Estimating Bite Force in Three Aquatic Turtle Species with Disparate Bite Strategies: Exploring the Impact of Assumptions on Theoretical Bite Force Modelling and Interpretation

Abstract Body : Bite force can have a direct effect on fitness and is a popular metric for ecological reconstructions and comparisons. In the absence of in vivo data, bite force is commonly estimated using static bite force models, though they may be highly sensitive to the input variables which can carry significant assumptions. Three such critical parameters are specific tension (Po), physiological cross-sectional area (PCSA), and mechanical advantage (MA). Due to their relative ease of measurement from cadaveric or fossil organisms, MA and PCSA are regularly the subject of comparative studies. In contrast, Po is treated as a standardized value, often between 25-40 Ncm-2 for jaw muscles, though it has long been known that specific tension is not a constant. Po requires in vitro/vivo measurement and thus is available for very few species. Here, we examine the relative effects of input variables on estimates by calculating and manipulating theoretical bite force in three aquatic cryptodiran species utilizing distinct bite strategies to capture and ingest prey: Trachemys scripta (nonspecialized), Malaclemys terrapin (static-forceful biting), and Chelydra serpentina (fast and forceful biting). Using diceCT, we measure the MA, normalized fiber length and angle and muscle volume to calculate PCSA of the internal jaw adductor muscles from 3D digital models. To estimate Po, we scale published in vivo bite forces to specimen size and algebraically rearrange the model to solve for Po. Whole sample averages of each variable were fixed as constants in the model while allowing one to vary as the measured value for each specimen, and the proportional effect measured. Greater architectural homogeneity gives the M. terrapin adductor mandibulae the highest PCSA/volume of the species, though not enough to explain the order of magnitude higher bite force over T. scripta. Meanwhile, functional regionalization into long anterior and short posterior fibers indicates a greater proportion of the muscle dedicated to fast jaw-closing in C. serpentina and a concommitant lower PCSA/volume than even T. scripta. The largest single determinant of bite force in the 3 species is the ratio of volume to fiber length (68.2%), which was similar to overall head size (PCSA scaled to jaw length, 78%), followed by the Po value chosen for the calculation (45.1% over the known range of vertebrate muscle, 5-60 Ncm-2), while fiber angle (5.8%) and MA (5.8%) had small effects. Po estimates suggests that contractile properties vary among species (T. scripta 35 Ncm-2, M. terrapin 53 Ncm-2, C. serpentina 40 Ncm-2), but not between sexes within a species. Varying fiber type and proportion optimizing different muscles to different tasks likely explains the variation in estimated Po. The large difference in estimated Po between T. scripta and M. terrapin explains their bite force disparity where PCSA does not, allowing the latter to consume armored molluscan prey despite similar body sizes. Large muscle volume ensures that C. serpentina can close its jaws at high velocity against water pressure, while the higher Po ensures that the bite is forceful enough for prey-capture and defensive bites. Among just three species we estimated large heterogeneity in Po, making comparisons to real-world fracture forces and in vivo measurements uninformative without the ability to predict Po. This significantly reduces the predictive power of theoretical bite force in turtles and indeed all vertebrates until there is wider sampling of jaw muscle physiology.