In the above Letter, the local surface density of galaxies, Sigma, was miscalculated in such a way that the number of galaxies was divided by the area squared rather than by the area itself. Because of this error, the readers are kindly requested to read off the correct densities (Sigmacorr) in such a way that logSigma(Mpc-2)=3 should have been logSigmacorr(Mpc-2)~2, logSigma=2 should have been logSigmacorr~1.5, and logSigma=1.5 should have been logSigmacorr~1.1. Therefore, the critical local density where the color distribution of galaxies changes abruptly should be ~30 galaxies Mpc-2 (logSigmacorr~1.5) rather than 100 galaxies Mpc-2 claimed in the Letter. Other than that, there is no significant change for the rest of the Letter.