Tree Grade Prediction for Mississippi Bottomland Hardwoods Using Discriminant Analysis
Information from 150 permanent plots in Mississippi bottomland hardwood stands was used to construct linear discriminant functions to predict individual tree grade for sweetgum (Liquidambar styraciflua L.), and three species of red oak: cherrybark (Quercus pagoda Raf.), water (Quercus nigra L.), and willow (Quercus phellos L.). Two versions of the functions were produced, one general set for all trees greater than 9.0 in. in diameter (including culls), and another set for trees grade 3 or better. The general set of discriminant functions successfully predicted tree grade for 50% to 63% of the trees in the database, depending on species. Prediction success improved to a range of 62% to 81% when culls were excluded from the analysis. South. J. Appl. For. 17(3):120-123.
No Supplementary Data
No Article Media
Document Type: Journal Article
Affiliations: USDA Forest Service, Southern Forest Experiment Station
Publication date: 01 August 1993