profile), strongly worn teeth (3° wear stage, W3) (IDAS 4, the second molar loses the inner profile) (Anders et al., 2011). It is well known that in herbivore the height of the tooth crown provides clues as regard to dietary apti- tudes, because reflects the resistance of tooth to the abrasive power of food responsible for tooth wear during the masticatory process (i.e. herbivores feeding on fi- brous material have a hypsodont dentition) (see inter alios Fortelius, 1985; Janis, 1988; Janis & Fortelius, 1988; Damuth & Janis, 2011; Kaiser et al., 2013). The quantitative measure of hypsodonty, (Hypsodonty Index, HI) can be calculated in different ways: on M2 by dividing the height of the crown by the occlusal total length of the tooth (Valli & Palombo, 2008), and on M3 by dividing the height of the crown by width of the tooth (Janis, 1988; Fortelius & Solounias, 2000; Kaiser et al., 2013). Appling both methods the value of the HI measured on unworn teeth is < 1.5 in brachydont, ranges from 1.5 to 3.0 in mesodont, and from 3.0 to 4.5 in hypsodont, and is > 4.5 in highly hypsodont animal. The HI of C. ex gr. C. ropalophorus has been evalu- ated measuring M2 and M3 at different wear stage. As regards to the parameters related to mesowear method (Fortelius & Solounias, 2000), the cusp height (Occlusal Relief-OR) was defined as the difference in height between cusp tips and inter cusp valley as seen in labial projection (M2, M3) or in lingual projection (M2, M3). The OR values is “high” when the result is greater than 0.1, while “low” when it is less than 0.1. The shape of the cusps was qualitatively evaluated according to three typologies: sharp, rounded, and blunt (Fig. 2). All mesowear results was converted into mesowear score (Kaiser et al., 2009) and plotted versus HI in order to verify whether this rating may be regarded as a valid habitat proxy (Kaiser et al., 2013). Data obtained for C. ex gr. C. ropalophorus have been compared with those of 63 extant species (data from Fortelius & Solounias, 2000; Mendoza & Palmqvist, 2008; Kaiser et al., 2009; Kaiser et al., 2013) (Tab. 1). To infer the grazer versus browser ability of C. ex gr. C. ropalophorus from Bate cave, a cluster analysis has been performed comparing C. ex gr. C. ropalophorus with 52 extant artiodactyl species (cervids, bovids, giraf- fids and camelids, data from Fortelius & Solounias, 2000; Franz-Odendaal & Kaiser, 2003; Louys et al., 2011) (Tab. 2). The species were assigned to dietary categories on the basis of personal observation and comparing results obtained by different authors (Fortelius & Solounias, 2000; Gagnon & Chew, 2000; Louys et al., 2011). Four extinct species having different trophic and ecological requirement have been added to the cluster analysis (Tab. 2) selected because they show a Dietary behaviour of Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete) 79 mesowear patter similar to that of C. ex gr. C. ropalo- phorus (Duboisia santeng, Metacervoceros rhenanus and Croizetoceros ramosus) and because inhabited on a Mediterranean island environments similar to those of the Cretan deer (Praemegaceros cazioti). Two Pleisto- cene insular species, Praemegaceros cazioti from Dragonara Cave (Sardinia) and Duboisia santeng from Java, and two continental Villafranchian cervids Meta- cervoceros rhenanus and Croizetoceros ramosus from Saint-Vallier (Drôme, France). The dwarf megacerine Preamegaceros cazioti from Dragonara Cave (Sardinia) was a mixed feeder inhabit- ing territories characterized by hard and bumpy sub- strate (Caloi & Palombo, 1991; Mujica, 2014) and a vegetation cover comparable to the Mediterranean scrub (Palombo, 2005). The dwarf bovid Duboisia santeng from Java, was a forest dweller, mostly feeding on leaves and only occasionally on harder vegetation (Rozzi et al., 2013). The Metacervoceros rhenanus and Croizetoceros ramosus Villafranchian deer from Saint- Tab. 2 - The extant and fossils species were compared in this study by cluster analysis with Candiacervus ex gr. C. ropalo- phorus from Bate Cave. Tab. 1 - Date used for the construction of bivariate plot: dietary classification (DC); mesowear score (MS); hypsodonty index (HI); habitat categories (Hab), 1=open, 2=intermediate, 3=closed. (data from Fortelius & Solounias (2000), Kaiser et al., 2009, Kaiser et al. (2013) and references therein). <<------------------------------------------ 80 Altamura S. Vallier (Drôme, France) were seasonal mixed feeder with a mainly browsing feeding behaviour (Valli & Palombo, 2008). Hierarchical cluster analysis was performed by means of Past 3.13 software (Hammer et al., 2001) with Euclidean distances and complete linkage, using as variables the percentage of OR and shape cusp result- ing from mesowear analysis. Three clusters analyses have been performed on three samples. In the first, cluster-a, M2 and the two- tooth model approach were used irrespectively of the wear stage of C. ex gr. C. ropalophorus teeth. In the second, cluster-b, the last upper and lower molars are individually examined taking into account their wear stage. In the third, cluster-c, the two-tooth model was used, taking into account wear stage of the teeth. Fig. 2 - Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete) (Bate 6), fragment of the left maxillary with P2 - M3 (on the left). The mesowear variables as in Fortelius and Solounias (2000) (on the right). Fig. 3 - Upper and lower molars of Candiacervus ex gr. Candiacervus ropalophorus from Bate Cave (Crete) grouped according to their wear stages: Wear stage 1: a) third upper molar (labial and occlusal views); b) second lower molar (lingual and occlusal views); Wear stage 2: c) third upper molar (labial and occlusal views); d) second and third lower molar (lingual and occlusal views); Wear stage 3: e) second and third upper molar (labial and occlusal views); f) second and third lower molar (lingual and occlusal views). Dietary behaviour of Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete) 81 second lower molars (Fig. 4). In teeth at first wear stage, cusps are in upper mo- lars only sharp, while in the lower molars rounded cusps reach the 30%. As the degree of wear augments the shape of cusps changes from sharp towards less- pointed, the percentage of rounded cusps increases and some blunt cusps may be occasionally present. It is worth noting, however, that cusps are still evident even in the worn teeth. Moreover, with the increasing of the wear degree, the percentage of the high cusps slightly decreases, but they are still the most abundant (Fig. 5, Tab. 5). The HI and MS values of the Cretan small deer are very low. As regards to the clues about dietary behaviour (Fig. 6) and habitat (Fig. 7) obtained by plotting HI versus MS, the C. ex gr. C. ropalophorus sample fall in the eco- space of browsers and species prevalently inhabiting close environments. The cluster analysis provides a quite good resolution in splitting extant artiodactyl species according to their dietary adaptation (Figg. 8a, b, c). Three main groups are generally detectable, the first includes a high per- centage of browsers, in the second includes mixed feed- ers and few browsers and grazers, in the third grazers dominate. In the cluster-a (Fig. 8a), C. ex gr C. ropalophorus (cRB)’s M2 falls in the mixed feeder spectrum (B.2) to- gether with Gazella granti (Gg), which is a mixed feeder 3. RESULTS 3.1. Wear Stage In the first wear stage teeth show thin streaks of dentin. The cusps, paracone and metacone, are well defined and the median groove shows a marked "V" shape (Fig. 3.a, 3.b). At the second wear stage, in the masticatory table the dentine area is wider and shows rhomboidal geome- try. The infundibula are evident though reduced with respect to the previous stage, and still connected in the upper molars. The median groove is evident, but the outline is "U-shape". The cusps are still well defined (Fig. 3.c, 3.d). In molars with an advanced wear degree, infundibula are greatly reduced, the dentine is exposed over the whole occlusal surface and the enamel band is thinner than that of previous wear stages. The bottom of the median groove can reach the neck, or even the roots. The cusps are still identifiable, the profile is an- gled or undulate, though height and shape of cusps show a quite significant variation (Fig. 3.e, 3.f). 3.2. Hypsodonty The average HI of C. ex gr. C. ropalophorus M2 and M3 shows values roughly consistent with each other, though HI values of M3 are generally higher than those obtained for M2. The lower third molar is, indeed, the last tooth to erupt and in extant deer the eruption is completed about 6-12 months (roe deer and red deer respectively) later than that of M2 (De Marinis & Toso, 2015). The average HI value in teeth at the first wear stage is < 1.5 confirming that C. ex gr. C. ropalophorus from Bate Cave has a brachyodont dentition (Tab. 3). As expected HI decreases as the wear augments. As ex- pected, the HI reduction is proportional to the degree of wear. Furthermore, the differences in HI between teeth at the first and the second stage of wear (HI) is slightly greater in M3 than in M 2, while the differences are al- most the same in M3 and M 2 at the second and third stages of wear. 3.3. Mesowear patterns and dietary behaviour Results obtained by the mesowear analysis on M2, M3, M2 and M3 of C. ex gr. C. ropalophorus from Bate Cave indicate that cusps have generally high (86%) and sharp/rounded (49%/48%) profiles (Tab. 4). The sharp shape prevails in the second upper mo- lar, while, rounded shapes are more frequent in the Tab. 3 - Hypsodonty Index in the last lower (M3) and second upper molars (M2) of Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete), grouped according to their wear stage. Tab. 4 - Summary of the percentage of mesowear variables in the molars of Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete), overall and grouped according to their wear stage. Tab. 5 - Percentage of mesowear variables in second and third upper and lower molars of Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete), grouped according to their wear stage. 82 Altamura S. but having a predominant browser behaviour (B.2.2.1.2). When the average value obtained from M2 and M3 analysis is considered (two-upper-tooth model), Cretan deer still falls in mixed feeder field (B.2.2.1.2), close to M2 average value, together with Ovibos mo- schatus (Om) which is an opportunistic specie. When the average value obtained from M2 and M3 analysis is considered (two-lower-tooth model), C. ex gr C. ropalo- phorus falls between the mixed feeders (B.2.2.1.1) but close to Antidorcas marsupialis (Ma), which is an inter- mediate seasonal browser. In the cluster-b (Fig. 8b), where the wear stages of C. ex gr. C. ropalophorus (cRB) have been taken into account, two main groups are identifiable. A small group (A) including browsers, and a large one (B) gathering both mixed feeders (B.1) and grazers (B.2), which, how- ever form two separate sister groups. The C. ex gr. C. ropalophorus M2, M3 (A.2.2) and M3 (A.1.2), at the first wear stage (W1) fall among browsers (A). In particular, M2 and M3 are linked with Alces alces (AA), and M3 with Odocoileus hemionus (OH), both typical browsers. Con- versely, M2 teeth fall among the mixed feeders (B.1.2.1.1) possibly because, although the cusps with a sharp shape predominate, the percentage of rounded cusps is not negligible (37.5%). The percentage of rounded cusps considerably increases in M2 at the third stage of wear, which fall in the grazer range (B.2.1.1.2) although the percentage of high cusps is unusually high (80%) for grazers. Most of the other C. ex gr. C. ropalo- phorus molars, showing various degrees of wear, fall in the large group of mixed feeders together with species (e.g. Gazella granti (Gg), G. thomsonii (Gt), G. dorcas (Gd), Ourebia ourebi (Oo), Tragelaphus strepsiceros (Tt), T. imberbis (Ti), but also Ammodorcas clarkei (EI)) that mainly eat leaves, fruit, or fresh grass, according to the season and the availability of vegetal resources. In the cluster-c (Fig. 8c), where only the two-tooth model with lower and upper molars of C. ex gr. C. ropa- lophorus (cRB) from the first to third stage of wear have been included, the distribution of extant species belong- ing to the three main trophic groups is similar to that of cluster-b. As regards to Cretan deer, molars at the first stage of wear fall in browser group (A), with Alces alces (AA) and Odocoileus hemionus (OH), while those at the second and third stage mainly fall among the mixed feeders tree. The sister group of the upper molars at the second and third wear stage and the lower molars at the third wear stage (B.1.1) includes Gazella soemmerringi (Gs), which feeds on bush leaves, grasses and herbs and G. dorcas (Gd), which eats leaves, pods, fruits and bulbs. The lower molars at the second wear stage has as sister group (B.1.2.2.1) Ovis canadensis (Oc), which consumes leaves, grasses, sedges and forbs, but it usu- ally takes some browse during winter, and Gazella granti (Gg), its diet consist mainly of leaves and stems. Fig. 6 - Bivariate plot of Mesowear Score (MS) and Hypsodonty Index (HI) of extant artiodactyls and Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete). Data of extant species from Fortelius & Solounias (2000), Kaiser et al. (2009) and Kaiser et al. (2013) (Tab.3). Fig. 7 - Bivariate plot of Mesowear Score (MS) and Hypsodonty Index (HI) with habitat as cofactor of extant artiodactyls and Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete). Data of extant species from Fortelius & Solounias (2000), Men- doza & Palmqvist (2008), Kaiser et al. (2009) and Kaiser et al. (2013) (Tab.3). Dietary behaviour of Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete) 83 Fig. 4 - Comparison among the percentage of mesowear vari- ables in second (M2u) and third (M3u) upper and lower (M2l- M3l) molars of Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete). Fig. 5 - Comparison among the percentage of mesowear vari- ables in upper and lower molars of Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete) grouped according to their wear stage. F ig . 8 .a - C lu st e r- a : H ie ra rc h ic a l cl u st e r d ia g ra m c o m p a ri n g t h e o n e -t o o th m o d e l u si n g u p p e r se co n d m o la rs o n ly a n d t h e u p p e r a n d l o w e r tw o -t o o th m o d e l u si n g s e co n d a n d t h ir d m o la rs o f C a n d ia ce rv u s e x g r. C . ro p a lo p h o ru s fr o m B a te C a ve , b a se d o n a s e t e xt a n t a n d f o ss il sp e ci e s (d a ta f ro m F o rt e liu s & S o lo u n ia s, 2 0 0 0 ; F ra n z- O d e n d a a l & K a is e r, 2 0 0 3 ; L o u ys e t a l., 2 0 1 1 ; P a lo m b o , 2 0 0 5 ; V a lli & P a lo m b o , 2 0 0 8 ; R o zz i e t a l., 2 0 1 3 ). T h e m e so w e a r va ri a b le s: p e rc e n t h ig h o cc lu sa l r e lie f, p e rc e n t sh a p e c u sp s (s h a rp , ro u n d e d a n d b lu n t) . B ro w se r= U P P E R C A S E a n d g re e n ; G ra ze r= lo w e r ca se a n d b ro w n ; M ix e d f e e d e r= M ix e d c a se a n d y e llo w ; F o ss il sp e ci e s= in ve rs e m ix e d C a se a n d p in k. c R B ( b lu e )= C . e x g r C . ro p a lo p h o ru s fr o m B a te C a ve . “D is ta n ce ” = E u cl id e a n d is ta n ce . 84 Altamura S. F ig . 8 .b - C lu st e r- b : H ie ra rc h ic a l cl u st e r d ia g ra m c o m p a ri n g t h e o n e -t o o th m o d e l u si n g u p p e r a n d l o w e r se co n d m o la rs a n d u p p e r a n d l o w e r th ir d m o la rs g ro u p e d a cc o rd in g t o t h e ir w e a r st a g e o f C a n d ia ce rv u s e x g r. C . ro p a lo p h o ru s fr o m B a te C a ve , b a se d o n a s e t e xt a n t a n d f o ss il sp e ci e s (d a ta f ro m F o rt e liu s & S o lo u n ia s, 2 0 0 0 ; F ra n z- O d e n d a a l & K a is e r, 2 0 0 3 ; L o u ys e t a l., 2 0 1 1 ; P a lo m b o , 2 0 0 5 ; V a lli & P a lo m b o , 2 0 0 8 ; R o zz i e t a l., 2 0 1 3 ). T h e m e so w e a r va ri a b le s: p e rc e n t h ig h o cc lu sa l re lie f, p e rc e n t sh a p e c u sp s (s h a rp , ro u n d e d a n d b lu n t) . B ro w se r= U P P E R C A S E a n d g re e n ; G ra ze r= lo w e r ca se a n d b ro w n ; M ix e d f e e d e r= M ix e d c a se a n d y e llo w ; F o ss il sp e ci e s= in ve rs e m ix e d C a se a n d p in k. c R B ( b lu e )= C . e x g r C . ro p a lo - p h o ru s fr o m B a te C a ve . “D is ta n ce ” = E u cl id e a n d is ta n ce . Dietary behaviour of Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete) 85 F ig . 8 .c - C lu st e r- c: H ie ra rc h ic a l cl u st e r d ia g ra m c o m p a ri n g t h e u p p e r a n d l o w e r tw o -t o o th m o d e l u si n g s e co n d a n d t h ir d m o la rs a cc o rd in g t o t h e ir w e a r st a g e o f C a n d ia ce rv u s e x g r. C . ro p a lo p h o ru s fr o m B a te C a ve , b a se d o n a s e t e xt a n t a n d f o ss il sp e ci e s (d a ta f ro m F o rt e liu s & S o lo u n ia s, 2 0 0 0 ; F ra n z- O d e n d a a l & K a is e r, 2 0 0 3 ; L o u ys e t a l., 2 0 1 1 ; P a lo m b o , 2 0 0 5 ; V a lli & P a lo m b o , 2 0 0 8 ; R o zz i e t a l., 2 0 1 3 ). T h e m e so w e a r va ri a b le s: p e rc e n t h ig h o cc lu sa l re lie f, p e rc e n t sh a p e c u sp s (s h a rp , ro u n d e d a n d b lu n t) . B ro w se r= U P P E R C A S E a n d g re e n ; G ra ze r= lo w e r ca se a n d b ro w n ; M ix e d f e e d e r= M ix e d c a se a n d y e llo w ; F o ss il sp e ci e s= in ve rs e m ix e d C a se a n d p in k. c R B ( b lu e )= C . e x g r C . ro p a lo p h o ru s fr o m B a te C a ve . “D is ta n ce ” = E u cl id e a n d is ta n ce . 86 Altamura S. 4. DISCUSSION AND CONCLUSION Results obtained by the mesowear analysis of mo- lars of C. ex gr. C. ropalophorus from Bate Cave sug- gest a chewing activity mainly characterised by a friction action associated with a milder abrasive component. The latter becomes more relevant as the degree of wear increases, although, in the cusps are still well delineated (Fig. 3.e, 3.f). Therefore, it is rational to suppose that the chewing process of C. ex gr. C. ropalophorus may have taken place in two phases. In the first phase a sharing component prevails, while in the second the sharing is progressively lost and the grinding prevails (Janis, 1979; Valli et al., 2012). The abrasive component equals or prevails over the frictional one if the wear is advanced. However, even at the third wear stage sharp cusps are still present (25%), suggesting that C. ex gr C. ropalo- phorus likely had an intermediate-browser feeding be- haviour. The variation of HI at different wear stage is not particularly high being the value of HI at the last stage 58% less than at the first. It is worth noting that the ty- pology of food may greatly affect brachyodont teeth, resulting in premature loss of its functionality (Anders et al., 2011). In deer, the extent of the wear stages is not characteristic of a species, conversely intraspecific variation has been observed according to the environ- mental characteristic and the resources availability. Accordingly, the variation in HI shown by C. ex gr. C. ropalophorus molars may be regarded as coherent with a diet consisting in no-abrasive food. This conclusion is supported by cluster analysis. The results obtained indicate that the small deer from Bate Cave had an intermediate-browser feeding strategy (mixed-feeder versus browser) and the diet possibly included more leaves and buds than hard grasses. The inferred feeding behaviour is supported by the cranial and mandibular features related to masticatory proc- esses (see Caloi & Palombo, 1996 for a discussion). In addition, it is interesting to note how mesowear pattern slightly differs in upper and lower molars having similar wear degree (Fig.8b). The cusps of second lower molar, indeed, are generally less sharp then those of last molars (Tab. 5), as indicate by the position of M2 (W1-W2-W3), which mainly fall among mixed-feeder, at the first and second stage of wear, and among grazers at the third stage. This result confirms for M2 what ob- served by Kaiser & Fortelius (2003) and Franz- Odendaal & Kaiser (2003) in some herbivores. Differ- ences in the mesowear pattern shown by upper and lower cheek teeth may depend on the fact that during the masticatory process food frequently is more in con- tact with the lower teeth than with the upper ones. The degree of functional anisodonty and differences in mesowear pattern seem to be more pronounced in mixed feeders then in specialized feeders, though such differences are less evident when a two-tooth model is applied (Fig. 8c). All in all, the obtained results indicate that C. ex gr. C. ropalophorus from Bate Cave had an opportunistic eating habit with a prevalent browser behaviour. These small deer on one hand possibly had a remarkable abil- ity to select the food and change their diet according to the nature and number of competitors, the environ- mental characteristics and their seasonal changes, on the other they likely did not inhabited environments with a low primary productivity. ACKNOWLEDGEMENTS I thank the director and staff of the Paleontological Museum of Sapienza University of Rome for giving me access to deer collections and their assistance during my research. I’m grateful to A.M.F. Valli and the anony- mous reviewer for their useful comments, and to M.R. Palombo for suggestions and valuable discussion. This research was funded by Italian MURST (Sapienza Uni- versity of Rome 2014, project C26A14BNRM led by M.R. Palombo). REFERENCES Anders U., von Koenigswald W., Ruf I., Smith B.H. (2011) - Generalized individual dental age stages for fossil and extant placental mammals. Paläontolo- gische Zeitschrift, 85, 3, 321-339. Caloi L., Palombo M.R. (1991) - Morfologia e funzione della mano e del piede del megacerino del Pleisto- cene superiore di Dragonara (Sardegna, Italia). Atti Acc.- Peloritana dei Pericolanti Classe FF. MM. NN., 57 (1989), suppl. 1, 319-341. Caloi L., Palombo M.R. (1996) - Functional aspects and ecological implications in Hippopotami and Cervids of Crete. In: Reese, D.S. (Ed.), Pleistocene and Holocene Fauna of Crete and its First settlers. Monographs in World Archaeology, Prehistory Press of Madison, 28,125-152. Clauss M., Franz-Odendaal T.A., Brasch J., Castell J. C., Kaiser T. (2007) - Tooth wear in captive giraffes (Giraffa camelopardalis): mesowear analysis classi- fies free-ranging specimens as browsers but captive ones as grazers. J. Zoo. Wild. Med., 38, 433-447. Damuth J., Janis C.M. (2011) - On the relationship be- tween hypsodonty and feeding ecology in ungulate mammals, and its utility in palaeoecology. Biological Reviews, 86, 3, 733-758. De Marinis A.M., Toso S. (2015) - Valutazione dell’età nei Cervidi tramite esame della dentatura. Manuali e Linee guida n. 90/2013, ISPRA. De Vos J. (1979) - The endemic Pleistocene deer of Crete. Proceedings of the Koninklijke Nederlandse Aludemie van Wetenschappen, Amsterdatn, Seies B, 82, 1, 59-90. De Vos J. (1984) - The endemic Pleistocene deer of Crete. Verhandeling der Koninklijke Nederlandse Akadernie van Wetenschappen, Afdeling Natuurkunde, Eerste reeks, Deel 31, 100. North- Holland Publishing Company, Amsterdam, Oxford, New York. De Vos J. (1996) - Taxonomy, ancestry and speciation of the endemic Pleistocene deer of Crete compared with the taxonomy, ancestry and speciation of Dar- win's finches. In: Reese, D.S. (Ed.), Pleistocene and Holocene Fauna of Crete and its First settlers. Monographs in World Archaeology, Prehistory Press of Madison, 28, 111-119. Dietary behaviour of Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete) 87 De Vos J. (2000) - Pleistocene Deer fauna in Crete: its adaptive radiation and extinction. Tropics 10, 1, 125- 134 [Special issue]. Fortelius M. (1985) - Ungulate check teeth: develop- mental, functional, and evolutionary interrelations. Acta Zoologica Fennica, 180, 1-76. Fortelius M., Solounias N. (2000) - Functional charac- terization of ungulate molars using the abrasion- attrition wear gradient: a new method for recon- structing paleodiets. American Museum Novitates, 1 -36. Franz-Odendaal T.A., Kaiser T.M. (2003) - Differential mesowear in the maxillary and mandibular cheek dentition of some ruminants (Artiodactyla). In An- nales Zoologici Fennici, 40, 5, 395-410. Gagnon M., Chew A.E. (2000) - Dietary preferences in extant African bovidae. J. Mammal., 81, 490-511. Geist V. (1999) - Deer of the world: their evolution, be- haviour and ecology. Swan Hill Press. Shrewsbury (UK), 421. van der Geer A., Dermitzakis M., De Vos J. (2006) - Relative growth of the Metapodals in a Juvenile island deer: Candiacervus (Mammalia, Cervidae) from the Pleistocene of Crete. Hellenic J Geosci, 41, 119-125. van der Geer A., Lyras G., De Vos J., Dermitzakis M. (2010) - Evolution of island mammals. Adaptation and extinction of placental mammals on islands. Wiley-Blackwell, London. van der Geer A., Lyras G., MacPhee R.D.E., Lomolino M.V., Drinia H. (2014) - Mortality in a predator-free insular environment: the dwarf deer of Crete, 3807. American Museum novitiates. Hammer Ø., Harper D.A.T., Ryan P.D. (2001) - PAST: Paleontological statistics software package for edu- cation and data analysis. Palaeontologia Elec- tronica, 4(1), 9pp. http://palaeo-electronica. org/2001_1/past/issue1_01.htm Janis C.M. (1979) - Mastication in the hyrax and its rele- vance to ungulate dental evolution. Paleobiology, 5, 1, 50-59. Janis C.M. (1988) - An estimation of tooth volume and hypsodonty indices in ungulate mammals, and the correlation of these factors with dietary preference. In Teeth revisited: proceedings of the VIIth interna- tional symposium on dental morphology (Vol. 53, pp. 367-387). Mémoires du Muséum national d'histoire Naturelle Paris (serie C). Janis C.M., Forteliu, M. (1988) - The means whereby mammals achieve increased functional durability of their dentitions, with special reference to limiting factors. Biol. Rev., 63, 197-230. Kaiser T. M. (2003) - The dietary regimes of two con- temporaneous populations of Hippotherium primi- genium (Perissodactyla, Equidae) from the Vallesian (Upper Miocene) of Southern Germany. Palaeo- geography, Palaeoclimatology, Palaeoecology, 198, 3, 381-402. Kaiser T.M., Croitor R. (2004) - Ecological interpreta- tions of early Pleistocene deer (Mammalia, Cervi- dae) from Ceyssaguet (Haute-Loire, France). Geodi- versitas, 26, 661-674. Kaiser T.M., Fortelius M. (2003) - Differential mesowear in occluding upper and lower molars: opening mesowear analysis for lower molars and premolars in hypsodont horses. Journal of Morphology, 258, 1, 67-83. Kaiser T.M., Kahlke R.D. (2005) - The highly flexible feeding strategy of Stephanorinus etruscus (Falconer, 1859) (Rhinocerotidae, Mammalia) during the early Middle Pleistocene in Central Europe. Berichte der Institutes für Erdwissenschaften Karl- Franzens-Universität Graz, 10, 50-53. Kaiser T.M., Solounias N. (2003) - Extending the tooth mesowear method to extinct and extant equids. Geodiversitas, 25, 321-345. Kaiser T.M., Brasch J., Castell J.C., Schulz E., Clauss M. (2009) - Tooth wear in captive wild ruminant spe- cies differs from that of free-ranging conspecifics. Mammalian Biology, 74, 425-437. Kaiser T.M., Müller D.W., Fortelius M., Schulz E., Codron D., Clauss, M. (2013). Hypsodonty and tooth facet development in relation to diet and habitat in herbivorous ungulates: implications for understand- ing tooth wear. Mammal Review, 43, 1,34-46. Loffredo L.F., DeSantis L.R. (2014) - Cautionary lessons from assessing dental mesowear observer variability and integrating paleoecological proxies of an ex- treme generalist Cormohipparion emsliei. Palaeo- geography, Palaeoclimatology, Palaeoecology, 395, 42-52. Louys J., Meloro C., Elton S., Ditchfield P., Bishop L.C. (2011) - Mesowear as a means of determining diets in African antelopes. Journal of Archaeological Sci- ence, 38, 7, 1485-1495. Mendoza M., Palmqvist P. (2008) - Hypsodonty in ungu- lates: an adaptation for grass consumption or for foraging in open habitat? Journal of Zoology, 274, 134-142. Mihlbachler M.C., Rivals F., Solounias N., Semprebon G.M. (2011) - Dietary change and evolution of horses in North America. Science, 331, 1178-1181. Mujica F. (2014) - Dinamiche evolutive dei cervi endemi- ci delle isole del mediterraneo: analisi eco- morfologiche. Tesi specialistica in Paleontologia, Univ. Roma “La Sapienza”. Palombo M.R. (2005) - Food habits of “Praemegaceros” cazioti (Depéret, 1897) from Dragonara Cave (NW Sardinia, Italy) inferred from cranial morphology and dental wear. In Proceedings of the International Symposium “Insular Vertebrate Evolution: the Palae- ontological Approach”. Monografies de la Societat d’Història Natural de les Balears, 12, 233-244. Palombo M.R., Kohler M., Sola S.M., Giovinazzo C. (2008) - Brain versus body mass in endemic rumi- nant artiodactyls: a case studied of Myotragus bale- aricus and smallest Candiacervus species from Mediterranean Islands. Quaternary International, 182, 1, 160-183. Palombo M.R., Zedda M. (2015) - Surviving in a preda- tor-free environment: Hints from a bone remodelling process in a dwarf Pleistocene deer from Crete. Comptes Rendus Palevol, http://dx.doi.org/10.1016/ j.crpv.2015.03.003 Reese D.S., Belluomini G., Ikeya M. (1996) - Absolute dates for the Pleistocene fauna of Crete. In: Reese, 88 Altamura S. D.S. (Ed.), Pleistocene and Holocene Fauna of Crete and its first Settlers. Monography in World Archaeology, 28, 47–51. Rivals F., Semprebon G.M. (2006) - A comparison of the dietary habits of a large sample of the Pleisto- cene pronghorn (Stockoceros onusrosagris) from the Papago Springs cave in Arizona to the modern Antilocapra americana. Journal of Vertebrate Pale- ontology, 26, 495-500. Rivals F., Solounias N., Schaller, G.B. (2011) - Diet of Mongolian gazelles and Tibetan antelopes from steppe habitats using premaxillary shape, tooth mesowear and microwear analyses. Mammalian Biology-Zeitschrift für Säugetierkunde, 76(3), 358- 364. Rozzi R., Winkler D.E., De Vos J., Schulz E., Palombo M. R. (2013) - The enigmatic bovid Duboisia san- teng (Dubois, 1891) from the Early-Middle Pleisto- cene of Java: a multiproxy approach to its pa- leoecology. Palaeogeography, Palaeoclimatology, Palaeoecology, 377, 73-85. Solounias N., Tariq M., Hou S., Danowitz M., Harrison M., (2014) - A new method of tooth mesowear and a test of it on domestic goats. In Annales Zoologici Fennici, 51, 111-118 Valli A.M., Palombo M.R. (2008) - Feeding behaviour of middle-size deer from the Upper Pliocene site of Saint-Vallier (France) inferred by morphological and micro/mesowear analysis. Palaeogeography, Pa- laeoclimatology, Palaeoecology, 257, 1, 106-122. Valli A.M., Palombo M.R., Alberdi M.T. (2012) - How homogeneous are microwear patterns on a fossil horse tooth? Preliminary test on a premolar of Equus altidens from Barranco Leon 5 (Spain). AMQ- Alpine and Mediterranean Quaternary, 25, 1, 25-33. Yamada E. (2012) - Mesowear analysis of the Japanese sika deer (Cervus nippon) in different food habits: its limitations and applicability. Mammal study, 37, 2, 93 -103. Dietary behaviour of Candiacervus ex gr. C. ropalophorus from Bate Cave (Crete) 89 Ms. received: June 28, 2016 Final text received: September 18, 2016 90 << /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles true /AutoRotatePages /None /Binding /Left /CalGrayProfile (Dot Gain 20%) /CalRGBProfile (sRGB IEC61966-2.1) /CalCMYKProfile (U.S. Web Coated \050SWOP\051 v2) /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Error /CompatibilityLevel 1.4 /CompressObjects /Tags /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages true /CreateJobTicket false /DefaultRenderingIntent /Default /DetectBlends true /DetectCurves 0.0000 /ColorConversionStrategy /CMYK /DoThumbnails false /EmbedAllFonts true /EmbedOpenType false /ParseICCProfilesInComments true /EmbedJobOptions true /DSCReportingLevel 0 /EmitDSCWarnings false /EndPage -1 /ImageMemory 1048576 /LockDistillerParams false /MaxSubsetPct 100 /Optimize true /OPM 1 /ParseDSCComments true /ParseDSCCommentsForDocInfo true /PreserveCopyPage true /PreserveDICMYKValues true /PreserveEPSInfo true /PreserveFlatness true /PreserveHalftoneInfo false /PreserveOPIComments true /PreserveOverprintSettings true /StartPage 1 /SubsetFonts true /TransferFunctionInfo /Apply /UCRandBGInfo /Preserve /UsePrologue false /ColorSettingsFile () /AlwaysEmbed [ true ] /NeverEmbed [ true ] /AntiAliasColorImages false /CropColorImages true /ColorImageMinResolution 300 /ColorImageMinResolutionPolicy /OK /DownsampleColorImages true /ColorImageDownsampleType /Bicubic /ColorImageResolution 300 /ColorImageDepth -1 /ColorImageMinDownsampleDepth 1 /ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true /ColorImageFilter /DCTEncode /AutoFilterColorImages true /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /ColorImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /GrayImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile () /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False /CreateJDFFile false /Description << /ARA /BGR /CHS /CHT /CZE /DAN /DEU /ESP /ETI /FRA /GRE /HEB /HRV (Za stvaranje Adobe PDF dokumenata najpogodnijih za visokokvalitetni ispis prije tiskanja koristite ove postavke. Stvoreni PDF dokumenti mogu se otvoriti Acrobat i Adobe Reader 5.0 i kasnijim verzijama.) /HUN /ITA /JPN /KOR /LTH /LVI /NLD (Gebruik deze instellingen om Adobe PDF-documenten te maken die zijn geoptimaliseerd voor prepress-afdrukken van hoge kwaliteit. De gemaakte PDF-documenten kunnen worden geopend met Acrobat en Adobe Reader 5.0 en hoger.) /NOR /POL /PTB /RUM /RUS /SKY /SLV /SUO /SVE /TUR /UKR /ENU (Use these settings to create Adobe PDF documents best suited for high-quality prepress printing. Created PDF documents can be opened with Acrobat and Adobe Reader 5.0 and later.) >> /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ << /AsReaderSpreads false /CropImagesToFrames true /ErrorControl /WarnAndContinue /FlattenerIgnoreSpreadOverrides false /IncludeGuidesGrids false /IncludeNonPrinting false /IncludeSlug false /Namespace [ (Adobe) (InDesign) (4.0) ] /OmitPlacedBitmaps false /OmitPlacedEPS false /OmitPlacedPDF false /SimulateOverprint /Legacy >> << /AddBleedMarks false /AddColorBars false /AddCropMarks false /AddPageInfo false /AddRegMarks false /ConvertColors /ConvertToCMYK /DestinationProfileName () /DestinationProfileSelector /DocumentCMYK /Downsample16BitImages true /FlattenerPreset << /PresetSelector /MediumResolution >> /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /DocumentCMYK /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /UseDocumentProfile /UseDocumentBleed false >> ] >> setdistillerparams << /HWResolution [2400 2400] /PageSize [612.000 792.000] >> setpagedevice