Growth of Juvenile and Adult Trees Across Productivity Gradients
Light, water and nutrient availability affect growth of understory and canopy trees in complex stands. Different site types have different levels of available belowground resources. Within a region of uniform climate, the ecological classification system in BC is based on identification of unique site types (site series) that cover a gradient from dry-poor to moist-rich sites (www.for.gov.bc.ca/hre/becweb). We are using this framework in the sub-boreal spruce zone around Smithers to study how growth of juvenile and adult trees is affected by site type and neighbourhood canopy spatial arrangement. Two projects are currently underway:
Erica Lilles, Rasmus Astrup, David Coates, and Marie-Lou Lefrançois
FIA-FSP project Y092022
This project investigates how light availability, nutrient availability, soil water availability, and canopy structure influence understory tree growth and physiological traits of sub-boreal tree species. This study expands on the earlier work of Wright et al. (1998) and Coates and Burton (1999), where juvenile tree growth as a function of light was predicted on mesic sites in northern temperate, sub-boreal and boreal forests.
A manuscript titled "Multiple resource limitation of sapling growth in sub-boreal forests" is in preparation.
2. The effect of site type on competitive interactions among trees in complex-structured mixed species sub-boreal forests
Dave Coates, Rasmus Astrup, Erica Close, and Aaron Trowbridge
FIA-FSP project Y082254
This project has two parts:
(1) Development and testing of a remote sensing technique to extract spatial information for use in forest research and management
There are several documented approaches for identification and extraction of individual tree crowns from high resolution aerial and satellite images. The most common techniques utilize high-contrast images acquired under sunny conditions during the summer months and localize individual crowns either by local brightness maxima, valley following, or using template matching techniques. However, these image types and techniques suffer common problems; difficulty in resolving touching and overlapping crowns, incorrect identification of understory vegetation in gaps as tree crowns and smaller trees lost in shadow of larger trees. To solve these problems we explore an object-oriented tree crown delineation approach carried out on very high resolution (5 cm pixel) visible spectrum aerial images acquired in the winter with overcast conditions where snow obscures ground cover and the overcast conditions minimises loss of smaller trees due to shading. The spectral characteristics of low-contrast winter conditions precluded complete reliance on one established technique. Therefore, we integrate a spectral local maxima approach with template methods based on morphological and relational characteristics. We are developing hierarchal object based algorithms in successive iterations using Definiens Developer 7.0. Rules are developed for young single-storied and mature two-storied stands populated by lodgepole pine, hybrid spruce and subalpine fir. Algorithm validation uses 12 manually stem-mapped plots consisting of 1145 trees.
(2) Development of individual tree growth functions utilizing spatial data
We are using likelihood methods and model selection criteria to examine how above- and below-ground competitive interactions affect tree growth in complex stands across a productivity gradient availability. Our spatially explicit (stem-mapped) dataset includes 1,952 sampled lodgepole pine, interior spruce, and subalpine fir trees and neighbouring tree species in the sub-boreal spruce forests of British Columbia. This study expands on the earlier work of Canham et al. (2004) and Coates et al. (2009), where adult tree growth was predicted on mesic sites in northern temperate forests at Date Creek.