The original statement "the observed increase could be largely due to this natural variability" seems to have come from a preliminary report in 1990. I wonder if you've had a chance to do more re-search since?