Dietary reconstruction of pygmy mammoths from Santa Rosa ... · Dietary reconstruction of pygmy...

lable at ScienceDirect

Quaternary International xxx (2015) 1e14

Contents lists avai

Quaternary International

journal homepage: www.elsevier .com/locate/quaint

Dietary reconstruction of pygmy mammoths from Santa Rosa Island ofCalifornia

Gina M. Semprebon a, *, Florent Rivals b, c, d, Julia M. Fahlke e, William J. Sanders f,Adrian M. Lister g, Ursula B. G€ohlich h

a Bay Path University, Longmeadow, MA, USAb Instituci�o Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spainc Institut Catal�a de Paleoecologia Humana i Evoluci�o Social (IPHES), Tarragona, Spaind Area de Prehistoria, Universitat Rovira i Virgili (URV), Tarragona, Spaine Museum für Naturkunde, Leibniz-Institut für Evolutions- und Biodiversit€atsforschung, Berlin, Germanyf Museum of Paleontology and Department of Anthropology, University of Michigan, MI, USAg Natural History Museum, London, UKh Natural History Museum of Vienna, Vienna, Austria

a r t i c l e i n f o

Article history:Available online xxx

Keywords:Island RuleNiche occupationProboscideaTooth microwearPaleodiet

* Corresponding author.E-mail addresses: [email protected],

(G.M. Semprebon).

http://dx.doi.org/10.1016/j.quaint.2015.10.1201040-6182/© 2015 Elsevier Ltd and INQUA. All rights

Please cite this article in press as: SemprebQuaternary International (2015), http://dx.d

a b s t r a c t

Microwear analyses have proven to be reliable for elucidating dietary differences in taxa with similargross tooth morphologies. We analyzed enamel microwear of a large sample of Channel Island pygmymammoth (Mammuthus exilis) molars from Santa Rosa Island, California and compared our results tothose of extant proboscideans, extant ungulates, and mainland fossil mammoths and mastodons fromNorth America and Europe. Our results show a distinct narrowing in mammoth dietary niche space aftermainland mammoths colonized Santa Rosa as M. exilis became more specialized on browsing on leavesand twigs than the Columbian mammoth and modern elephant pattern of switching more betweenbrowse and grass. Scratch numbers and scratch width scores support this interpretation as does thePleistocene vegetation history of Santa Rosa Island whereby extensive conifer forests were availableduring the last glacial when M. exilis flourished. The ecological disturbances and alteration of thisvegetation (i.e., diminishing conifer forests) as the climate warmed suggests that climatic factors mayhave been a contributing factor to the extinction of M. exilis on Santa Rosa Island in the Late Pleistocene.

© 2015 Elsevier Ltd and INQUA. All rights reserved.

1. Introduction

1.1. Background

Endemic to the California Channel Islands, the pygmymammoth(Mammuthus exilis, Maglio, 1970) was initially discovered on theIsland of Santa Rosa and later on Santa Cruz and San Miguel in theChannel Island archipelago of California. M. exilis is a smallmammoth considered to be a dwarfed form of its likely ancestor,the Columbian mammoth (M. columbi), which occupied the main-land of North America (Madden, 1977, 1981; Johnson, 1978). Today,the Channel Islands are comprised of eight islands (Fig. 1). In theLate Pleistocene, the four Northern Channel Islands formed a single

[email protected]

reserved.

on, G.M., et al., Dietary reconoi.org/10.1016/j.quaint.2015.1

super-island, dubbed Santarosae by Orr (1968) and lay closer to themainland than today's Channel Islands. Even during periods ofglaciation in the Pleistocenewhen sea levels weremuch lower thanthey are today, the islands were separated from the California coastby a relatively small water gap of around 6.5e8 km (Roth, 1996;Muhs et al., 2015). As sea levels rose due to the melting of conti-nental ice, 76% of Santarosae disappeared (Johnson, 1972) leavingonly the highest elevations exposed e now known as the islands ofSan Miguel, Santa Cruz, Santa Rosa, and Anacapa. Of these modernislands, all but Anacapa have produced mammoth remains(Agenbroad, 2001). The breakup of Santarosae is believed to havetaken place about 11,000 cal. BP (Kennett et al., 2008).

Researchers have long been interested in the Channel Islands forseveral reasons. First, the islands are part of one of the richestmarine ecosystems in the world and are home to over 150 endemicspecies such as the island fox (Urocyon litteralis) e a small fox withsix subspecies each unique to the island it lives on. Hence, theChannel Islands are often referred to as the North American

struction of pygmy mammoths from Santa Rosa Island of California,0.120

mailto:[email protected]

mailto:[email protected]

www.sciencedirect.com/science/journal/10406182

http://www.elsevier.com/locate/quaint

http://dx.doi.org/10.1016/j.quaint.2015.10.120



Fig. 1. The Channel lslands and Southern California Coast (Modified from Rick et al. (2012) (Fig. 1)). Dotted line indicates Late Pleistocene shoreline of the “super island” ofSantarosae.

G.M. Semprebon et al. / Quaternary International xxx (2015) 1e142

Galapagos. Second, because island species are generally regarded asmore susceptible to human-induced extinctions than those oncontinents, and the Channel Islands were initially occupied byhumans during a period of extensive extinction in North America(and elsewhere) around 13,000 cal. BP (Erlandson et al., 2011), datafrom such islands has proven useful for providing context for lateQuaternary extinctions on continents such as the Americas andAustralia (Steadman and Martin, 2003; Wroe et al., 2006). Third,islands are invaluable for studying evolution and diversification,including the effects of insularity on fauna (Palombo, 2008) such asthe so-called “Island Rule”. The Island Rulewas first stated by Foster(1964) after comparing numerous island species to their mainlandvarieties. He proposed that the body size of a species becomessmaller or larger depending on the resources available to it in itsenvironment. The Island Rule posits that certain island speciesevolve larger size when predation pressure is relaxed (due to theabsence of some mainland predators), while others evolve smallersize due to resource constraints regarding availability of food andland area (Whittaker, 1998; McNab, 2010).

The history of the discovery and excavation of pygmy mam-moths from the islands is chronicled in Agenbroad (2001).Mammoth remains have been known from the Northern ChannelIslands of Santa Rosa, San Miguel, and Santa Cruz since 1856(Stearns,1873). A spectacular findwasmade in 1994 by Park Serviceresearchers led by Larry Agenbroad (a Santa Barbara Museumresearch associate at the time) of a nearly complete adult male M.exilis skeletone amaturemale of about 50 years of age (Agenbroad,1998). After this discovery, a thorough pedestrian survey of theislands using GPS was begun, to document and pinpoint each dis-covery. More than 160 new localities were recorded with the ma-jority of localities found on Santa Rosa Island (Agenbroad et al.,2007). Several mainland-size mammoth (Mammuthus columbi) el-ements in addition to remains ofM. exiliswere recovered as a resultof this survey. The survey of Santa Rosa revealed a ratio ofapproximately 3:10 large mammoth:small mammoth remains(Agenbroad, 2012).

As many as three species ostensibly of different sizes have beenproposed by researchers over the years (Orr, 1956a,b, 1968; Roth,1982, 1996). However, Agenbroad (2009) showed convincingly,using less fragmentary material, that only two species are likelypresent e M. columbi and M. exilis (Fig. 2).

Please cite this article in press as: Semprebon, G.M., et al., Dietary reconQuaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.1

There has been much speculation regarding the origin ofmammoths on the Northern Channel Islands, and the finding ofColumbian mammoth remains is intriguing given that these re-mains may represent remnants of an ancestral population, unlessthey represent later migrants to the island. The oldest remains arefound in the basal conglomerate of the Garanon Member of theSanta Rosa Island Formation. U/Th results formerly suggested anage of at least 200 ka (Orr, 1968), but these are not now consideredreliable and new U/Th data indicate an age of ca. 80 ka (Muhs et al.,2015). Mammoth remains attributed to both M. columbi and M.exilis have been found throughout the entire Formation (the latestcalibrated direct radiocarbon date onM. exilis being 10,700 ± 90 BP(B-14660), equivalent to c. 12,600 cal BP (Agenbroad, 2012).

Also intriguing is the speculation as to how mainland mam-moths arrived on the islands. Initially, it was assumed that theancestral form of M. exilis was either Mammuthus (formerly Archi-diskodon) imperator orMammuthus columbi. It has now been shownthat M. imperator and M. columbi are conspecific (Slaughter et al.,1962; Miller, 1971, 1976; Agenbroad, 2003) and further researchhas suggested that M. columbi represents the likely ancestor of M.exilis (Johnson, 1978; Madden, 1981; Roth, 1982, 1996). This sug-gests that M. exilis evolved according to the Island Rule (Foster,1964). That is, a large continental species (M. columbi) adapted toan island environment becoming in time a new smaller species e

M. exilis (Fig. 2).The question remains e how did Columbian mammoths reach

the islands? Historically, it was assumed that ancestral mammothscould not have swum to the islands. Consequently, various landbridges linking the Northern Channel Islands to the mainland havelong been hypothesized (e.g., Clements, 1955; Van Gelder, 1965;Valentine and Lipps, 1967; von Bloeker, 1967; Remington, 1971).This early idea, that insular mammoth remains proved the exis-tence of a land bridge, persisted for many decades (see synopsis byJohnson, 1978). The idea was deeply entrenched but began to giveway in light of accumulating geological and biological evidence(Savage, 1967; Johnson, 1978). For example, Johnson (1972, 1978)pointed out that a land bridge was not a sine qua non for explain-ing mammoths on the Northern Channel Islands, citing researchthat elephants are excellent distance swimmers, among the best ofall land mammals and highly skilled at crossing water gaps. Hence,currently it is now understood that sea-level fall brought the


Fig. 2. Comparative sizes of the mainland mammoth (Mammuthus columbi) and the pygmy mammoth of the Channel Islands (Mammuthus exilis). Modified from Agenbroad (2012)(Fig. 2). Estimated height data from Agenbroad (2009).

G.M. Semprebon et al. / Quaternary International xxx (2015) 1e14 3

islands only a few kilometers from the shore, enabling mainlandmammoths an opportunity to gain access even if a land bridge wasnot present.

Today the local vegetation of the Northern Channel Islands isdominated by grasslands and shrublands with some patchywoodlands, and mild temperatures with little annual temperaturefluctuation (Junak et al., 2007). To date, studies of the vegetationhistory of the island have included pollen analyses on Santa RosaIsland (Cole and Liu 1994; Kennett et al., 2008), San Miguel Island(West and Erlandson, 1994), continuous offshore pollen cores fromthe Santa Rosa Basin (Heusser, 1995, 2000) and paleobotanicalstudies on Santa Cruz Island (Cheney and Mason, 1930). During theLate Pleistocene, when these islands existed as the single super-island of Santarosae, forest vegetation was present including cy-press, pine, and Douglas fir (Cheney and Mason, 1930; Andersonet al., 2008). Climatic changes at the end of the Pleistocene havebeen invoked as the cause for the loss of coastal forest (Andersonet al., 2008), and major fires coincident with the last knownoccurrence of pygmy mammoths (around 13,000 cal BP) have beendemonstrated (Kennett et al., 2008). Anderson et al. (2010) com-bined pollen analysis with an analysis of the fire disturbanceregime of Santa Rosa Island and reported that a coastal coniferforest covered the highlands of Santa Rosa during the last glacialbut that by ca. 11,800 cal. BP Pinus stands, coastal sage scrub andgrassland replaced the forest as the climate warmed.

1.2. Aims of the study

The purpose of this study is twofold: 1) to explore the paleo-dietary ecology of Mammuthus exilis from Santa Rosa Island, Cali-fornia using stereomicrowear analysis, and 2) to comparemicrowear patterns of M. exilis to those of extant elephants and toother mammoths and mastodons from North America and Europe,especially its ancestor M. columbi. To test the hypothesis that M.exilis was exploiting the forest vegetation during the Late Pleisto-cene, we employed light microscopy dental microwear on a largesample of pygmy mammoth teeth from Santa Rosa Island. If forestvegetation including woody browse was exploited on Santa RosaIsland during the Late Pleistocene, we predict that: 1. M. exiliswould exhibit more scratches in the low scratch range (i.e., be-tween 0 and 17) typical of extant mammals that feed on leaves; 2.Some individuals of M. exilis would display the unusually deep andwide scratches (i.e., hypercoarse scratches) found in extant


proboscideans and other mammals known to ingest large quanti-ties of twigs and bark.

1.3. Microwear

Microwear has been employed for over three decades as atechnique to visualize scars in dental enamel caused by food itemssuch as plant phytoliths or by exogenous grit or dust coating thesurface of vegetation (e.g., Rensberger, 1978; Walker et al., 1978;Solounias and Semprebon, 2002; Semprebon et al., 2004;Merceron et al., 2004, 2005; Scott et al., 2005, 2006; Ungar et al.,2008, 2010) including studies of mammoths and other pro-boscideans (Rivals et al., 2012, 2015). Microwear turns over rela-tively rapidly, thus giving insight into the dietary behavior of thelast days, or weeks, before an animal's death e the so-called “LastSupper Effect” (Grine, 1986). As such, it provides a window into theshort-term dietary behavior of a taxon as opposed to the deep-timeadaptation of gross tooth morphology, giving insight into what ataxon was actually eating despite what it might originally havebeen adapted to eat (Rivals and Semprebon, 2011). Importantly, thissnapshot into the habitat and dietary behavior of a taxon is largelytaxon-independent.

While microwear analyses have proven to be very reliable indiagnosing dietary behavior of mammals, two considerations arerelevant to all microwear approaches: 1) potential error involved inthe analyses, and 2) the causal factors producing microwear scars.Different microwear methodologies have different strengths andweaknesses, but all have value depending on questions beingstudied. The three methods currently used involve scanning elec-tron microscopy (SEM), light microscopy for dental microwear(LDM), and dental microwear texture analysis (DMTA). BecauseSEM and LDM rely on observer measurements, while DMTA em-ploys an automated approach, studies have been done to identifyand/or quantify both inter- and intra-observer error involved ineach (e.g., Grine et al., 2002; Galbany et al., 2005; Fraser et al., 2009;Mihlbachler et al., 2012 DeSantis et al., 2013; Williams and Geissler,2014; Hoffman et al., 2015).

While these studies have provided valuable insight regardingelucidating potential error in microwear studies, it is important tonot generalize results utilizing one methodology to another. Thus,microwear error studies aimed at testing LDM (other thanSemprebon et al., 2004), while adding to the cannon of knowledge,have not duplicated the original LDM methodological regime (i.e.,



have employed different magnifications, used different countingareas, have not counted all features visible, and/or assessed micro-wear from photographs e sometimes using untrained observers).

The original methodology developed by Semprebon, 2002 andSolounias and Semprebon, 2002 (and utilized in this study) in-volves counting pits and scratches directly using a stereoscope at 35times magnification following the same standard microscope cali-bration and systematic counting methods employed every day indiagnostic medicine and life science research to accurately countsubstances such as blood cells and platelets using light microscopy(where differential light refraction as used in this study is used toidentify the objects being counted).

For example, the procedure used to construct the comparativeextant ungulate and proboscidean databases (Semprebon, 2002;Rivals et al., 2012) and to analyze the fossil proboscideans used inthis study relies on employing a calibrated ocular reticle subdividedinto 9 subsquares which greatly enhances the ability to not losetrack of features counted rather than attempting to count numerousfeatures scattered within a larger, undivided field of view. Addi-tionally, features are counted according to a strict protocol thatensures that features present are not skipped over or counted morethan once (Estridge and Reynolds, 2012). Also, a relatively largecounting area is employed and features are counted twice (toreduce error due to potential variable scar distribution). When thisprotocol is followed, both interobserver and intraobserver error hasbeen demonstrated to be low (Semprebon et al., 2004).

In some studies, however, feature counts have been made fromimages using much higher magnifications than used in Semprebonet al. (2004) and counting error would naturally be expected to behigher due to the greater number of features visible at highermagnifications. For example, Mihlbachler et al. (2012) reported thatinterobserver error was higher than that reported by Semprebonet al., 2004, although evidence was found (2012) that error dimin-ished significantly with increased experience of the observer anddata collected by all observers was highly correlated (Mihlbachleret al., 2012). Also, while DMTA has been employed to extract 2Dmicrowear data from photosimulations of scanned areas of DMTA,(e,g., DeSantis et al., 2013), stereomicroscopy was invented to pro-vide depth offield and a three dimensional visualization of an objectbeing examined and thus iswidely used inmicrosurgery, dissection,forensics, medicine, industry, and quality control. In microwearanalysis as employed here and in Solounias and Semprebon (2002)and Semprebon et al. (2004), stereomicroscopy allows relativedepth and 3D shape of different features to be assessed.

While fully automated methodologies such as DMTA have beendeveloped and have proven useful to reduce subjectivity andinterobserver error (Ungar et al., 2003; Galbany et al., 2005) it isimportant to note, that DMTA reduces interobserver error onlywhen the same exact location on the same tooth surface is scannedtwice, as results obtained on the same tooth and same facet may bequite different when different regions of that tooth and facet arescanned (see Fig. 8 in Scott et al., 2006). This difference is due to thenatural heterogeneity of microwear on tooth surfaces, a propensitywhich affects all microwearmethodologies but is more problematicat higher versus lower magnifications and when only small areas ofa tooth are analyzed. If relatively low magnification and a largecounting area is assessed such as in the Solounias and Semprebon(2002) methodology, both inter- and intra-observer error may beminimized (see Williams and Geissler, 2014).

Regardless, the effects of potential error should be evaluatedwithin the context of whether accurate dietary categorization ofliving mammals with known diets can be obtained by single ob-servers. A variety of different variations on the original LDMmethodology have produced great fidelity in terms of partitioningliving mammals (e.g., carnivores, cetaceans, proboscideans,


primates, marsupials, ungulates, and rodents) into their knowntrophic groups (Semprebon et al., 2004; Merceron et al., 2004,2005; Rivals et al., 2007, 2012; Townsend and Croft, 2008; Fraseret al., 2009; Williams and Patterson, 2010; Fraser and Theodor,2011, 2013; Williams and Holmes, 2011; Bastl et al., 2012; Fahlkeet al., 2013; Christensen, 2014; Williams and Geissler, 2014).

The causal factors involved in producing microwear scars areonly beginning to be understood. Microwear patterns must beinfluenced by the differing physical properties of enamel andingested abrasives such as biogenic silica (i.e., phytoliths) fromplants, or abiotic silica from dust or grit. While it is also possiblethat differing tooth morphologies and masticatory biomechanicsmay impact the nature of microwear scars, a similar pattern ofmicrowear scars have been observed across various orders ofmammals consuming similar dietary items which seems to arguethat such effects are relatively minor (e.g., Semprebon et al., 2004)although further study is needed. A number of studies have testedthe effects of abrasives in producing microwear scars and haveexpanded our thinking away from considering phytoliths alone ascontributing to microwear scar topography (Covert and Kay, 1981;Kay and Covert, 1983; Maas, 1991, 1994; Gügel et al., 2001;Mainland, 2003; Sanson et al., 2007; Lucas et al., 2013; Schulzet al., 2013; Müller et al., 2014; Hoffman et al., 2015). Studiesexamining the relative hardness of enamel versus phytoliths have,however, produced contradictory results (Baker et al., 1959; Sansonet al., 2007; Rabenold and Pearson, 2014; Xia et al., 2015). The mostpromising studies regarding teasing apart causal factors involvecontrolled feeding experiments. These studies have elucidatedpatterns in such widely diverse taxa as American opossums (Covertand Kay, 1981; Kay and Covert, 1983), rabbits (Schulz et al., 2013),and ungulates (Hoffman et al., 2015). Such studies have docu-mented an increase in scratch numbers in rabbits when their dietcontained grass with more biogenic abrasives (Schulz et al., 2013)and in opossums that were fed fine-grained pumice (Covert andKay, 1981; Kay and Covert, 1983). Hoffman et al. (2015) providedthe controlled feeding experiment with domesticated sheep, andtested the effects of abiotic abrasives of different sizes. This studyreported a significant increase in pit features that was correlatedwith an increase in grain size. This observed grit effect supports theinterpretation of increased grit consumption by extant ungulatesfrom semi-arid and arid environments due to increased pittingobserved in these forms (Solounias and Semprebon, 2002).

LDM (after Semprebon, 2002; Solounias and Semprebon, 2002)was used in this study as it allows for the attainment of relativelylarge sample sizes (numbers of specimens), and a very largecomparative stereomicrowear database exists which is comprisedof extant artiodactyls, perissodactyls, and proboscideans of knowndiets compiled by a single observer (GMS). Relatively few studieshave concerned themselves with proboscidean microwear patterns(examples are Palombo and Curiel (2003), Palombo (2005), Greenet al. (2005), Todd et al. (2007) and Calandra et al. (2008, 2010)).Rivals et al. (2012, 2015) studied dietary diversity patterns inPleistocene representatives of Mammut, Anancus, Palaeoloxodonand Mammuthus from a variety of localities in Europe and NorthAmerica. In this study, we test the effect of insular dwarfism ondietary niche occupation by comparing Pleistocene mammothsfrom the North American mainland with island mammoths fromthe Channel Islands off the coast of California.

2. Materials and methods

2.1. Sample

Molar teeth of Mammuthus exilis from the Channel Islands ofCalifornia were sampled from the collections of the Santa


Fig. 3. Mandible of Mammuthus exilis from Santa Rosa Island, California (Specimenfrom Santa Barbara Museum of Natural History e catalogue number SBMNH-VP 1078).Arrow depicts area sampled for microwear analysis (i.e., the central portion of thecentral enamel bands of the grinding facet of the occlusal surface). Scale bar ¼ 5 cm.


Barbara Museum of Natural History, California. The specimenswere screened under a stereomicroscope for potential tapho-nomic alteration of dental surfaces. Those with badly preservedenamel or taphonomic defects (i.e., features displaying unusualmorphologies or fresh features made during the collecting pro-cess or during storage) were removed from the analysisfollowing King et al. (1999). We also excluded slightly or heavilyworn teeth.

Fifty-one M. exilis adult molar casts from Santa Barbara Island,two from SanMiguel Island, and one from Santa Cruz Island, as wellas one tooth ofM. columbi from Santa Rosa Island, were determinedto be suitable for analysis and subjected to microwear analysis by asingle observer (GMS) who also analyzed the extant ungulate andproboscidean comparative samples.

2.2. Microwear technique

Tooth surfaces were cleaned, molded, cast and examined at 35times magnification with a Zeiss Stemi-2000C stereomicroscopeafter the methodology of Solounias and Semprebon (2002). Theanalysis was made from the central portion of the central enamelbands of the occlusal surface (Fig. 3). Microscopic scars werevisualized using external oblique illumination. An M1-150 high-

Table 1Microwear summary data for the second molars of extant and fossil proboscidean samp

Species N Pit mean Pit SD Scratch mean Scrat

Loxodonta africana 33 22.9 3.9 17.4 5.3Loxodonta cyclotis 6 29.8 4.0 12.9 5.1Elephas maximus 10 20.9 2.6 18.3 4.3Mammuthus exilis e Santa Rosa Island 51 26.9 5.99 10.78 4.59Mammuthus exilis e San Miguel Island 2 24.0 e 7.8 e

Mammuthus exilis e Santa Cruz Island 1 22.0 e 8.5 e

Mammuthus columbi e Santa Rosa Island 1 31.5 e 19.5 e

Abbreviations: N¼Number of specimens; Pit Mean ¼ Mean number of pits; Pit SD ¼ standeviation of scratches; SWS¼Mean Scratch width score from 0 to 4 (0¼ fine scratches onlarge pits; %G ¼ Percentage of individuals per taxon with gouges; %F¼ Percentage of indicoarse scratches; %M¼ Percentage of individuals per taxonwith amixture of fine and coarhypercoarse scratches; %H¼ Percentage of individuals per taxon with hypercoarse scratcscratch counts between 0 and 17; B ¼ leaf browser; MF ¼ mixed feeder; G ¼ grazer. All


intensity fiber-optic light source (Dolan-Jenner) was directed acrossthe surface of casts at a shallow angle to the occlusal surface.

The mean number of pits (rounded features) versus meannumber of scratches (elongated features) per taxon were countedwithin a 0.4mm square area (0.16mm2) by GMS. It was also noted ifmore than four large pits (i.e., deep pits with low refractivity thatare at least twice the diameter of highly refractive small pits) werepresent or absent per microscope field (within the 0.4 mm squarearea) and whether gouges (i.e., depressions with ragged, irregularedges that are 2e3 times larger than large pits) were present orabsent (after Semprebon, 2002; Solounias and Semprebon, 2002).Results were compared with extant ungulate and proboscideanmicrowear databases to determine the dietary categories ofbrowser versus grazer (Semprebon, 2002; Solounias andSemprebon, 2002). Scratch textures were assessed as being eitherfine, coarse, hypercoarse, a mixture of fine and coarse, or a mixtureof coarse and hypercoarse scratch types per tooth surface using thedifferential light refractivity of each type of scratch to categorizethem (after Semprebon et al., 2004). In addition, a scratch widthscore (SWS) was obtained by assigning a score of 0 to teeth withpredominantly fine scratches per tooth surface, 1 to those with amixture of fine and coarse types of textures, 2 to those with pre-dominantly coarse scratches, 3 to those with a mixture of coarseand hypercoarse textures, and 4 to those with predominantlyhypercoarse scratches. Univariate statistics, ANOVA, and Tukey'spost-hoc test for honest significant differences were calculatedusing PAST software (Hammer et al., 2001).

Because extant seasonal mixed feeders and those that alternatetheir diet when migrating to different regions alternately feed onbrowse and grass may overlap the browsing or grazing meanscratch/pit morphospaces, the percentage of individuals within ataxon that possess raw scratch values falling into the low scratchrange (i.e., between 0 and 17) were calculated, since extant leafbrowsers, grazers, and mixed feeders show distinctive anddiscriminating low scratch range patterns (Semprebon, 2002;Semprebon and Rivals, 2007).

3. Results

3.1. Comparative extant proboscidean microwear

Results of the microwear analysis are shown in Table 1. It is clearfrom Fig. 4A that extant Loxodonta africana and Elephas maximusexhibit a varied dietary behavior e taking both grass and browse,whereas the forest elephant, Loxodonta cyclotis, exhibits less vari-ability and is more focused on the browse end of the spectrum(note, however, that fewer specimens were available for L. cyclotisthan for the others). This corroborates what is known from fieldstudies (Sukumar, 2003).

les.

ch SD SWS % LP % G %F % C % M % CH % H % 0e17 S Diet

3.1 54.6 36.4 0.0 0.0 0.0 87.9 12.1 39.4 MF2.8 50.0 33.3 0.0 0.0 16.7 66.7 16.7 83.3 B3.1 70.0 50.0 0.0 0.0 0.0 90.0 10.0 40.0 MF2.0 37.3 13.7 7.84 11.77 33.33 47.06 0.0 94.1 B4.0 50.0 0.0 0.0 0.0 0.0 0.0 100 100 B1.0 0.0 0.0 0.0 0.0 100 0.0 0.0 100 B2 100 0 0 100 0 0 0 0 G

dard deviation of Pits; Scratch ¼ Mean number of scratches; Scratch SD ¼ standardly, to 4¼ hypercoarse scratches only; %LP¼ Percentage of individuals per taxonwithviduals per taxon with fine scratches; %C¼ Percentage of individuals per taxon withse scratches; %CH¼ Percentage of individuals per taxonwith amixture of coarse andhes; %0-17S ¼ low scratch percentage (i.e., Percentage of individuals per taxon withextant and extinct specimens were scored by GMS.



3.2. Pygmy mammoth molar microwear compared with extantproboscideans

The raw scratch distribution of M. exilis is skewed more towardthe low scratch browsing range, much like that of Loxodonta cyclotis(Fig. 4B), indicating a less heterogeneous dietary regime than thatfound in L. africana or E. maximus. Two teeth of M. exilis from SanMiguel Island and one from Santa Cruz Island were available andthese specimens also plot within the extant ungulate browsingecospace.

Table 2ANOVA and Tukey's HSD test results for differences in numbers of pits, numbers scratch

Number of scratches

ANOVA results e Test for Equal Means:Sum of Squares df

Between groups: 1107.22 3Within groups: 2259.03 96Total: 3366.25 99Tukey's honest significance test: Pair-wise comparisons e q values below the diagon

L. cyclotis L. africanaLoxodonta cyclotis e 0.0972Loxodonta africana 3.303 e

Elephas maximus 3.945 0.6417Mammuthus exilis 1.57 4.873

Number of pits


Between groups: 636.545 3Within groups: 2428.51 96Total: 3065.06 99Tukey's honest significance test:



Scratch width score





Gouges





Large pits






An ANOVA (Table 2) indicates that M. exilis scratch numbers aresignificantly different (i.e., lower) from those of L. africana (Tukey'sHSD Test, p ¼ 0.0047) and E. maximus (Tukey's HSD Test,p¼ 0.0011) but are not significantly different from scratch numbersof L. cyclotis (Tukey's HSD Test, p ¼ 0.6844). Also, M. exilis pitnumbers are significantly different (i.e., higher) from those of E.maximus (Tukey's HSD Test, p ¼ 0.0189) but not significantlydifferent from those of L. africana (Tukey's HSD Test, p ¼ 0.193) orL. cyclotis (Tukey's HSD Test, p ¼ 0.4508).

es, and scratch width score. Significant differences are indicated in bold.

Mean square F p369.075 15.68 <0.00123.5315

al, p values above the diagonal.E. maximus M. exilis0.0319 0.68440.9688 0.0047e 0.00115.514 e

Mean square F p212.182 8.388 <0.00125.297

E. maximus M. exilis0.0002 0.45080.7648 0.193e 0.018934.214 e

Mean square F p10.2437 14.85 <0.0010.689884


Mean square F p0.577029 3.164 0.02800.182385


Mean square F p0.402204 1.629 0.18770.24691



Fig. 4. A. Bivariate plot of the number of pits versus the number of scratches for theextant proboscideans analyzed graphed in relation to Gaussian confidence ellipses(p ¼ 0.95) on the centroid for extant ungulate browsers (B) and grazers (G) (convexhulls) adjusted by sample size (ungulate data from Semprebon (2002) and Solouniasand Semprebon (2002); extant proboscidean data compiled by GMS). B. Bivariateplot showing the raw scratch and pit distribution ofM. exilis compared to that of extantungulates from Semprebon (2002) and Solounias and Semprebon (2002). C. Meanscratch versus mean pit results for M. exilis from Santa Rosa Island compared to extantungulates (convex hulls), and living and fossil proboscidean mean scratch/pit values.Key: 1 ¼ Mammut americanum (Phosphate Beds, SC USA); 2 ¼ M. americanum(Ingleside, San Patricio Co., TX, USA); 3 ¼ M. americanum (Various Localities, FL,USA);4 ¼ M. exilis, Santa Rosa Island, CA, USA; 5 ¼ M. columbi (Phosphate Beds, SC, USA);6 ¼ M. columbi (Santa Rosa Island, CA USA); 7 ¼ M. columbi (Ingleside, San Patricio Col,



Among Mammuthus exilis from Santa Rosa Island (Table 1,Fig. 5A), 94.1% of individuals have scratches that fall between 0 and17, a pattern typical of browsing ungulates and those of the extantLoxodonta cyclotis individuals analyzed here, but distinctive fromextant species with a greater breadth of scratch values such asungulate mixed feeders, Loxodonta africana and Elephas maximus.

When additional microwear variables are considered, Mam-muthus exilis (Fig. 6), exhibits some distinct patterns from those ofextant elephants. Differences in large pitting and gouging (Fig 6A)are not statistically significant when M. exilis is compared to livingelephants (Table 2) but there are statistically significant scratchtextural differences between M. exilis and Loxodonta africana andElephas maximus (Fig. 6B), while scratch textural differences be-tweenM. exilis and Loxodonta cyclotis are not statistically significant(Table 2). Scratch texture is a reflection of differences in widths anddepths of scratches e i.e., fine scratches are the narrowest andshallowest while hypercoarse scratches are thewidest and deepest.Thus, the enamel surface of M. exilis typically displays finelytextured and mixtures of finely textured and coarse scratches, andfewer mixtures of coarse and hypercoarse scratches or purelyhypercoarse scratches than seen in the African savannah and Asianelephants (Table 1, Fig. 6B).

Fig. 7 shows that typical extant leaf-dominated ungulatebrowsers as a group have scratch width scores that range between0 and 0.46 (i.e., they possess mostly finely textured scratches);whereas typical grazers range from 0.70 to 1.67 (i.e., they havefewer fine and more coarse scratches) and do not overlap leafbrowsers. Mixed feeders that consume browse and grass regionallyor seasonally predictably overlap both leaf browsers and grazersand range from 0 to 1.14 (data from Semprebon, 2002). What isstriking about Fig. 7, is that both modern proboscideans and M.exilis are truly unique in terms of their larger mean scratch widthscores when compared to other herbivorous mammals studiedwith stereomicrowear data thus far (but note that M. exilis hasnarrower scratches than modern proboscideans).

3.3. Pygmy mammoth microwear compared to mainlandmammoths and mastodons

Mammuthus exilis clearly hasmean scratch and pit results nearlyidentical to the forest elephant (Loxodonta cyclotis) and distinctfrom L. africana and Elephas maximus. Mammuthus exilis has meanscratch and pit results that are more similar to those found in priorstudies for the mastodonMammut americanum than those found inmainland mammoths (Figs. 4C and 5B), the latter taxa typicallydisplaying values either intermediate between the browsing andgrazing extant morphospaces, or skewed towards grazing,depending on the species (Green et al., 2005; Rivals et al., 2012).

In addition, an examination of enamel surfaces (Fig. 8) suggestsa more attritive diet for M. exilis compared to most other mam-moths studied. Note the relatively polished, flat, and shiny enamelsurface with relatively few food scars in M. exilis compared to therelatively dull and more irregular enamel surface with many foodscars in M. primigenius. Fortelius and Solounias (2000) provide acomprehensive study and discussion of attritive versus abrasivemolar wear due to mastication of various types of food items. Taxathat consume relatively soft food items such as browse experiencemuch tooth-on-tooth (attritive) polishing of opposing wear facetsas molars cut completely through relatively soft food and produce

TX, USA); 8 ¼ M. columbi (Quarry G., Sheridan Co., NE, USA); 9 ¼ M. columbi (Grayson,Sheridan Co., NE, USA); 10 ¼ M. primigenius (Fairbanks Area, AK, USA);11 ¼ M. primigenius (Brown Bank, North Sea); 12 ¼ M. trogontherii (Crayford - SladeGreen, UK); 13 ¼ M. trogontherii (Ilford, UK). Data: 1e2, 5, 7e14 from Rivals et al.(2012); 3 from Green et al. (2005); 4, 6 from this study).


Fig. 5. Box plot of the percentage of scratches per taxon that fall within the low scratch range (i.e., between 0 and 17). Boxes represent results from extant ungulates. A. Low scratchrange scores forMammuthus exilis (from Santa Rosa Island) compared to results for extant ungulates and extant proboscideans. B. Low scratch range scores for mainland mammoths(key as in Fig. 7 legend). The box represents the central 50% of the scratch values and the bars represent the range (minimal and maximal values).


relatively polished and flat enamel facets. Taxa that consumerelatively abrasive foods such as grass experience more abrasivefood-on-tooth wear which leads to a less polished and moreirregular enamel surface.

4. Discussion

4.1. Pygmy mammoth microwear compared to extant elephants

In this study, M. exilis exhibited a microwear scratch patternsimilar to (and not significantly different from) that of Loxodontacyclotis (i.e., scratch results concentrated toward the browsingmorphospace), in contrast to that of L. africana or Elephas max-imus (scratch results that extend more into the mixed feeding andgrazing zones). These results are consistent with known differ-ences in habitat and dietary habits of L. cyclotis, L. africana and E.maximus.

Loxodonta cyclotis (African forest elephant), found today inwestern and central Africa, mainly occupies lowland tropical rain-forests, swamps, and semi-evergreen and semi-deciduous tropicalrainforests. Forest elephants are known to change habitatsseasonally, occupying lowland rainforest during thewet season andmoving to swampy areas during the dry season (Merz, 1986a,b; Fayand Agnagna, 1991; Dudley et al., 1992; White et al., 1993; Tangley,1997). Loxodonta cyclotis subsists mainly on leaves, bark, fruit, and


twigs of rainforest trees and varies its diet regionally according tothe availability of trees and fruits (White et al., 1993; Tangley, 1997;Eggert et al., 2003). An extensive seven-year observational study ofL. cyclotis from the Lop�e Reserve in Gabon (White et al., 1993)determined that 70% of the diet of L. cyclotis in terms of the numberof species and amounts eaten came from leaves and bark.

In contrast, Loxodonta africana (African savannah elephant) andElephas maximus (Asian elephant) are more generalist feeders,relying on more widely distributed resources including grasses,shrubs, herbs, palms and vines and broadleaved trees (Nowak,1999; Osborn, 2005; Archie et al., 2006; Rode et al., 2006;Wittemyer et al., 2007). Loxodonta africana extended historicallyfrom north of the Sahara Desert to the southern tip of Africa but isnow found in increasingly fragmented habitats (deserts, savannas,forests, river valleys and marshes (Estes, 1999). E.maximus is foundtoday in Southeast Asia from India to Borneo but was formerlymore widely distributed from the Levant, throughout SoutheastAsia, and in China as far north as the Yangtze River. Like L. africana,E.maximus used a greater variety of habitats in the past than it doestoday, as currently ecotones with an interdigitation of grass, forests,and low woody plants are favored as continuously forested areaswithout regular clearings do not support high densities (Shoshaniand Eisenberg, 1982). Both L. africana and E. maximus alternatebetween a predominantly browsing strategy with that of grazingbecause of seasonal variation of food plants, seasonally changing


Fig. 6. Bar charts showing percentages of individuals per taxon in M. exilis and extantproboscideans with large pits and gouges (Fig. 6A) and the percentage of individualsper taxon of M. exilis and extant proboscideans displaying fine, mixed fine and coarse,coarse, mixed coarse and hypercoarse and hypercoarse scratch textures (Fig. 6B).


mineral content, and migration habits (Joshi and Singh, 2008;Mohaptra et al., 2013).

Our microwear results suggest that pygmy mammoths fromSanta Rosa engaged in a fairly restricted dietary regime more like

Fig. 7. Box plot of scratch width scores for extant ungulates and proboscideans and M. exilicoarse scratches, 2 ¼ those with mostly coarse scratches, 3 ¼ those with a mixture of coarseplot, the center vertical line marks the median of the sample, the central 50% of the scratch vof scratch values.


that of the extant L. cyclotis than the more highly migratory L.africana and E. maximus. Thus, some microwear features inscribedinto enamel of M. exilis, such as scratch numbers and textures andthe appearance of the enamel surfaces themselves, are morecongruent with a less abrasive diet than African savannah andAsian elephants. The enamel surface of M. exilis is flatter (enamelband edges sharp and less rounded), more polished and less scarredthan African savannah and Asian elephants and more similar to L.cyclotis and other predominantly browsing forms (Solounias andSemprebon, 2002).

Interestingly, no individual of M. exilis exhibits mostly hyper-coarse scratches as is seen in some individuals of all species of livingelephants (Solounias and Semprebon, 2002). Hypercoarsescratches are relatively unusual features that have thus far mainlybeen reported in those taxa known to consume appreciableamounts of bark (Solounias and Semprebon, 2002; Green et al.,2005) or are known hard-object processors such as primate hard-fruit and seed consumers (Semprebon et al., 2004). It is possiblethat these scratch textural results are due to bark consumption, aspuncture-like large pits typical of fruit and seed consumers werenot observed. The fact that fewer individuals of M. exilis possessmixed coarse and hypercoarse or purely hypercoarse scratches thanthe African savannah and Asian elephants raises interesting ques-tions regarding the relative importance of bark consumption inthese forms.

It is conceivable, however, that because proboscideans are largeanimals with large jaw muscles, they might concentrate morepressure when biting on a hard object compared to smaller moreselective-feeding mammals and that this, rather than bark con-sumption, might produce hypercoarse scratches. While this shouldnot be discounted as a possibility or as a contributing factor,hypercoarse scratches are not simply deeper than coarse or finescratches, they are also wider. This augmented scratch width isconcordant with the tendency of bark to include clustered silicawhereby soil minerals are concentrated and cemented into silicaaggregates (Sangster and Hodson, 1992; Albert and Weiner, 2001).Also, if bite force alone were responsible for producing hypercoarsescratches, other microwear scars would be expected to be deeperand we have not found this to be the case. In addition, an importantfinding that bears on this question is the presence of hypercoarsescratches in the enamel of Florida mastodons with preserved

s. A score of 0 ¼ teeth with mostly fine scratches, 1 ¼ those with a mixture of fine andand hypercoarse scratches, and 4 ¼ those with mostly hypercoarse scratches. In eachalues fall within the range of the box, and the top and bottom bars represent the range


Fig. 8. Photomicrographs of molar enamel surfaces (@35X magnification) of M. exilis e SBMNH e VPN 838 (left) compared toM. primigenius e AMNH 651012-1952 (right). Note therelatively flat and polished enamel surface of M. exilis, typical of attritive feeders versus the uneven topography and duller enamel surface of M. primigenius which is typical ofabrasion-dominated feeders. Both specimens were analyzed using the same lighting regime. Scale bar ¼ 0.4 mm.


digesta consisting of a heavy concentration of cypress twigs (Webbet al., 1992; Mihlbachler, 1998; Green et al., 2005). While the aboveare consistent with hypercoarse scratches being caused by barkconsumption, further study is needed to tease apart the contrib-uting factors that produce such distinctive scratches inproboscideans.

4.2. Pygmy mammoth microwear compared to mainlandmammoths and mastodons

Our results indicate thatM. exilismost likely engaged in a diet ofrelatively low abrasion (polished enamel) and less dietary breadth(few scratches) than the majority of mainland mammoths studiedthus far, including its likely ancestor e M. columbi.

Our data show (Table 1 and Fig. 4B) that 94% of individuals in oursample ofM. exilis have scratch numbers that fall within a relativelynarrow low scratch (browsing type) range while only the forestelephant (L. cyclotis) and one of the mammoth samples we studied(Mammuthus columbi from the Phosphate Beds of South Carolina e

Fig. 4C) also fall within this range. Two questions to consider wheninterpreting these data are: (1) do the relatively narrow browse-dominated scratch results of M. exilis reflects its actual dietaryrange or were the specimens analyzed collected from a particularhorizon or facies that may underrepresent the diversity of habitats(and diets), and (2) since microwear represents a snapshot in timeof dietary behavior (i.e., the diet at the time of an animal's death), isit possible that a concentration of deaths in a particular seasonmight have led to a “last supper effect” (Grine, 1986)?

Firstly, the sample available to us was drawn from a variety offacies (Agenbroad, 2001) so our results most likely show fidelity tothe actual dietary range of M. exilis. However, if food supply orclimate were highly seasonal on Santa Rosa Island during thetenure of M. exilis on the island, it is quite plausible that deathswould not have been spread evenly throughout the year, especiallyon an island where seasonal migration is not possible.

Interestingly, the enamel surfaces andmean scratch/pit patternsof M. exilis are more similar to those found in prior studies for theAmerican Mastodon e Mammut americanum (Green et al., 2005)than to the results reported for Middle to Late Pleistocene Mam-muthus from North America and Europe (Perez-Crespo et al., 2012;


Rivals et al., 2012) which indicate more seasonal or regional mixedfeeding (i.e., alternate consumption of browse and grass) in themajority of forms, although exceptions do exist. For example,Mammuthus columbi from the Phosphate Beds of South Carolina inNorth America apparently browsed (Fig. 4C, point labeled 5; Rivalset al., 2012).

4.3. Channel Island Paleoecology

Today the local vegetation of the Northern Channel Islands isdominated by grasslands and shrublands with some patchywoodlands, and mild temperatures with little annual temperaturefluctuation (Junak et al., 2007). To date, studies of the vegetationhistory of the island have included pollen analyses on Santa RosaIsland (Cole and Liu 1994; Kennett et al., 2008) and San MiguelIsland (West and Erlandson, 1994), and paleobotanical studies onSanta Cruz Island (Cheney and Mason, 1930). During the LatePleistocene, when these islands existed as the single super-island ofSantarosae, forest vegetation was present including cypress, pine,and Douglas fir (Cheney and Mason, 1930; Anderson et al., 2008).Climatic changes at the end of the Pleistocene have been invoked asthe cause for the loss of coastal forest (Anderson et al., 2008), andmajor fires coincident with the last known occurrence of pygmymammoths (around 13,000 cal. BP) have been demonstrated(Kennett et al., 2008). Anderson et al. (2010) combined pollenanalysis with an analysis of the fire disturbance regime of SantaRosa Island and reported that a coastal conifer forest covered thehighlands of Santa Rosa during the last glacial but that by ca.11,800 cal. BP Pinus stands, coastal sage scrub and grasslandreplaced the forest as the climate warmed.

Our molar microwear results support our hypothesis that M.exiliswas exploiting forest vegetation on Santa Rosa during the LatePleistocene sinceM. exilis individuals possessed scratchesmainly inthe low scratch range typical of leaf browsing extant ungulates andother mammals as well as the unusually deep and wide scratchestypical of bark and twig consumers such as extant proboscideansand black rhinos (see Solounias and Semprebon, 2002). These re-sults are consistent with what is known of the Pleistocene vege-tation history of Santa Rosa Island whereby extensive coastalconifer forests were available during the last glacial as well as Pinus



stands and sage scrub as the climate warmed (Anderson et al.,2010). Therefore, abundant browse was apparently available to M.exilis, and consumption of leaves and twigs, as hypothesized here iscompatible with available Pleistocene vegetation on the island.

4.4. Extinction of M. exilis

There is much debate surrounding the extinction of Late Qua-ternary megafauna. The leading hypotheses are human overkill,climatic fluctuations, or vegetational change, with widespreaddisease, or extraterrestrial impact events sometimes invoked(Martin and Klein,1984; Grayson, 2001, 2007; Grayson andMeltzer,2002; Koch and Barnosky, 2006; Firestone et al., 2007; Haynes,2007; Faith and Surovell, 2009). It has long been thought that is-land species are more susceptible to human-induced extinctionsthan those on continents and, consequently, the Channel Islandshave generatedmuch interest to proponents of overkill hypotheses.While early researchers reported that the Northern Channel Islandswere colonized by humans at least 40,000 years ago and claimed anassociation between human artifacts and mammoth remains aswell as what were interpreted as “fire features” (Berger and Orr,1966; Orr and Berger, 1966; Orr, 1968), closer scrutiny puts thedate of the first occupation of the islands by humans at around13,000 cal. BP), and what were interpreted as “mammoth roastingpits” are now thought to be natural burn features or due to othernatural processes (Wendorf, 1982; Cushing et al., 1986; Rick et al.,2012; Pigati et al., 2014).

While a possible minimum temporal overlap betweenmammoth and Homo sapiens of 250 years has been suggested forSanta Rosa Island using radiocarbon dating (Agenbroad et al.,2005), faunal remains associated with archeological sites datedbetween 12,200 and 11,500 cal. BP on the Channel Islands aredevoid of large mammals (Erlandson et al., 2007; Rick et al., 2012)and there is no direct evidence of human hunting of mammoths onthe Channel Islands (Rick et al., 2012).

Much is known about Channel Islandmammoths. It is likely thatColumbian mammoths gained access to the Northern ChannelIslands during a low stand of MIS 6, ca. 130 ka (Muhs et al., 2015)and gave rise to island pygmy mammoths via an approximately 50percent linear size reduction (Agenbroad, 2009). This decrease insize would provide the advantage of a lesser need for food intakewhich might have proven to be an advantage in times of environ-mental stress (McNab, 2010; Lister and Hall, 2014).

However, the survival ability of the pygmy mammoths wouldhave been compromised by the vast reduction of island land area assea levels rose at the end of the Pleistocene. The super-island ofSantarosae shrank to about 75% of its original landmass as sea levelsrose (Kennett et al., 2008) causing inevitable ecological distur-bances and alteration of available vegetation and freshwater sour-ces. In addition, the results of this study clearly show that M. exilismainly browsed on leaves and twigs, a clear shift in niche from itsmainland ancestor. As warming ensued, conifer forests werereplaced by sage scrub, Pinus stands and grasslands by about11,800 cal. BP (Heusser, 1995, 2000; Anderson et al., 2010). Ourmicrowear results are consistent with the last glacial (MIS4-2)vegetative history of Santa Rosa Island and suggest that climaticfactors contributed to the extinction of Channel Island PygmyMammoths on Santa Rosa Island as the shrinking of the foresthabitat would have impacted a species that relied heavily on leaves,and twigs.

Muhs et al. (2015) point out that since the earliest dated remainsofMammuthus from the California Channel islands are in late MIS 5(ca. 80ka), it is likely thatMammuthus arrived in a low-stand of MIS6 (or conceivably MIS 8). In this case, they survived the warmphases of MIS 5 where vegetation changes similar to those of the


MIS2/1 transition occurred (reduction in pine and other trees;expansion of open vegetation). Muhs et al. (2015) conclude thatclimate and vegetation change may not have been a majorcontributor to M. exilis extinction. It remains plausible, however,that habitat change, especially in the situation of a small islandarea, would have reduced population size, making the species morelikely to go extinct from these causes either stochastically or viaanother agent (Lister and Stuart, 2008). While our results suggestthat climatic factors may have played a role in the extinction ofChannel Island pygmy mammoths on Santa Rosa Island, thedetermination of the relative contribution of climatic factors versushuman overkill would benefit from futher study. These suggestionsrequire further work including on the diet of the early (M. columbi)immigrants in comparison with M. exilis.

5. Conclusions

This analysis has provided for the first time microwear evidenceof feeding behavior (i.e., mainly browsing on leaves and twigs) inM.exilis from Santa Rosa Island of California. This strongly suggeststhat a constrained niche was imposed upon M. exilis, compared tothe broader dietary breadth apparent in its mainland ancestor (M.columbi). Lastly, our analysis offers a glimpse into a possiblecontributing factor for the extinction of Channel Island pygmymammoths, as the woodland dietary preferences of this taxonwerebeginning to diminish and be replaced by grasslands as the climatewarmed.

Acknowledgements

We thank Paul Collins from the Santa Barbara Museum of Nat-ural History for access to the specimens and Jelena Hasjanova,Michaela Tarnowicz, Kristina Rogers, and Mia Russo from Bay PathUniversity for assistance in making casts.

References

Agenbroad, L.D., 1998. New pygmy mammoth (Mammuthus exilis) localities andradiocarbon dates from San Miguel, Santa Rosa, and Santa Cruz Islands, Cali-fornia. In: Weigand, P. (Ed.), Contributions to the Geology of the NorthernChannel Islands, Southern California. Pacific Section of the American Associa-tion of Petroleum Geologists, Bakersfield, California, pp. 169e175.

Agenbroad, L., 2003. New absolute dates and comparisons for California's Mam-muthus exilis. Deinsea 9, 1e16.

Agenbroad, L.D., 2001. Channel Islands (USA) pygmy mammoths (Mammuthusexilis) compared and contrasted with M. columbi, their continental ancestralstock. In: Cavarretta, G., Gioia, P., Mussi, M., Palombo, M.R. (Eds.), The World ofElephants: International Conference, Rome, pp. 473e475.

Agenbroad, L.D., 2009. Mammuthus exilis from the Channel Islands: height, mass,and geologic age. In: Damiani, C.C., Garcelon, D.K. (Eds.), Proceedings of theSeventh California Islands Symposium. Institute for Wildlife Studies, Arcata,pp. 15e19.

Agenbroad, L.D., 2012. Giants and pygmies: mammoths of Santa Rosa Island, Cali-fornia (USA). Quaternary International 255, 2e8.

Agenbroad, L.D., Johnson, J.R., Morris, D., Stafford, T.W., 2005. Mammoths andhumans as Late Pleistocene contemporaries on Santa Rosa island. In:Garcelon, D., Schwemm, C. (Eds.), Proceedings of the Sixth California IslandsSymposium (National Park Service Technical Publication CHIS-05-01. Institutefor Wildlife Studies, pp. 3e7.

Agenbroad, L.D., Johnson, J., Morris, D., Stafford, T.W., 2007. Mammoths and humansas late Pleistocene contemporaries on Santa Rosa Island. In: AmericanGeophysical Union Spring Meeting, pp. 3e7.

Albert, R.M., Weiner, S., 2001. Study of phytolith in prehistoric ash layers fromKebara and Tabun caves using a quantitative approach. In: Meunier, J.D.,Colin, F. (Eds.), Phytolith: Applications in Earth Sciences and Human History.A.A. Balkema Publishers, Lisse, pp. 251e256.

Anderson, R.L., Byrne, R., Dawson, T., 2008. Stable isotope evidence for a foggyclimate on Santa Cruz Island, California at̴ 16,600 cal. yr. B.P. Palaeogeography,Palaeoclimatology, Palaeoecology 262, 176e181.

Anderson, R.S., Starratt, S., Jass, R.M.B., Pinter, N., 2010. Fire and vegetation historyon Santa Rosa Island, Channel Islands, and long-term environmental change insouthern California. Journal of Quaternary Science 25, 782e797.


http://refhub.elsevier.com/S1040-6182(15)01402-0/sref1














































Archie, E.A., Morrison, T.A., Foley, A.H., 2006. Dominance rank relationships amongwild female African elephants, Loxodonta africana. Journal of Animal Behaviour71, 117e127.

Baker, G., Jones, L.H.P., Wardrop, I.D., 1959. Cause of wear in sheeps' teeth. Nature184, 1583e1584.

Bastl, K., Semprebon, G., Nagel, D., 2012. Low magnification microwear in Carnivoraand dietary diversity in Hyaenodon (Mammalia: Hyaenodontidae) with addi-tional information on its enamel microstructure. Palaeogeography, Palae-oclimatology, Palaeoecology 348e349, 13e20.

Berger, R., Orr, P., 1966. The fire areas of Santa Rosa Island, California II. Proceedingsof the National Academy of Sciences 56, 1678e1682.

Calandra, I., G€ohlich, U.B., Merceron, G., 2008. How could sympatric megaherbivorescoexist? Example of niche partitioning within a proboscidean community fromthe Miocene of Europe. Naturwissenschaften 95, 831e838.

Calandra, I., G€ohlich, U.B., Merceron, G., 2010. Feeding preferences of Gomphothe-rium subtapiroideum (Proboscidea, Mammalia) from the Miocene of Sandelz-hausen (Northern Alpine Foreland Basin, southern Germany) through life andgeological time: evidence from dental microwear analysis. Pal€aontologischeZeitschrift 84, 205e215.

Cheney, R.W., Mason, H.L., 1930. A Pleistocene Flora from Santa Cruz Island, Cali-fornia. Publication 415. Carnegie Institute of Washington, Washington, D.C.,pp. 1e24

Christensen, H.B., 2014. Similar associations of tooth microwear and morphologyindicate similar diet across marsupial and placental mammals. PLoS ONE 9 (8),e102789.

Clements, T., 1955. The Pleistocene History of the Channel Islands Region eSouthern California. In: Essays in the Natural Sciences in Honor of Capt.Alan Hancock, Univ. of Southern California Press, Los Angeles,pp. 311e323.

Cole, K.L., Liu, G., 1994. Holocene paleoecology of an estuary on Santa Rosa Island,California. Quaternary Research 41, 326e335.

Covert, H.H., Kay, R.F., 1981. Dental microwear and diet: implications for deter-mining the feeding behaviors of extinct primates, with a comment on the di-etary pattern of Sivapithecus. American Journal of Physical Anthropology 55,331e336.

Cushing, J., Wenner, A., Noble, E., Daily, M., 1986. A groundwater hypothesis for theorigins of ‘fire areas’ on the northern Channel Islands, California. QuaternaryResearch 23, 207e217.

DeSantis, L.R.G., Scott, J.R., Schubert, B.W., Donohue, S.L., McCray, B.M., VanStolk, C.A., Winburn, A.A., 2013. Direct comparisons of 2D and 3D dentalmicrowear proxies in extant herbivorous and carnivorous mammals. PLoS ONE8 (8), e71428.

Dudley, J., Mensah-Ntiamoah, A., Kpelle, D., 1992. Forest elephants in a rainforestfragment: preliminary findings from a wildlife conservation project in southernGhana. African Journal of Ecology 30, 116e126.

Eggert, L., Eggert, J., Woodruff, D., 2003. Estimating population sizes for elusiveanimals: the forest elephants of Kakum National Park, Ghana. Molecular Ecol-ogy 12, 1389e1402.

Erlandson, J.M., Graham, M.H., Bourque, B.J., Corbett, D., Estes, J.A., Steneck, R.S.,2007. The Kelp highway hypothesis: Marine ecology, the coastal migrationtheory, and the Peopling of the Americas. Journal of Island and CoastalArchaeology 2, 161e174.

Erlandson, J.M., Rick, T.C., Braje, T.J., Casperson, M., Culleton, B., Fulfrost, B., Garcia, T.,Guthrie, D.A., Jew, N., Kennett, D.J., Moss, M., Reeder, L., Skinner, C., Watts, J.,Willis, L., 2011. Paleocoastal seafaring, maritime technologies, and coastalforaging on California's Channel Islands. Science 331, 1181e1185.

Estes, R., 1999. African Mammal Data Base 1998. IUCN Publications Services,Cambridge.

Estridge, B.H., Reynolds, A.P., 2012. Basic Clinical Laboratory Techniques. DelmarCengage Learning, New York, p. 816.

Fahlke, J.M., Bastl, K., Semprebon, G.M., 2013. Paleoecology of archaeocete whalesthroughout the Eocene: dietary adaptations revealed by microwear analysis.Palaeogeography, Palaeoclimatology, Palaeoecology 386, 690e701.

Faith, J.T., Surovell, T., 2009. Synchronous extinction of North America's Pleisto-cene mammals. Proceedings of the National Academy of Sciences 106,20641e20645.

Fay, M., Agnagna, M., 1991. A population survey of forest elephants (Loxodontaafricana cyclotis) in northern Congo. African Journal of Ecology 29, 177e187.

Firestone, R.B., West, A., Kennett, J.P., Becker, L., Bunch, T.E., Revay, Z.S., Schultz, P.H.,Belgya, T., Kennett, D.J., Erlandson, J.M., Dickenson, O.J., Goodyear, A.A.,Harris, R.S., Howard, G.A., Kloosterman, J.P., Lechler, P., Mayewski, P.A.,Montgomery, J., Poreda, R., Darrah, T., Que Hee, S.S., Smith, A.R., Stich, A.,Topping, W., Wittke, J.H., Wolbach, W.S., 2007. Evidence for an extraterrestrialimpact 12,900 years ago that contributed to the megafaunal extinctions andYounger Dryas cooling. Proceedings of the National Academy of Sciences 104,16016e16021.

Fortelius, M., Solounias, N., 2000. Functional characterization of ungulate molarsusing the abrasion-attrition wear gradient: a new method for reconstructingpaleodiets. American Museum Novitates 3301, 1e36.

Foster, J.B., 1964. The evolution of mammals on islands. Nature 202, 234e235.Fraser, D., Theodor, J.M., 2011. Comparing ungulate dietary proxies using discrimi-

nant function analysis. Journal of Morphology 272, 1513e1526.Fraser, D., Theodor, J.M., 2013. Ungulate diets reveal patterns of grassland evolution

in North America. Palaeogeography, Palaeoclimatology, Palaeoecology 369,409e421.


Fraser, D., Mallon, J.C., Furr, R., Theodor, J.M., 2009. Improving the repeatability oflow magnification microwear methods using high dynamic range imaging.Palaios 24, 818e825.

Galbany, J., Martinez, L., Amor, H.L., 2005. Error rates in buccal-dental microwearquantification using scanning electron microscopy. Scanning 27, 23e29.

Grayson, D.K., 2001. The archaeological record of human impacts on animal pop-ulations. Journal of World Prehistory 15, 1e68.

Grayson, D.K., 2007. Deciphering North American Pleistocene extinctions. Journal ofAnthropological Research 63, 185e213.

Grayson, D.K., Meltzer, D.J., 2002. The human colonization of North America, Clovishunting, and large mammal extinction. Journal of World Prehistory 16,313e359.

Green, J.L., Semprebon, G.M., Solounias, N., 2005. Reconstructing the paleodiet ofFlorida Mammut americanum via low-magnification stereomicroscopy. Palae-ogeography, Palaeoclimatology, Palaeoecology 223, 34e48.

Grine, F.E., 1986. Dental evidence for dietary differences in Australopithecus andParanthropus: an analysis of permanent molar microwear. Journal of HumanEvolution 15, 783e822.

Grine, F., Ungar, P., Teaford, M., 2002. Error rates in dental microwear quantificationusing scanning electron microscopy. Scanning 24, 144e153.

Gügel, I.L., Grupe, G., Kunzelmann, K.-H., 2001. Simulation of dental microwear:characteristic traces by opal phytoliths give clues to ancient human dietarybehavior. American Journal of Physical Anthropology 114, 124e138.

Hammer, Ø., Harper, D.A.T., Ryan, P.D., 2001. PAST: Paleontological statistics soft-ware package for education and data analysis. Palaeontologica Electronica 4 (1),9 pp.

Haynes, G., 2007. A review of some attacks on the overkill hypothesis with specialattention to misrepresentations and doubletalk. Quaternary International169e70, 84e94.

Heusser, L.E., 1995. Pollen stratigraphy and paleoecologic interpretation of the 160k.y. record from Santa Barbara Basin, Hole 893A. In: Kennett, J.P., Baldauf, J.G.,Lyle, M. (Eds.), Proceedings of the Ocean Drilling Program, Scientific Results, vol.146, pp. 265e279 (2),.

Heusser, L.E., 2000. Rapid Oscillations in Western North America Vegetation andClimate during Oxygen Isotope Stage 5 Inferred from Pollen Data from SantaBarbara Basin (Hole 893A).

Hoffman, J.M., Fraser, D., Clementz, M.T., 2015. Controlled feeding trials with un-gulates: a new application of in vivo dental molding to assess the abrasivefactors of microwear. The Journal of Experimental Biology 218, 1538e1547.

Johnson, D., 1972. Landscape Evolution on San Miguel Island, California. Ph.D.dissertation. University of Kansas, Lawrence, 378 pp.

Junak, S., Knapp, D.A., Haller, F.R., Philbrick, R., Shoenherr, A., Keeler-Wolf, T., 2007.The California Channel Islands. In: Barbour, M.G., Keeler-Wolf, T.,Schoenherr, A.A. (Eds.), Terrestrial Vegetation of California. University of Cali-fornia Press, Berkeley, CA, pp. 229e252.

Johnson, D., 1978. The origin of island mammoths and the quaternary land bridgehistory of the northern Channel Islands, California. Quaternary Research 10,204e225.

Joshi, R., Singh, R., 2008. Feeding behaviour of wild Asian elephants (Elephasmaximus) in the Rajaji National Park. The Journal of American Science 4, 34e48.

Kay, R.F., Covert, H.H., 1983. True grit: a microwear experiment. American Journal ofPhysical Anthropology 61, 33e38.

Kennett, D.J., Kennett, J.Pl, West, G.J., Erlandson, J.M., Johnson, J.R., Hendy, I.L.,West, A., Culleton, B.J., Jones, T.L., Stafford, T.W., 2008. Wildfire and abruptecosystem disruption on California's northern Channel Islands at the Ållerød-Younger Dryas Boundary (13.-12.9ka). Quaternary Science Reviews 27,2530e2545.

King, T., Andrews, P., Boz, B., 1999. Effect of taphonomic processes on dentalmicrowear. American Journal of Physical Anthropology 108, 359e373.

Koch, P.L., Barnosky, A., 2006. Late Quaternary extinctions: state of the debate.Annual Review of Ecology, Evolution, and Systematics 37, 350e360.

Lister, A.M., Hall, C., 2014. Variation in body and tooth size with island area in smallmammals: a study of Scottish and Faroese house mice (Mus musculus). AnnalesZoologici Fennici 51, 95e110.

Lister, A.M., Stuart, A.J., 2008. The impact of climate change on large mammaldistribution and extinction: evidence from the last glacial/interglacial transi-tion. Comptes Rendus Geoscience 340, 615e620.

Lucas, P.W., Omar, R., Al-Fadhalah, K., Almusallam, A.S., Henry, A.G., Michael, S.,Thai, L.A., Watzke, J., Strait, D.S., Atkins, A.G., 2013. Mechanisms and causes ofwear in tooth enamel: implications for hominin diets. Journal of the RoyalSociety Interface 10, 20120923.

Maas, M.C., 1991. Enamel structure and microwear: an experimental study of theresponse of enamel to shearing force. American Journal of Physical Anthro-pology 85, 31e49.

Maas, M.C., 1994. A scanning electron-microscopic study of in vitro abrasion ofmammalian tooth enamel under compressive loads. Archives of Oral Biology 39,1e11.

Madden, C.T., 1977. Elephants of the Santa Barbara Channel Islands, Southern Cal-ifornia. Geological Society of America (Cordilleran Section) Abstracts withPrograms, pp. 458e459.

Madden, C.T., 1981. Origin(s) of mammoths from northern Channel Islands, Cali-fornia. Quaternary Research 15, 101e104.

Mainland, I.L., 2003. Dental microwear in grazing and browsing Gotland sheep (Ovisaries) and its implications for dietary reconstruction. Journal of AchaeologicalScience 30, 1513e1527.











































http://refhub.elsevier.com/S1040-6182(15)01402-0/sref133a


























































































































































































Martin, P.L., Klein, R.G., 1984. Quaternary Extinctions: a Prehistoric Revolution.University of Arizona Press, Tucson, Arizona.

McNab, B.K., 2010. The geographic and temporal correlations of mammalian sizereconsidered: a resource rule. Oecologia 164, 13e23.

Merceron, G., Blondel, C., Brunet, M., Sen, S., Solounias, N., Viriot, L., Heintz, E., 2004.The Late Miocene paleoenvironoment of Afghanistan as inferred from dentalmicrowear in artiodactyls. Palaeogeography, Palaeoclimatology, Palaeoecology207, 143e163.

Merceron, G., Blondel, C., Bonis, L., Koufos, G.D., Viriot, L., 2005. A new method ofdental microwear: application to extant primates and Ouranopithecus mace-doniensis (Late Miocene of Greece). Palaios 20, 551e561.

Merz, G., 1986a. Counting elephants (Loxodonta africana cyclotis) in tropical rainforests with particular reference to the Tai National Park, Ivory Coast. AfricanJournal of Ecology 24, 61e68.

Merz, G., 1986b. Movement patterns and group size of the African forest elephantLoxodonta africana cyclotis in the Tai National Park, Ivory Coast. African Journalof Ecology 24, 133e136.

Mihlbachler, M.C., 1998. Late-Pleistocene mastodon and digesta from Little RiverNorth Florida. Current Research in the Pleistocene 15, 116e118.

Mihlbachler, M.C., Beatty, B.L., Caldera-Siu, A., Chan, D., Lee, R., 2012. Error rates andobserver bias in dental microwear analysis using light microscopy. Palae-ontolgia Electronica 15, 1e22.

Miller, W.E., 1971. Pleistocene vertebrates of the Los Angeles Basin and vicinity(exclusive of Rancho La Brea). Bulletin of the Los Angeles County Museum ofNatural History 10, 1e224.

Miller, W.E., 1976. Late Pleistocene vertebrates of the Silver Creek local fauna fromnorth- central Utah. Great Basin Naturalist 36, 387e424.

Mohaptra, K.K., Patra, A.K., Paramanik, D.S., 2013. Food and feeding behavior ofAsiatic elephant (Elephas maximus Linn.) in Kuldiha Wild Life Santuary, Odisha,India. Journal of Environmental Biology 34, 87e92.

Muhs, D.R., Simmons, K.R., Groves, L.T., McGeehin, J.P., Schumann, R.R.,Agenbroad, L.D., 2015. Late Quaternary sea-level history and the antiquity ofmammoths (Mammuthus exilis and Mammuthus columbi), Channel Islands Na-tional Park, California, USA. Quaternary Research 83, 502e521.

Müller, J., Clauss, M., Codron, D., Schulz, E., Hummel, J., Fortelius, M., Kircher, P.,Hatt, J.-M., 2014. Growth and wear of incisor and cheek teeth in domesticrabbits (Oryctolagus cuniculus) fed diets of different abrasiveness. Journal ofExperimental Zoology Part A: Ecological Genetics and Physiology 321, 283e298.

Nowak, R.M., 1999. Walkers Mammals of the World. The Johns Hopkins UniversityPress, Baltimore.

Orr, P.C., 1956a. Radiocarbon mammoths, and man on Santa Rosa Island. GeologicalSociety of America Bulletin 67, 1777.

Orr, P.C., 1956b. Radiocarbon dates from Santa Rosa Island. Bulletin of the SantaBarbara Museum of Natural History 2, 1e10.

Orr, P.C., 1968. Prehistory of Santa Rosa Island. Santa Barbara Museum of NaturalHistory, Santa Barbara.

Orr, P.C., Berger, R., 1966. The fire areas of Santa Rosa Island, California. Proceedingsof the National Academy of Sciences 56, 1409e1416.

Osborn, F.V., 2005. Habitat selection by bull elephants in central Zimbabwe.Pachyderm 39, 63e66.

Palombo, M.R., 2005. Coupling tooth microwear and stable isotope analyses forpalaeodiet reconstruction: the case study of Late Middle Pleistocene Elephas(Palaeoloxodon) antiquus teeth from Central Italy (Rome area). Quaternary In-ternational 126e128, 153e170.

Palombo, M.R., 2008. Insularity and its Effects. Quaternary International 182, 1e183.Palombo, M.R., Curiel, V., 2003. Tooth Microwear Analysis of Mammuthus (Archi-

diskodon) meridionalis gromovi (Alexeeva and Garutt, 1965) from Montopoli(Lower Valdarno, Tuscany, Italy): a Methodological Approach. Bollettino dellaSociet�a Paleontologica Italiana, pp. 151e155.

Perez-Crespo, V.A., Arroyo-Cabrales, J., Benammi, M., Johnson, E., Polaco, O.J., San-tos- Moreno, A., Morales-Puente, P., Cienfuegos-Alvarado, E., 2012. Geographicvariation of diet and habitat of the Mexican populations of ColumbianMammoth (Mammuthus columbi). Quaternary International 276e277, 8e16.

Pigati, J.S., Mcgeehin, J.P., Skipp, G.L., Muhs, D.R., 2014. Evidence of repeated wild-fires prior to human occupation on San Nicolas Island, California. Monographsof the Western North American Naturalist 7, 35e47.

Rabenold, D., Pearson, O.M., 2014. Scratching the surface: a critique of Lucas et al.(2013)'s conclusion that phytoliths do not abrade enamel. Journal of HumanEvolution 74, 130e133.

Remington, C.L., 1971. Natural history and evolutionary genetics of the CaliforniaIslands. Discovery 7, 2e18.

Rensberger, J.M., 1978. Scanning electron microscopy of wear and occlusal events insome small herbivores. In: Butler, P.M., Joysey, K.A. (Eds.), Development,Function and Evolution of Teeth. Academic Press, London, pp. 415e438.

Rick, T.C., Hofman, C.A., Braje, T.J., Maldonado, J.E., Sillett, T.S., Danchisko, K.,Erlandson, J.M., 2012. Flightless ducks, giant mice and pygmy mammoths: LateQuaternary extinctions on California's Channel Islands. World Archaeology 44,3e20.

Rivals, F., Solounias, N., Mihlbachler, M.C., 2007. Evidence for geographic variation inthe diets of Late Pleistocene and early Holocene Bison in North America, anddifferences from the diets of recent Bison. Quaternary Research 68, 338e346.

Rivals, F., Semprebon, G.M., 2011. Dietary plasticity in ungulates: insight from toothmicrowear analysis. Quaternary International 245, 279e284.

Rivals, F., Semprebon, G.M., Lister, A., 2012. An examination of dietary diversitypatterns in Pleistocene proboscideans (Mammuthus, Palaeoloxodon, and


Mammut) from Europe and North America as revealed by dental microwear.Quaternary International 255, 188e195.

Rivals, F., Mol, D., Lacombat, F., Lister, A.M., Semprebon, G.M., 2015. Resource par-titioning and niche separation between mammoths (Mammuthus rumanus andMammuthus meridionalis) and gomphotheres (Anancus arvernensis) in the EarlyPleistocene of Europe. Quaternary International 379, 164e170.

Rode, K.D., Chiyo, P.I., Chapman, C.A., McDowell, L.R., 2006. Nutritional ecology ofelephants in Kibale National Park, Uganda and its relationship with crop raidingbehavior. Journal of Topical Ecology 22, 441e449.

Roth, V.L., 1982. Dwarf mammoths from the Santa Barbara Channel Islands: Size,Shape, Development, and Evolution. Ph.D. dissertation. Yale University, NewHaven, CT.

Roth, V.L., 1996. Pleistocene dwarf elephants from the California Islands. In:Shoshani, J.H., Tassy, P. (Eds.), The Proboscidea. Oxford University Press, Oxford,pp. 249e253.

Sangster, A.G., Hodson, M.J., 1992. Deposition in Subterranean organs. In: Rapp, G.,Mulholland (Eds.), Phytolith Systematics: Emerging Issues. Plenum Press, NewYork, pp. 239e251.

Sanson, G.D., Kerr, S.A., Gross, K.A., 2007. Do silica phytoliths really wearmammalian teeth? Journal of Archaeological Science 34, 526e531.

Savage, J.M., 1967. Evolution of the insular herpetofaunas. In: Philbrick, R.N. (Ed.),Proceedings of the Symposium of the Biology of the California Islands. SantaBarbara Botanic Garden, Santa Barbara, pp. 219e227.

Scott, R.S., Ungar, P.S., Bergstrom, T.S., Brown, C.A., Grine, F.E., Teaford, M.F.,Walker, A., 2005. Dental microwear texture analysis shows within-species dietvariability in fossil hominins. Nature 436, 693e695.

Scott, R.S., Ungar, P.S., Bergstrom, T.S., Brown, C.A., Childs, B.E., Teaford, M.F.,Walker, A., 2006. Dental microwear texture analysis; technical considerations.Journal of Human Evolution 51, 339e349.

Schulz, E., Piotrowski, V., Clauss, M., Mau, M., Merceron, G., Kaiser, T.M., 2013. Di-etary abrasiveness is associated with variability of microwear and dental sur-face texture in rabbits. PLoS One 8, e56167.

Semprebon, G.M., 2002. Advances in the Reconstruction of Extant Ungulate Eco-morphology with Applications to Fossil Ungulates. Ph.D. Dissertation. Univer-sity of Massachusetts, Amherst.

Semprebon, G.M., Godfrey, L.R., Solounias, N., Sutherland, M.R., Jungers, W.L., 2004.Can low-magnification stereomicroscopy reveal diet? Journal of Human Evo-lution 47, 115e144.

Semprebon, G.M., Rivals, F., 2007. Was grass more prevalent in the pronghorn past?An assessment of the dietary adaptations of Miocene to recent Antilocapridae(Mammalia: Artiodactyla). Palaeogeography, Palaeoclimatology, Palaeoecology253, 332e347.

Shoshani, J., Eisenberg, J.F., 1982. Elephas maximus. Mammalian Species 182,1e8.

Slaughter, B.H., Crook, W.W., Harris, R.K., Allen, D.C., Seifert, M., 1962. The Hull-Schuller Local Faunas of the Upper Trinity River, Dallas and DentonCounties, Texas. Texas Bureau of Economic Geology, pp. 1e75. Report of In-vestigations 48.

Solounias, N., Semprebon, G., 2002. Advances in the reconstruction of ungulateecomorphology with application to early fossil equids. American MuseumNovitates 366, 1e49.

Steadman, D.W., Martin, P.S., 2003. The late Quaternary extinction and futureresurrection of birds on Pacific islands. Earth Science Reviews 61, 133e147.

Stearns, R.E.C., 1873. [no title: paraphrase of comments attributed to Stearns inminutes of Academy's regular meeting]. Proceedings of the California Academyof Sciences 5, 152.

Sukumar, R., 2003. The Living Elephants: Evolutionary Ecology, Behavior andConservation. Oxford University Press, New York.

Tangley, L., 1997. In search of Africa's Forgotten Forest elephant. Science 275,1417e1419.

Todd, N.E., Falco, N., Silva, N., Sanchez, C., 2007. Dental microwear variation incomplete molars of Loxodonta africana and Elephas maximus. Quaternary In-ternational 169e170, 192e202.

Townsend, K.E.B., Croft, D.A., 2008. Diets of notoungulates from the Santa CruzFormation, Argentina: new evidence from enamel microwear. Journal ofVertebrate Paleontology 28, 217e230.

Ungar, P.S., Brown, C.A., Berstrom, T.S., Alan, W., 2003. Quantification of dentalmicrowear by tandem scanning confocal microscopy and scale-sensitive fractalanalysis. Scanning 25, 185e193.

Ungar, P.S., Scott, J.R., Schubert, B.W., Stynder, D.D., 2010. Carnivoran dentalmicrowear textures: comparability of carnassial facets and functional differ-entiation of postcanine teeth. Mammalia 74, 219e224.

Ungar, P.S., Scott, R.S., Scott, J.R., Teaford, M., 2008. Dental microwear analysis:historical perspectives and new approaches. In: Irish, J.D., Nelson, G.C. (Eds.),Technique and Application in Dental Anthropology. Cambridge University Press,Cambridge, pp. 389e425.

Valentine, J.W., Lipps, J.H., 1967. Late Cenozoic history of the southern CaliforniaIslands. In: Philbrick, R.N. (Ed.), Proceedings of the Symposium of the Biology ofthe California Islands. Santa Barbara Botanic Garden, Santa Barbara, California,pp. 21e35.

Van Gelder, R.G., 1965. Channel Island skunk. Natural History 74, 30e35.von Bloeker, J.R., 1967. The land mammals of the southern California Islands. In:

Philbrick, R.N. (Ed.), Proceedings of the Symposium of the Biology of the Cali-fornia Islands. Santa Barbara Botanic Garden, Santa Barbara, California,pp. 245e263.



































































































































































































































Walker, A., Hoek, H.N., Perez, L., 1978. Microwear of mammalian teeth as an indi-cator of diet. Science 201, 908e910.

Webb, S.D., Dunbar, J., Newsom, L., 1992. Mastodon digesta from North Florida.Current Research in the Pleistocene 9, 114e116.

Wendorf, M., 1982. The fire areas of Santa Rosa Island: an interpretation. NorthAmerican Archaelologist 3, 173e180.

West, G.J., Erlandson, J.M., 1994. A Late Pleistocene Pollen Record from San MiguelIsland, California: Preliminary Results. AMQUA Abstracts. University of Min-nesota, p. 256.

White, L., Tutin, C., Fernandez, M., 1993. Group composition and diet of forest el-ephants, Loxodonta africana cyclotis Matschie 1900, in the Lope Reserve, Gabon.African Journal of Ecology 31, 181e199.

Whittaker, R.J., 1998. Island Biogeography: Ecology, Evolution, and Conservation.Oxford University Press, Oxford, pp. 73e75.

Williams, F.L., Geissler, E., 2014. Reconstructing the diet and palecology of Plio-Pleistocene Cercopithecoides williamsi from Sterkfontein, South Africa. Palaios29, 483e494.


Williams, F.L., Holmes, N.A., 2011. Evidence of terrestrial diets in Pliocene Eurasianpapionins (Mammalia: Primates) inferred from low-magnification stereo-microscopy of molar enamel use-wear scars. Palaios 26, 720e729.

Williams, F.L., Patterson, J.W., 2010. Reconstructing the paleoecology of Taung,South Africa from low-magnification of dental microwear features in fossilprimates. Palaios 25, 439e448.

Wittemyer, G., Getz, W.M., Vollrath, F., Hamilton, I.D., 2007. Social dominance,seasonal movements, and spatial segregation in African elephants: a contri-bution to conservation behavior. Behavioural Ecology and Sociobiology 61,1919e1931.

Wroe, S., Field, J., Grayson, D.K., 2006. Megafaunal extinction: climate, humans, andassumptions. Trends in Ecology and Evolution 21, 61e62.

Xia, J., Zheng, J., Diaodiao, H., Tian, Z.R., Chen, L., Zhongrong, Z., Ungar, P.S.,Linmao, Q., 2015. New model to explain tooth wear with implications formicrowear formation and diet reconstruction. Proceedings, National Academyof Science 112 (34), 10669e10672.














































Dietary reconstruction of pygmy mammoths from Santa Rosa ... · Dietary reconstruction of pygmy...

Documents

Transcript of Dietary reconstruction of pygmy mammoths from Santa Rosa ... · Dietary reconstruction of pygmy...