Variance of size-age curves:: Bootstrapping with autocorrelation

We modify a method of estimating size-age relations from a minimal set of individual increment data, recognizing that growth depends not only on size but also varies greatly among individuals and is consistent within an individual for several to many time intervals. The method is exemplified with data from a long-lived desert plant and a range of autocorrelation factors encompassing field-measured values. The results suggest that age estimates based on size and growth rates with only moderate autocorrelation are subject to large variation, which raises major problems for prediction or hindcasting for ecological analysis or management.