KEY WORDS: Swimming, undulations, technique, movement.

Finswimming is a speed competition sport practiced at the surface or underwater with different monofins of variable rigidity. Alike the motion of dolphins (Videler, 1981), propulsion bears on the vertical displacement of the whole body. The use of the upper body is forbidden for propulsion purposes. The vertical displacement of the body during the stroke cycle has been described as wave-like (Ungerechts, 1982). Since such motion could be characterized by specific amplitudes of oscillations. Such oscillations were also specified by a particular frequency and phase relationship (Sanders et al., 1995).

Our knowledge on finswimming bears mostly on underwater finswimming. It was shown that the wave-like motion traveled along the body in the caudal direction and that finswimmers adapted their undulations in frequency and amplitude, starting the propulsive motion at the hip level (Baly et al., 2002). Nevertheless, because of the air-water interface, finswimmers' motion is quite different in amplitude and frequency at the surface than underwater.

Concerning finswimming at the surface, Zamparo et al. (2002) performed a kinematic study that quantified the efficiency of human swimming in using fins. Another study revealed the effect of the monofin shape on the propulsive forces by analyzing the change in the swimmer's velocity over one cycle of the monofin's motion (Tamura et al., 2002).

Finally, a finswimmer is at one and the same time a propelling and a propelled body. Given the speed reached (3.89 m·s^-1), body pitch and knee bending must be weak in order to limit the frontal surface area, as defined by Vogel (1994), a main factor of hydrodynamic resistance. The gliding and the propulsion must be finely tuned in order to optimize performance. A previous study at the surface showed that the race distance and practice level increase the stroke frequency and the ankle vertical amplitude (Gautier et al., 2004). However the gender influence on kinematic parameters was not taken into consideration.

In the present study, we hypothesized that the gender affects finswimming performance in association with the practice level and the race distance. Thus, we aimed to quantify finswimming at the surface in terms of kinematic parameters such as the swimming speed, the movement excursion, and the degree of knee bending according to gender, race distance and practice level. Their analysis may provide information useful to coaches and technicians in order to reduce drag of the finswimmer.

The experiment was carried out in a 50 m pool. Six elite finswimmers members of French National Team (3 males, 3 females) and six novices (3 males, 3 females) who started competition six months ago participated on a voluntary basis. For the elite group, average age was 20 ± 4 years, body mass 69.1 ± 8.1 kg, and stature 1.74 ± 0.8 m. For the novices, average age was 25 ± 3 years, body mass 64.6 ± 5.2 kg, and stature 1.69 ± 0.4 m. Finswimmers were instructed to perform 3 random trials of 25 m per race distance (100 m and 800 m). Using a snorkel, finswimmers performed trials with their standard monofins.

Passive disk markers were applied on selected joints to facilitate their tracking and video analysis: wrist, elbow, shoulder, hip, knee, and ankle were visible to the camera throughout the stroke cycle. Video images were collected from two digital cameras (SONY VX-2000E, 50Hz). One digital camera was immerged and securely fixed in a watertight box at 0.15 m below the water surface and recorded in sagittal plane. A calibration was carried out separately for each trial, at the middle tank and 5.34 m away from the camera. The width of the optical field of view in the plane of motion was 6 m. The other digital camera (SONY VX-2000E, 50Hz) was placed in the axe of the first to record motion above the water in order to control whether the whole body was kept underwater, except for the snorkel, vertex, and heel.

The markers were semi automatically tracked (3D Vision, Kihopsys). The analysis period comprised one complete stroke cycle starting from the video frame corresponding to the first whole body appearance. Body marks were seized to determine a 5-segment human model: forearm, arm, trunk, thighs, and shank. The knee bending angle was determined using the relative angle between the distal extension joint center of the thigh and the shank. As for all biomechanical data, random error from the digitizing process was reduced using a recursive fourth-order Butterworth digital filter with a frequency cut off at 4 Hz.

The analysis consisted of a repeated measures ANOVA with gender (male vs. female) and expertise level (novice vs. expert) as between-subject factors, and distance (100 m vs. 800 m) as within-subject factor (N = 12). The dependant variables were the velocity, the amplitude and the frequency of oscillation, and the knee angle. All effects were declared significant at a threshold of p < 0.05.

The average speed for 800 m was lower than for 100 m for the experts (2.09 m·s^-1 ± 0.17 m·s^-1 and 2.66 m·s^-1 ± 0.26 m·s^-1, respectively). Novices were slower than experts (1.69 m·s^-1 ± 0.11 m·s^-1 and 2.07 m·s^-1 ± 0.18 m·s^-1 respectively). Irrespective of the distance, males were faster than females (2.58 m·s^-1 ± 0.40 m·s^-1 vs. 2.15 m·s^-1 ± 0.31 m·s^-1).

Table 1 shows that the average frequency of stroke cycle decreased with increasing distance for both experts and novices (p < 0.01). On 800 m, novices were slower than elite finswimmers, although their frequency was higher. Such frequency was higher for males (1.61 Hz) than female (1.22 Hz) on 100 m as well as on 800 m males (0.81 Hz vs. 0.61 Hz, respectively).

Figure 1 indicates that the amplitude of body oscillation increased as a function of race distance (p < 0.01) for all expertise levels. The upper limbs amplitude of oscillation (i.e., wrist, elbow, and shoulder) was smaller for experts than for novices. Concerning the lower limbs (hip, knee, and ankle), only the amplitude at the ankle level was larger for experts than for novices. There was non difference between male and female in vertical ankle amplitude (0.56 m vs. 0.51 m, respectively). However, wrist amplitude was significantly lower in female (0.22 m) than in male (0.26 m).

Knee bending was a factor of performance as a function of the expertise (p < 0.01) associated with the important vertical amplitude of the ankle. Bending for experts was smaller than for novices (119.25 ± 3.31 deg. and 104.60 ± 4.96 deg. respectively on 100m). On the 800 m, novices' knee bending was higher than the elite's one (94.85 ± 7.30 deg. vs. 108.87 ± 6.61 deg.). However, such main effects on knee bending were comprised within a 4-way gender x race distance x practice interaction. This result will be discussed in the following part.

• Finswimmers are at one and the same time a propelling and a propelled body. This study investigates the undulatory motion performed by finswimmers at the surface.
• Elite male finswimmers were pitching more acutely than female swimmers and showed a smaller knee bending than both novices and elite female swimmers.
• Finswimmers tended to perform a dolphin-like motion, which is used underwater situation and optimizes hydrodynamics.