An experimental analysis of lunule function in the sand dollar Mellita quinquiesperforata

Abstract

Coefficients of lift and drag for fifty lunulate sand dollars [Mellita quinquiesperforata (Leske 1778)], poised over a flat, sandy surface, were determined at seven different air stream velocities. Coefficients of static friction were obtained experimentally on level sand in standing water. Critical velocity, at which a sand dollar would be dislodged, is defined as that current velocity in which the force of drag is equal to the product of residual weight (i.e. weight-lift) and the coefficient of static friction. Small individuals have lower critical velocities than large ones: 26 cm s⁻¹ and 43 cm s⁻¹ for individuals 48 and 94 mm in diameter, respectively. Measurements were repeated for 20 of the controls, their lunules smoothly filled with modelling clay, and 20 from which lunule spines were removed. In every case, blocking the lunules decreased critical velocity, by an average of 15%; this was shown to be statistically significant (P>0.01). Removal of spines, which partially occlude the lunules, significantly (P>0.01) increased critical velocity by an average of 13%. Direct measurements in a water tunnel showed that calculated values of critical velocity were slightly too high, but otherwise confirmed the role of the lunules. Allometric analysis showed that as sand dollars increase in size, the distances of the lunules from the center of the disk increase less rapidly than the radii. A principal components' analysis of profile measurements of nine species indicated that lunulate and non-lunulate sand dollars are quite different in form. Lunulate species tend to be thinedged, flat-domed and larger in diameter. Some alternative ideas of lunule function are discussed and it is shown that there are no allometric or experimental data to support any of the suggested feeding hypotheses.