Evolutionary Biology

Jacobs, Craig T.; Dominguez, Jaron; Devine, Jay; Vidal-García, Marta; David Aponte, J.; Da Silva, Cassidy; Marcucio, Ralph S.; Hallgrímsson, Benedikt

doi: 10.1007/s11692-026-09667-0pmid: N/A

Robustness is a common feature of many developmental systems. This is perhaps best demonstrated by tolerance to heterozygous loss of function for many mutations. This non-linear interaction acts protectively, preventing major deleterious effects and preserving correct development. Critical cell signalling pathways that are reused throughout development, such as Hedgehog (Hh) signalling may tend to exhibit high degrees of robustness to variation in gene expression level due to selection against the widespread deleterious effects of their functional perturbation. Here, we investigate the robustness of Hh signalling in zebrafish by selectively inhibiting Hh signalling at a specific timepoint with varying concentrations of the small molecule inhibitor cyclopamine. Through increasing the concentration, we alter the amount of Hh signalling the embryo receives from low inhibition to complete inhibition. Using 3D geometric morphometrics, we uncovered a clear threshold at which Hh inhibition generates severe phenotypes and further defined a non-linear relationship between Hh signalling output and craniofacial morphology. Interestingly, this non-linear relationship also appears to allow for new directions of shape change that differ from the major “Hh loss” axis. This is particularly noticeable at below-threshold levels of inhibition, where the phenotypic changes are subtle and non-severe, but the direction is distinctly different. This change in covariance structure along the non-linear response curve suggests that the probing of limits of robustness for developmental processes may be a mechanism for overcoming developmentally constrained covariation structure to establish novel directions of evolutionary change.

Allen, Jonathan G.; Fisher, Taylor L.; Albert, James S.

doi: 10.1007/s11692-026-09668-zpmid: N/A

The Navajini is a clade of Neotropical electric fishes specialized to inhabit deep river channels in the lowland Amazonian, with species that exhibit high craniofacial disparity and a highly conservative post-cranial body shape. This study investigates the morphological, ecological, and developmental factors contributing to phenotypic diversification in this clade. By integrating 3D geometric morphometrics, phylogenetic comparative methods, and ancestral state estimation within a robust phylogenetic framework, we examine the evolution of skull shape across the group. Our approach combines high-resolution micro-CT scanning with multivariate analyses to identify major axes of craniofacial disparity, and to assess how these patterns relate to ecology, sexual dimorphism, and phylogenetic history. Rather than being randomly distributed or purely environmentally plastic, key traits such as lower jaw elongation, opercle morphology, and maxilla orientation exhibit strong phylogenetic signal and structural consistency within genera. We explore how certain lineages have accessed distinct regions of craniofacial morphospace, likely by means of trophic specialization and developmental shifts. Sexual dimorphism emerges as a major contributor to morphological disparity in some but not all subclades. These results highlight repeated evolutionary trajectories among taxa that share similar ecological roles or mating strategies. Together, these findings provide new insights into the processes driving phenotypic evolution in one of the most morphologically diverse lineages of Neotropical freshwater fishes.

Guillaumot, Charlène; Sklab, Youcef; Dubied, Morgane; Laffont, Rémi; Navarro, Nicolas

doi: 10.1007/s11692-026-09669-ypmid: N/A

Geometric morphometrics has become a widely used toolkit for quantitative morphology across diverse biological fields. It relies on the precise placement of anatomical landmarks in 2D or 3D. However, manual landmarking, especially in 3D, is labour-intensive, time-consuming, prone to intra- and inter-observer variation and requires significant expertise. These limitations are increasingly incompatible with the demands of high-throughput imaging workflows. This bottleneck has spurred efforts to automate landmarking, often through deformable registration of 3D images, surfaces, or point clouds. While these methods can effectively capture overall shape, they may introduce biases in landmark localization and distort patterns of variances and covariances. Previous attempts to mitigate these issues with deep learning (DL) have generally focused on specific anatomical parts of certain species, most notably human faces and craniofacial anatomy in model-organisms. In this study, we introduce a DL-based pipeline for refining landmark placement. The pipeline begins with approximate landmark predictions derived from a simple rigid registration. These initial predictions define local 3D surface regions, which are then parametrized into 2D using least-squares conformal maps and enhanced with colorization informed by diverse geometry and ambient lighting. The resulting standardized representations are then used to train Transformer and Convolutional Neural Network (CNN) architectures. The approach was evaluated using an open-access dataset of 3D mouse skull models with manually digitized landmarks. Focusing on ten distinct landmarks, we compared our rigid + DL predictions with those from a mainstream global registration approach, ALPACA. We assessed both distance accuracy to the manual groundtruth and preservation of the shape variation patterns. We also examined the effects of various hyperparameter settings strategies, offering practical guidelines for future implementations. Our results demonstrate the feasibility of using 2D colored-enhanced local geometric representations for accurate landmark position refinement with DL. The proposed approach mitigates biases associated with the considered deformable registration method and establishes a methodological foundation for more generalizable, high-throughput morphometric analyses using DL.