Dydo Jeremy,

Yes Anova programs should work fine. Theoretically you can also map the time series directly onto the surface (using 3dVol2Surf), then do 3dDeconvolve followed by ANOVA and the like (see comment on time-series 1D formats on SUMA's documentation website). I say theoretically because I have not carried out the analysis myself but the tools are there.
By the way, the new AFNI release contains two new programs SurfMeasures and SurfSmooth. The formal announcement will be made in a day or two but I figured I'd let you know since you seem to be one of them pioneers and might find those programs useful.

Let us know how the analysis goes.

Cheers,
-ziad