Linking morphology and motion: Testing multibody simulations against in vivo cranial kinematics in suction feeding fishes using XROMM

The heads of ray‐finned fishes contain over 20 mobile skeletal elements, interconnected by a morphologically and functionally varied set of articulations that allow a high degree of mobility. Many fish use this mobility to expand the buccal cavity during feeding to create suction and draw in prey items. Due to this high degree of mobility, a varied set of articulations, and the diversity of cranial forms among fishes, fish skulls provide an excellent system for relating skeletal and articular morphology to skeletal motion. The motion of fish skulls is traditionally simulated using planar mechanical linkage models, such as the 4‐bar linkage. While linkage models accurately predict the in vivo kinematics for particular subsets of bones within the skull, there are other aspects of cranial motion for which planar 4‐bar linkage models have proven insufficient. Additionally, these models have rarely been tested against three‐dimensional kinematics. We used CT data and the R package linkR to build multibody simulations of cranial motion in three species of ray‐finned fishes: largemouth bass ( Micropterus salmoides ), channel catfish ( Ictalurus punctatus ), and common carp ( Cyprinus carpio ). For each species, we created a series of simulations with varying numbers of parameters and degrees of freedom. We then tested the accuracy of these simulations against three‐dimensional in vivo kinematics, collected using X‐ray Reconstruction of Moving Morphology (XROMM). In evaluating model accuracy, we apply a new approach that measures how model error changes as a function of model complexity. We demonstrate the potential, and limitations, of using morphology to predict in vivo kinematics and ultimately to test how the evolution of organismal form influences the evolution of organismal motion and performance. Support or Funding Information This work was funded by NSF DBI‐1612230.