If you talk to Bryan Nowroozi, who has just completed his PhD investigating the role of the fish's spinal column, he will readily confess the reason for undertaking his PhD: ‘I was always interested in how anatomy plays a role in animal performance and how changes in anatomy across various species, with regards to the vertebral column, have an impact on the different forms of locomotion.’ Nowroozi goes on to explain that his father, a medical doctor also interested in spines, may have sparked his interest in the field but it was his PhD supervisor, Elizabeth Brainerd at Brown University, USA, who encouraged him to focus his attention on how bending of the vertebral column could help propel a swimming striped bass (p. 2833).Nowroozi was particularly interested by how much a curved spine could contribute to body stiffness during swimming, explaining that ‘by increasing stiffness, the force required to bend the body increases, and ultimately swimming speed can also increase’. But how much bending is needed to generate stiffness? Nowroozi recounts some of his earlier studies: ‘I dissected out joints from all regions [of the vertebral column] – the cervical region, the abdominal region and the caudal region – and subjected them to mechanical testing experiments, where I was able to quantify the amount of stiffness that they generated during a lateral [sideways] bending motion.’ He found that laterally bending the joints more than 15 deg resulted in substantial body stiffness, with abdominal joints becoming the most stiff. However, the question remained – how much did they bend during swimming?To investigate, Nowroozi decided to use 2D X-ray videoing. So, donning a lead apron, he used a small rod to give a swimming bass a small fright and induce the characteristic C-shaped startle response. By filming the vertebral movements using an X-ray emitter and a camera placed above and below the tank, he could then calculate how much each vertebral joint was bending: ‘You see a fair amount of bending in the vertebra 8 to vertebra 15 region, which corresponds to more or less the anterior abdominal region, and that correlated nicely with what's thought to occur in previous studies’, he explains. To his surprise, in some cases he also found that the cervical joints also bent substantially, which he suspects is due to the fright occurring close to their heads. The biggest surprise, however, was that none of the joints were laterally bending more that 12 deg.During his study, Nowroozi also used a 3D X-ray videoing system developed by his supervisor to investigate the movements in more detail. He recalls it wasn't easy, as he had to surgically implant six tantalum metal markers on to two adjacent vertebrae and although most fish made it out of the ‘operating room’, Nowroozi found that the markers just wouldn't stay in place. However, after 4 years of hard work, Nowroozi had perfected the technique, and he could track the rotations of the markers laterally, as well as dorsoventrally (up and down) and axially (around the spine). He confirmed that the joints did not bend laterally more than 12 deg, but he also found that laterally bending was the only substantial movement the vertebrae made; both axial and dorsoventral rotations were less than 2 deg.In conclusion, Nowroozi thinks the spinal bending doesn't contribute to whole-body stiffness during swimming, at least in striped bass. Upon reflection, he suggests that his study highlights that you need to combine kinematic and mechanical testing studies to understand the role a joint plays – just because the vertebrae can physically bend more than 15 deg doesn't mean they will!