Air pollution and Pinus strobus height growth: a soil/site modelling approach

The hypothesis that subtle height-growth reductions can be detected in eastern white pine ( Pinus strobus L.) near a major urban/industrial area was tested in northeastern Ohio, U.S.A. Total height and 20-year height intercept above 1.4 m were compared for dominant and co-dominant plantation white pines within the Cleveland-Akron area on the Cuyahoga Valley National Recreation Area (CVNRA) and at similar sites in a remote locale. Predictive height equations specific to the two locations were developed using stepwise multiple-regression analysis on principal-component scores extracted from an original set of 38 soil/site variables. The resulting models were used reciprocally to compare and actual and predicted tree heights for plots at each location. Multivariate comparisons of soil/site factors indicated the two locations were not highly analogous, particularly with regard to soil chemical and physical properties. Use of the models to test the original hypothesis was deemed inappropriate because some of these differences in inherent site variables also had a large effect on height estimates; however, the models were unbiased estimators for the specific regions for which they were derived. Height was primarily a function of tree age, thickness of surface soil horizons, soil moisture-holding capacity, and various expressions of cation presence. Heights of white pines on the CVNRA (and on the remote site) exhibited no unusual or unexplained deviations from the normal range based on our exerience in Ohio. Soils of the CVNRA exhibited lower surface horizon pH than the published ranges for the soil series and were comparatively low in exchangeable bases and available P, leaving open the question of possible acid-deposition-mediated soil changes on the CVNRA white-pine sites. The multivariate techniques used here provided useful insights into comparative soil/site properties and their differential effects on white-pine height growth at the two locations.