Use of a non-linear model in examining growth responses of loblolly pine to ozone and acid precipitation

Abstract

Monthly diameter 2 × height ( d 2h ) data were measured over two years in open-top chambers at Auburn University, Alabama. This study exposed seedlings from two half-sibling loblolly pine ( Pinus taeda L.) families to ozone and acid precipitation treatments. For these data, the accumulation of d 2h ) by individual trees over two years was efficiently represented by a six-parameter non-linear model of ln ( d 2h ) as a function of time. Multivariate analysis of variance using these six estimated parameters for each seedling resulted in greater sensitivity to treatment differences as measured by tests of hypotheses than did analysis of covariance on ln (final d 2h ). This result illustrates the importance of utilizing appropriate analyses that can bring as much of the data as is possible to bear on the question at hand. A test for additional information indicated that five of the six parameters contributed important information concerning treatment differences for at least one of the two families tested. It may be inferred that the treatments have an important effect on the nature of d 2h accumulation within a growing season as well as on the d 2h at the end of the growing season.