Error rates and observer bias in dental microwear analysis using light microscopy

Abstract

Despite the increased use of light microscopy in microwear analysis, studies that recognize observer error are scarce. Nonetheless, microwear analysis based on light microscopy may be more prone to observer bias than SEM or confocal microscopy. We measured observer error among five observers, who independently analyzed identical sets of dental wear surfaces on digital micrographs taken through a light microscope. For experienced microscopists, error in light-microscope-based microwear was of a similar magnitude to error rates for SEM-based microwear methods. Significant intraobserver error was rare among the most experienced observers and higher among inexperienced observers. However, because observers develop familiarity when repeatedly analyzing the same image, intraobserver error measured here and in other similar SEM studies may be artificially low. Interobserver error was highly significant and did not appear to diminish with experience. Nonetheless, the data collected by all observers was highly correlated. Essentially all observers found similar microwear differences between the species analyzed, even though the absolute values in the data were observer dependent. Going forward, microwear results will be more robust if observers adopt methods that ensure observer blindness, and avoid the common practice of combining data from different observers, and even from one observer when the observarions are separated by time and experience. Rather than using pre-published microwear databases of extant species as a basis for interpreting paleodiet, researchers may be better served by building shared microwear image libraries, with which observers could generate their own data for the basis of making paleodietary inferences.

Feeding Ecology in Oligocene Mylodontoid Sloths (Mammalia, Xenarthra) as Revealed by Orthodentine Microwear Analysis

“…For these reasons, our goal was not to reproduce absolute feature counts between the two methods, as such is neither possible nor necessary. In user-based microwear methods, relative differences in quantitative variables among taxa is more informative than relying on absolute benchmarks to define dietary differences (Mihlbachler et al 2012). Thus, below we interpret the results from the two techniques separately (in the context of relative differences to other sloths within each technique) and then combine our separate interpretations to hypothesize the feeding ecology of Octodontotherium and Orophodon .…”

A large carnivorous mammal from the Late Cretaceous and the North American origin of marsupials

“…By quantifying the microwear of species with known diets, a comparative data set can be built and then used to estimate the diet of taxa for which food intake is unknown, including extinct species. Although multiple studies have quantified the microwear of numerous extant mammals whose diet is known, analyses of inter-observer error demonstrate that each investigation of the diet of fossil taxa should develop its own comparative data set74.…”

Direct Comparisons of 2D and 3D Dental Microwear Proxies in Extant Herbivorous and Carnivorous Mammals

“…As dental microwear records food consumption during the last few days to weeks of an animal’s life, it can be used to clarify ancient diets and assess dietary responses to changing climates and environments. While microwear has been commonly used by anthropologists and paleontologists since the late 1970s (e.g., [1–3]), the methodologies used to quantify tooth surfaces are highly variable and still debated among researchers, and results generated between and within methods are not directly comparable [4–7]. …”

“…Observer identification and quantification of individual wear features, whether from an SEM micrograph or using light microscopy, is prone to high observer biases, particularly between observers of different experience levels [4,5,7,16]. Grine and co-authors [4] found a 9% error in measurements between observers while Galbany and co-authors [5] found a 6% error in observers with five or more years of experience when quantifying SEM micrographs.…”

“…Grine and co-authors [4] found a 9% error in measurements between observers while Galbany and co-authors [5] found a 6% error in observers with five or more years of experience when quantifying SEM micrographs. Most recently, Mihlbachler and colleagues [7] documented a 45% interobserver error among experienced and inexperienced individuals, which was reduced to 8-12% in experienced individuals after multiple iterations. While experienced observers yielded similarly shaped trophic triangles consistent with prior work [11], inexperienced microwear observers failed to generate a trophic triangle using light microscopy [7].…”