R help - Nov 2006 - finding significant change points after fitting GAM using TPRS

Does anybody know if code or functions exist for determining significant
change points after fitting GAM in 'mgcv' using thin plate regression
spline and 'ts' bases?  I would like to identify years where the
curvature
or change in trajectory of a longterm annual population curve is
significant.  Thank you.

Deahn DonnerWright
email: ddonnerwright at fs.fed.us