There has long been controversy over the reputed high efficiency of certain swimming animals, dolphins in particular. Observations of their swimming speed, and calculations of the likely drag on a well-streamlined body under those conditions implied a tremendous power output would be required -- well above what could reasonably be produced by the available musculature. The only conclusion seemed to be that the friction and/or form drags must in fact be very much smaller than usual, and hopes were raised that hitherto unknown drag reduction mechanisms could be found in nature, and then emulated in engineering designs. Unfortunately, the original observations were themselves flawed, as were some of the calculations, and the entire phenomenon was shown to be a chimera that dematerialised upon close scrutiny.

Much more fruitful, turned out to be some of the research work on the fundamental hydrodynamic mechanisms of aquatic propulsion. Following the initial impetus given by Prof. Sir James Lighthill in his groundbreaking (or should that be fluid-shearing?) review/research articles on aquatic animal propulsion (see [Li69], [Li70]), H.K. Cheng and colleagues ([HK76], [CM84], [KSC90]) outlined the design principles, and geometry and kinematics of oscillating, lifting surfaces of high hydrodynamic efficiency. In particular the offsetting effects of sweep and local centreline curvature were identified, and the two most visible planforms of the tails of fast fish, the crescent-shaped lunate tail, and the swept-back v-tail were both shown to be potentially highly efficient solutions the problem. The next time you see a picture in a sport fishing paper of a marlin, swordfish, tuna, or any of the fastest, most powerful fish, note the lunate tail with it's curved, thin (high aspect ratio) shape.

Information provided by: http://ae-www.usc.edu