Dry deposition and canopy leaching rates in deciduous and coniferous forests of the Georgia Piedmont: an assessment of a regression model

Abstract

Simple and multiple linear regression models were used to estimate dry deposition and canopy leaching sources of constituents in throughfall and stemflow for adjacent deciduous and coniferous forests at the Panola Mountain Research Watershed (PMRW), Georgia. The models were evaluated for their comparability and statistical significance, and model fluxes were compared with fluxes derived from field collection and a dry deposition inferential model, i.e. field-inferred fluxes. The multiple regression model used antecedent dry period and event quantity as independent variables. Field-inferred fluxes indicated that dry deposition was a major source of SO 4 2−, N03 and H+ in throughfall, and canopy leaching was the major source of Ca 2+, Mgt +, K + and Cl − in both deciduous and coniferous forests. The canopy leaching flux predicted by the multiple linear regression model was:(1) within 30% of the field-inferred fluxes for NH 4 + in the deciduous and coniferous forest during the growing season; (2) within 40% of field-inferred fluxes for Ca 2+, Mgt +, and K + in the deciduous forest during the growing season and 50% in the dormant season; (3) within 67% of field-inferred fluxes for H + in the deciduous forest and Mg 2+ and K + in the coniferous forest. Similarly, dry deposition predicted by multiple linear regression model when compared with the field-inferred flux for the growing season was:(1) the same for Na + in the deciduous forest; (2) 10% more for Ca 2+ in either forest; (3) about 20% less for NH 4 + in the deciduous forest; (4) 70% more for Mg 2+ in the deciduous forest and K + in the coniferous forest; (5) 50% less for NH 4 + in the coniferous forest; (6) 70% SO 4 2− less for SO 4 2− in either forest. The multiple regression model, in general, was best applied to estimate dry deposition and canopy leaching of some constituents for the deciduous forest in the growing season, but it had limited applicability and cannot substitute for field collections for analysis of constituents with significant dry deposition at PMRW.