Dental microwear texture analysis of extant koalas: clarifying causal agents of microwear

Abstract

Microscopic wear patterns on teeth, that is, dental microwear, are capable of recording observed dietary behaviour in a diversity of extant and extinct animals. However, recent work has questioned the utility of dental microwear at clarifying dietary behaviour, instead suggesting that dental microwear textures are reflective of grit consumed and not the dietary properties of ingested food. Some suggest that dental microwear cannot reflect the textural properties of vegetation consumed because phytoliths are too soft to form microwear. Koalas (Phascolarctos cinereus) are model organisms for examining dental microwear formation because they consume a specialized diet consisting almost exclusively of eucalyptus leaves, which notably lack phytoliths. Here, we assess if koala dental microwear records a diet consistent with the consumption of tough leaves – despite the absence of phytoliths in their primary food source. Dental microwear texture data of koalas are consistent with tough folivorous diets, with high anisotropy values indistinguishable from folivorous primates and grazing bovids. However, koalas have significantly higher complexity (indicative of hard object feeding) than the folivorous primate Alouatta palliata, and are indistinguishable in this variable from mixed feeding and browsing bovids. While higher complexity values in koalas may result from increased dust and/or grit on the landscape in comparison to folivores that occur in wetter environments, it may also result from the mastication of woody browse. We also determined that complexity and textural fill volume are not significantly greater in drier environments – as one would predict if grit was the major contributor to dental microwear formation. Koalas may instead be selecting softer younger leaves in drier regions where water is more limited. Collectively, microwear associated with tough object feeding can be formed in the absence of phytoliths and grit may interact with food to form microwear reflective of diet, as opposed to overprinting dietary signals.

Phylogenetic signal in tooth wear dietary niche proxies

“…For these reasons, tooth wear is often considered to be taxon or phylogeny free, as it is thought to reflect feeding behavior under the implicit assumption that this behavior is highly labile (Blomberg, Garland, & Ives, ). Despite the assumption of lability, one of the most common applications of analytical tooth wear methods is for the inference of average species diets (Fortelius & Solounias, ; Fraser & Theodor, ; Semprebon et al., ; Solounias & Semprebon, ), an approach that has been applied with considerable success (Donohue et al., ; Fortelius & Solounias, ; Fraser & Theodor, ; Haupt et al., ; Hedberg & DeSantis, ; Semprebon et al., ; Solounias & Semprebon, ). We show herein that mammalian diets, summarized at the species‐level, show high phylogenetic signal (Table ; Figure a–b) and that mammals may show some degree of trophic PNC (although it is outside the scope of this paper to definitively test for PNC in mammal diets; Münkemüller, Boucher, Thuiller, & Lavergne, ).…”

“…If tooth wear proxies are reliable reflections of average species’ diets, as many studies suggest they are (Donohue et al., ; Fortelius & Solounias, ; Fraser & Theodor, ; Haupt et al., ; Hedberg & DeSantis, ; Semprebon et al., ; Solounias & Semprebon, ), then logic would dictate that they must show a similar degree of phylogenetic signal. We demonstrate that all of the analytical tooth wear data analyzed herein show strong phylogenetic signal (Table ; Figure ); mean H 2 and Pagel's λ are consistently high for mesowear and microwear.…”

“…For mammals, the two most commonly employed sets of proxies are tooth macrowear (also referred to as mesowear) and microwear (microscopic tooth wear). Analytical tooth wear methods such as mesowear and microwear involve quantification of macroscopic and microscopic features on the occlusal surfaces of teeth in an attempt to characterize the diets of individuals and, in turn, the Eltonian niches of entire species (DeSantis & Haupt, ; DeSantis et al., ; Donohue, DeSantis, Schubert, & Ungar, ; Fortelius & Solounias, ; Fraser, Mallon, Furr, & Theodor, ; Fraser & Theodor, , , ; Fraser, Zybutz, Lightner, & Theodor, ; Grine & Kay, ; Haupt, DeSantis, Green, & Ungar, ; Hedberg & DeSantis, ; Semprebon, Godfrey, Solounias, Sutherland, & Jungers, ; Solounias, Moelleken, & Plavcan, ). Tooth wear data are then typically compared to observed dietary data (e.g., gut contents, fecal contents, personal observations) in a reference population or sample of species (i.e., the training dataset) to create a set of regression coefficients or discriminant functions that can be used to infer the diets of species and individuals for which independent dietary data are unavailable (Barr & Scott, ; Fraser & Theodor, ).…”

“…Tooth wear data are then typically compared to observed dietary data (e.g., gut contents, fecal contents, personal observations) in a reference population or sample of species (i.e., the training dataset) to create a set of regression coefficients or discriminant functions that can be used to infer the diets of species and individuals for which independent dietary data are unavailable (Barr & Scott, ; Fraser & Theodor, ). Many studies have found analytical tooth wear methods to be statistically powerful means of dietary inference (Donohue et al., ; Fortelius & Solounias, ; Fraser & Theodor, ; Haupt et al., ; Hedberg & DeSantis, ; Semprebon et al., ; Solounias & Semprebon, ); rates of correct classification of species with known diets are often fairly high (~80% or higher, depending on the study) (Fraser & Theodor, ), allowing dietary inference from tooth wear with a reasonable degree of confidence.…”

“…The most widely employed methods for dietary inference include mesowear, the visual scoring of macroscopic tooth wear (Fortelius & Solounias, ; Fraser et al., ; Kaiser & Fortelius, ; Kaiser & Solounias, ), and microwear, the quantification of microscopic marks on the chewing surfaces of the teeth at either low or high magnification (Grine & Kay, ; Merceron, Blondel, Bonis, Koufos, & Viriot, ; Semprebon et al., ; Solounias & Semprebon, ; Ungar et al., ). Both mesowear and microwear have proven to be extremely useful tools for inferring the dietary preferences of extant and extinct mammals (and nonmammals) from numerous families (DeSantis & Haupt, ; DeSantis et al., ; Donohue et al., ; Fortelius & Solounias, ; Fraser & Theodor, , , ; Fraser et al., , ; Grine & Kay, ; Haupt et al., ; Hedberg & DeSantis, ; Semprebon et al., ; Solounias et al., ).…”