Surface texture analysis in Toothfrax and MountainsMap® SSFA module: Different software packages, different results?

Abstract

The scale-sensitive fractal analysis (SSFA) of dental microwear textures is traditionally performed using the software Toothfrax. SSFA has been recently integrated to the software MountainsMap® as an optional module. Meanwhile, Toothfrax support has ended. Before switching to the new module, the outputs between the two software packages must be compared for consistency. We have performed such a test using Bayesian modelling on three datasets including dental surfaces of sheep (Merceron, Ramdarshan, et al., 2016) and guinea pigs (Winkler, Schulz-Kornas, Kaiser, Cuyper, et al., 2019) from controlled feeding experiments, as well as surfaces of quartzite and flint flakes used in an actualistic archeological experiment on cleaning procedures (Pedergnana, Calandra, Bob, et al., 2020). We found that the two software packages calculate significantly different values for the SSFA parameters epLsar, Asfc, HAsfc9 and R 2 , even when the same settings are used. Nevertheless, the treatments (different diets or cleaning procedures) are discriminated similarly within each dataset. While the new software module is as good as the original software to differentiate treatments, our results imply that the outputs from the two software packages are not directly comparable and, as such, cannot be merged. Surface texture analysts should therefore consider re-analyzing published surfaces before integrating them in their studies.

Life in a Central European warm-temperate to subtropical open forest: paleoecology of the rhinocerotids from Ulm-Westtangente (Aquitanian, early Miocene, Germany)

“…The consumption of leaves, that are abrasive but soft items, could reconciliate both, although anisotropy is normally high in folivore (Scott, 2012; Hullot et al, 2019). This last observation could however be biased by the previous use of epLsar, which had an error in the code for its extraction (Calandra et al, 2022).…”

“…We conducted scale-sensitive fractal analyses on the selected surfaces (200 × 200 μm; see Supplementary 3) using MountainsMaps® (v.8.2). Our study will focus on the following texture variables, described in detail in Scott et al (2006): anisotropy or exact proportion of length-scale anisotropy of relief (epLsar) is a measure of the orientation concentration of surface roughness; in MountainsMaps®, this parameter has been corrected (NewepLsar), as there was an error in the code of Toothfrax (software previously used for DMTA but not supported anymore; Calandra et al, 2022); complexity or area-scale fractal complexity (Asfc) assesses the roughness at a given scale; heterogeneity of the complexity (Hasfc) gives information of the variation of complexity at a given scale (here 3 × 3 and 9 × 9) within the studied 200 × 200 μm zone. …”

“…• anisotropy or exact proportion of length-scale anisotropy of relief (epLsar) is a measure of the orientation concentration of surface roughness; in MountainsMaps®, this parameter has been corrected (NewepLsar), as there was an error in the code of Toothfrax (software previously used for DMTA but not supported anymore; Calandra et al, 2022);…”

Evolutionary paleoecology of European rhinocerotids across the Oligocene-Miocene transition

“…Several parameters can indicate the anisotropy of the surface: - epLsar (exact proportion of length-scale anisotropy of relief) or NewepLsar. The latter is the corrected value of epLsar in MountainsMaps® compared to Toothfrax (software previously used for DMTA but not supported anymore) as there was an error in the code to calculate this parameter in Toothfrax (Calandra et al, 2022); - Str is a spatial parameter of the international standard ISO 25178 (specification and measurement of 3D surface textures). It is the ratio of Rmin/Rmax, where Rmin and Rmax are respectively the minor and major axes of the intersection ellipse between the plane z = s with the autocorrelation function f ACF (tx, ty).…”

“…Facets -epLsar (exact proportion of length-scale anisotropy of relief) or NewepLsar. The latter is the corrected value of epLsar in MountainsMaps® compared to Toothfrax (software previously used for DMTA but not supported anymore) as there was an error in the code to calculate this parameter in Toothfrax (Calandra et al, 2022);…”