Because of the nonlinear pattern of RSG in Douglas-fir trees, we
examined a variety of functions that have concave up followed by
concave down behavior, have asymptotes that would represent the
initial low RSG in rings near the pith and eventual high RSG
reached in older rings, and have first and second derivatives from
which the inflection point at the change in concavity can be
determined. After review of the literature and preliminary
analyses, we concluded that the four-parameter logistic function,
would best model the data. Mora et al. (2007) also found that the
four-parameter logistic function was best for modeling site and
treatment effects on the mean pattern of inter-annual RSG of
loblolly pine (Pinus taeda L.) trees, a species which also has an Sshaped
RSG pattern.