In this paper, the use of a channel shortening prefilter in conjunction with a maximum a-posteriori probability (MAP) based turbo equalizer is considered. The prefilter shortens the effective channel, thereby reducing the number of equalizer states. As a result of channel shortening, residual intersymbol interference (ISI) appears at the input to the turbo equalizer and the noise becomes colored. To account for the ensuing performance loss, two enhancements to the scheme are proposed. Firstly, a feedback path is used to cancel residual ISI. Secondly, a carefully selected value for the variance of the noise assumed by the MAP-based turbo equalizer is used. Simulations are performed over the highly dispersive Proakis C channel. It is shown that the proposed enhancements give an improvement of approximately 0.65 dB with respect to the unmodified channel shortened turbo equalizer at a bit error rate (BER) of 10E-5.