Performance of the FDTD and lossy material

Does anyone have any reference to articles dicussing performance related issues when modelling lossy material with FDTD, not just with the boundary conditions. According to the conventional FDTD formulation I have assumed that loss is well managed by the coefficients of the FDTD equations, yet this may not be the case accoring to a discussion with some geophysicists.

Any anecdotes or experience with this are much appreciated.