MODELING THE REGENERATION OF OAK STANDS IN THE MISSOURI OZARK HIGHLANDS

This paper describes a method for modeling the regeneration of even-ag ed oak stands in the Ozark Highlands of southern Missouri. The approac h is based on (i) a growth model that is applicable to both oak sprout s and advance reproduction and (ii) a method for probabilistically est imating future size distributions of trees. The modeling method is ill ustrated using sprouting frequency, survival, and 5th-year height data for stump sprouts of five oak species. To consider the large residual variation in estimates of future sprout heights, nonlinear regression estimates of heights and their prediction errors are simultaneously u sed to estimate the probability that a sprout originating from a paren t tree of a given species and diameter will grow into a specified 5th- year height class. To account for sprouting and survival failures, tho se probabilities are multiplied by logistic regression estimates of th e probability that a parent tree will produce a sprout that survives t o age 5. The resulting joint probabilities facilitate predicting futur e height distributions of surviving sprouts when the model is applied to a preharvest inventory of overstory trees.