Protist, Vol. 152, 329–338, December 2001 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/protist Introduction Most dinoflagellates have a characteristic flagellar arrangement: a transversal flagellum that encircles the cell and, perpendicular to it, a trailing flagellum. Both flagella are inserted on what has been defined as the ventral side of the cell. The transversal flagel- lum is usually situated within an equatorial or slightly descending groove (the girdle or cingulum) and runs to the left, usually encircling the cell completely. The proximal part of the longitudinal flagellum lies freely in another groove referred to as the sulcus. The mechanisms by which this arrangement causes the cells to swim and the adaptive arrange- ment of the flagella have been debated for almost a century, but this has not so far provided a convincing model of dinoflagellate swimming and some contra- dictory observations and ideas have prevailed in the literature. Levandowsky and Kaneta (1987) and Goldstein (1992) have most recently reviewed the topic. Most authors seem to agree that the flagel- lates swim in a helical path in which the ventral side is directed towards the axis of the helix and that the cells rotate around their longitudinal axis. However, there has been no consensus as to whether it is a right or left helix or in which direction cells rotate, and it has even been suggested that cells can alter- nate between these rotational directions. In practise it is difficult to determine rotational directions micro- scopically because different focal planes may leave the impression that directions change. Following Lindemann (1928) (cited in Levandowsky and Kaneta 1987) who claimed that propulsion is unaf- fected by loss of the trailing flagellum in Hemidinium cells, it has generally been assumed that the transversal flagellum is at least in part responsible for translational motion whereas the trailing flagel- lum serves mainly for “orientation”. Hand and Schmidt (1975) also observed residual translational motion in Gyrodinium dorsum cells deprived of the longitudinal flagellum. Jahn et al. (1963) observed flow-lines in the form of a torus encircling the girdle How Dinoflagellates Swim Tom Fenchel 1 Marine Biological Laboratory (University of Copenhagen) Strandpromenaden 5, DK-3000 Helsingør, Denmark Submitted June 22, 2001; Accepted August 30, 2001 Monitoring Editor: Donald M. Anderson Dinoflagellates possess two flagella; usually these are directed perpendicular to one another consti- tuting a transversal flagellum and a longitudinal, trailing flagellum, respectively. The transversal flag- ellum causes the cell to rotate around its length axis. The trailing flagellum is responsible for the translation of the cell; due to its asymmetric insertion it also causes a rotation of the cell around an axis perpendicular to the longitudinal axis. Together, these two rotational components result in a he- lical swimming path. Cells can vary the two rotational components independently as well as the translational velocity. With these three degrees of freedom, cells can vary the parameters of their he- lical swimming paths for steering. Dinoflagellates use this mechanism for orientation in chemical concentration gradients (“helical klinotaxis”). Protist 1434-4610/01/152/04-329 $ 15.00/0 ORIGINAL PAPER 1 fax 45-49 26 11 65 e-mail tfenchel@zi.ku.dk