RESUMO
Dental microwear studies often analyze casts rather than original surfaces, although the information loss associated with reproduction is rarely considered. To investigate the sensitivity of high magnification (150x) microwear analysis to common surface replication materials and methods, we compared areal surface texture parameters (ISO 25178-2) and traditional microwear variables (pits and scratches) generated from teeth and casts of rat molars exposed to experimental diets involving hard and soft foods in which abrasive materials had been added. Although the data from the original and replicated surfaces were correlated, many significant differences were found between the resulting data of the casts and original teeth. Both areal surface texture parameters and traditional microwear variables showed diminished ability to discriminate between the eight diet treatments when casts were analyzed. When areal surface texture parameters and traditional microwear variables were combined into a single discriminant function analysis, the cast data and original data produced the most similar results. Microwear researchers tend to favor either texture analysis or traditional microwear methods, better results may be generated by combining them. Although surface textures were not accurately reproduced by the casts, they retained sufficient information to discriminate between microwear of the experimental diets to a degree similar to the original teeth.