Predicting understory light microclimate from stand parameters in young paper birch ( Betula papyrifera Marsh.) stands

Abstract

On productive sites, young paper birch ( Betula papyrifera Marsh.) outgrows most of the conifers with which it is associated. Due to its dominant position in the canopy, and the potentially high densities of birch regeneration found on some sites, its presence may lead to a reduction in the amount of light reaching subordinate conifers. To evaluate the potential use of stand attributes for estimating competition for light and understory light levels, we measured fractional transmittance of light (DIFN 1) using Licor LAI-2000 Plant Canopy Analyzers, at a height of 1 m under the canopies of nine 10–40-year-old paper birch stands in southern interior BC. Various measures of broadleaf and conifer abundance were made at each of the 55 sample points where light measurements were taken. Vertical light gradients were also measured to a height of 8 m within birch canopies at four of the nine sites. Regression models were developed to predict understory light availability from broadleaf and conifer abundance. The models that best-fit the data were based on broadleaf basal area (BA b), and included either conifer cover ( C c) ( R 2=0.83) or conifer basal area (BA c) ( R 2=0.80). Broadleaf density (TPH b) alone was a poor predictor of understory light, but in combination with broadleaf quadratic mean diameter (QMD b) or height of the tallest broadleaf (HT b) the models improved ( R 2>0.85). Light availability increased rapidly with relative height in birch canopies, especially where transmittance to the understory was low; at 35% of the height from live crown base to top of the canopy (relative height of 0.35), approximately 80% full light was available. A diagrammatic representation of the relationship between DIFN 1, QMD b and TPH b is presented as a potential tool for the development of stand-tending prescriptions. DIFN 1, BA b and TPH b were all highly variable across individual sites, indicating that the models should be applied at the microsite level.