Continuum robots always exhibit strong adaptability for exploring unstructured environments and high safety for achieving human-machine interactions. To further enhance their agility, it is imperative that these robots integrate both bending and elongation motion patterns. However, employing a unified actuation strategy will inevitably lead to motion coupling between these patterns, which poses a significant challenge in predicting and regulating morphing profiles. In this letter, inspired by the motion regulation observed in muscle fibers of elephant trunks, we proposed a hybrid-driven modular continuum robot. This robotic configuration enables the modulation of motion patterns through various approaches, thus achieving independent control of bending and elongation motions. Here, each backbone of the module not only produces an elongation rate of 62.76 ± 0.01% but also morphs into a bending profile with an angle of 35.53 ± 1.18°. Then, through theoretical and experimental investigations on a one-segment continuum robot composed of three modules, we illustrated that segregating the actuation approaches for bending and elongation motion patterns effectively mitigates motion coupling, enhancing the motion accuracy of continuum robots. In addition, we also demonstrated the versatility of a two-segment continuum robot, which morphs into diverse profiles for grasping, moving, and releasing items by combining different motion patterns.