Dental complexity and diet in amniotes: A meta-analysis

Abstract

Tooth morphology is among the most well-studied indicators of ecology. For decades, researchers have examined the gross morphology and wear patterns of teeth as indicators of diet, and recent advances in scanning and computer analysis have allowed the development of new and more quantitative measures of tooth morphology. One of the most popular of these new methods is orientation patch count (OPC). OPC, a measure of surface complexity, was originally developed to distinguish the more complex tooth crowns of herbivores from the less complex tooth crowns of faunivores. OPC and a similar method derived from it, orientation patch count rotated (OPCR), have become commonplace in analyses of both modern and fossil amniote dietary ecology. The widespread use of these techniques makes it possible to now re-assess the utility of OPC and OPCR. Here, we undertake a comprehensive review of OPC(R) and diet and perform a meta-analysis to determine the overall difference in complexity between herbivores and faunivores. We find that the relationship between faunivore and herbivore OPC or OPCR values differs substantially across studies, and although some support the initial assessment of greater complexity in herbivores, others do not. Our meta-analysis does not support an overall pattern of greater complexity in herbivores than faunivores across terrestrial amniotes. It appears that the relationship of OPC or OPCR to diet is taxon-specific and dependent on the type of faunivory of the group in question, with insectivores often having values similar to herbivores. We suggest extreme caution in comparing OPC and OPCR values across studies and offer suggestions for how OPCR can constructively be used in future research.

Fig 1. PRISMA diagram showing paper screening and selection process conducted in November 2021.

Three additional studies were added to the analysis in March 2023 using a similar process, filtering for publication since 2021.

Fig 2. Dendrogram showing similarity among sample groups included in meta-analysis, measured in mean nearest taxon distance.

Scale is in units of millions of years, indicating the average number of years of divergence for the taxa in one sample group to the mostly closely-related taxa in other sample groups.

Fig 3. Funnel plots showing the distribution of SMD compared to standard error.

SMD is the standardized mean difference. Open circles indicate studies conducted on a single tooth. Filled circles indicate studies conducted on multiple teeth. Note that standard error is shown with 0 at the top of the y axis. (A) Raw SMD values before sample similarity correction. Shaded region represents a 99.5% pseudo-confidence interval based on standard error. (B) Adjusted SMD values after sample similarity correction.

Fig 4. Forest diagrams showing effect sizes for each study sample used in meta-analyses, before sample similarity correction.

Standardized mean difference is calculated between the mean herbivore and mean faunivore OPC(R) values. Studies of multiple teeth are shown in red, and studies of single teeth are shown in black. Size of dot indicates relative study weight. Error bars show 95% confidence intervals. Samples are ordered by increasing precision.

Fig 5. Forest diagrams showing effect sizes for each study sample used in meta-analyses, after sample similarity correction.

Standardized mean difference is calculated between the mean herbivore and mean faunivore OPC(R) values, and subsequently adjusted using the mean nearest taxon distance among samples. Studies of multiple teeth are shown in red, and studies of single teeth are shown in black. Size of dot indicates relative study weight. Error bars show 95% confidence intervals. Samples are ordered by increasing precision.

Fig 6. Log-transformed mean OPC(R) values for herbivores and faunivores.

Each point represents a comparison of an herbivore sample to a faunivore sample within a single study, with open and filled indicating whether OPC(R) values were calculated for a single tooth or a tooth row. Dotted line shows 1:1 ratio, where points would lie if herbivores and faunivores had equal OPC(R).