A comparison of alternative data sources for modeling site index in loblolly pine plantations

Abstract

Site index curves are generally developed from one of three data sources: (i) permanent sample plots, (ii) temporary sample plots, or (iii) stem analysis data. Permanent sample plots are considered to be the best data source for modeling height–age relationships; however, due to time and cost constraints, analysts may consider using temporary sample plots or stem analysis data for equation fitting. Temporary sample plot and stem analysis data, although more quickly obtained, require assumptions that are often not met when modeling site index. The question becomes how models developed from temporary sample plot or stem analysis data compare with models developed from permanent sample plot data. Data from a region-wide study in loblolly pine (Pinus taeda L.) plantations were used to develop site index curves from each of the three data sources. A form of the Chapman–Richards model was used for all three data sources to guard against confounding and to discern differences among the data sources when modeling height–age relationships. For the comparison and evaluation of behavior of different functions, the Schumacher model was also fitted to the three data sources. Curves developed from temporary sample plot and stem analysis data did not reproduce the height growth pattern exhibited in permanent sample plots, although curves derived from temporary plots were closer to the height growth pattern.