Habitat depth (= sampling depth) was represented by mid-range values for discrete samplings. The minimum depth of occurrence (MDO), defined as the depth below which 90% of the population of a given species is distributed (Torres et al., 1979; Donnelly and Torres. 1988; Seibel et al., 1997; Seibel, 2007), was assumed equivalent to the habitat depth of the species. The depth of the near-surface comllections was assigned 1 m. The attributes of these variables were analyzed simuitaneously using a stepwise multiple regression (forward selection) method (Sokal and Rohlf, 1995). Independent variables were added and removed at the p = 0.05 level; therefore. partial regtession coefficients from the resultant equations are all significant (p < 0.05). unless otherwise noted. The calculations were conducted using a linear regression program in SYSTAT version 10.2.