How do fish break the speed limit?

Many studies have shown that tail beat frequency of teleost fish is closely related to their swimming speed. These findings were developed further by Wardle1, who showed that maximum tail beat frequency and thus swimming speed could be predicted by measuring the twitch contraction time of the fast swimming muscles. The contraction time increases with the length of the fish and decreases with increase of temperature. Several authors have observed and recorded swimming speeds greater than the predicted speed limit. For example, Stevens and Neil2 report that skipjack tuna (Katsuwonus pelamis (L), length (L) 0.5 m) could swim at a speed of 10m s−1 at 32 °C, whereas the maximum predicted speed (extrapolating Wardle's1 data) was only 8m s−1. Walters and Fiersteine3 recorded a yellowfin tuna (Thunnus albacores (Bonnatare), L = 0.98m) moving at 20.8m −1 and Wahoo (Acanthocybium solandin (Cuvier), L = 1.13 m) at 21.4 m s−1. These two records show just twice the predicted maximum swimming speeds. Unfortunately no observations of the tail beat frequency were made during these high-speed swim records. Brill and Dizon4 have shown that five skipjack tuna (L 0.37–0.43 m) had muscle contraction times similar to non-scombroid teleost species of this size measured by Wardle1. It seems therefore that the tuna family (Scombridae) do not have faster maximum tail beat frequency than other teleosts of the same size and temperature. We show here how this paradox can be solved by the fish using a different swimming style, requiring greater power and efficient interaction between the propelling surfaces and the water, but allowing the fish to move twice as far for each tail beat.