TESTING MODELS OF UNTHINNED RED PINE PLANTATION DYNAMICS USING A MODIFIED BAKUZIS MATRIX OF STAND PROPERTIES

The comprehensive graphical matrix of even-aged stand property inter-d ependence, first developed by E.V. Bakuzis, facilitates identification and understanding of the relationship among stand properties. Bakuzis ' original matrix of eight stand properties, contained 64 cells, was s ymmetric, but only about 11 of the 64 relationships had shown enough r egularity among species to have been named. In this paper I simplify t he Bakuzis matrix by reordering the rows and columns to make a mof-e c ompact, lower triangular arrangement of eight rules or law-like relati onships. I then demonstrate matrix use by looking for structural flaws in two models of unthinned red pine (Pinus resinosa) plantation dynam ics in the Lake States (Minnesota, Wisconsin, and Michigan) (STEMS and REDPINE), and one from Ontario (yield tables from Petawawa). Flaws we re found in each source of projections. REDPINE violates the Sukachev effect, predicts trees will have larger diameters on poor sites than o n good sites, and that site has a significant effect on the mean heigh t-stem frequency relation All bi-variate relations for the Petawawa da ta are identical, which violates several rules. The STEMS program does the poorest job of the three methods of predicting unthinned red pine plantation stand development, primarily because site index has very l ittle effect on any of the stand variables. (C) 1997 Elsevier Science B.V.