Traditionally, modeling of permanent plot individual tree survival has not considered the multiple sources of heterogeneity and correlation that may arise due to the multilevel data structure inherent in the design (e.g., clustering of trees within a plot). Permanent plots are sampled periodically; therefore, data are interval-censored because it is only known that a tree died between two successive measurement occasions. Here...

A large source of variability in yield predictions is due to estimation of future surviving trees per unit area. Previous whole-stand survival modeling efforts have concentrated on modeling the empirical survival curve. Modeling hazard functions, an approach to survival analysis commonly used in fields such as medicine and sociology, can be applicable to plantation survival estimation. We offer a generalized method for derivin...