Simulation and quantification of the fine-scale spatial pattern and heterogeneity of forest canopy structure: A lacunarity-based method designed for analysis of continuous canopy heights

• Published in: Forest ecology and management Vol. 214; no. 1; pp. 65 - 90
• Main Authors: Frazer, Gordon W., Wulder, Michael A., Niemann, K. Olaf
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 03.08.2005; Elsevier
• Subjects: Animal and plant ecology; Animal, plant and microbial ecology; Biological and medical sciences; canopy gaps; Canopy height models; Canopy structure; Forest management; Forest management. Stand types and stand dynamics. Silvicultural treatments. Tending of stands. Natural regeneration; Forest structure; Forestry; forests; Fundamental and applied biological sciences. Psychology; height; Lacunarity analysis; Laser remote sensing; LiDAR; overstory; Principal component analysis; remote sensing; simulation models; Spatial heterogeneity; Spatial pattern analysis; spatial variation; stand structure; surfaces; Synecology; Terrestrial ecosystems; Canopy structure; Forest structure; Laser remote sensing; Lacunarity analysis; LiDAR; Canopy height models; Spatial heterogeneity; Spatial pattern analysis; Principal component analysis; Forests; Statistical analysis; Forest management; Canopy height models: Canopy structure; Spatial scale; Biodiversity; Spatial variation; Spatial distribution; Ecological niche; Spatial heterogeneity: Spatial pattern analysis; Environmental monitoring; Canopy(vegetation)
• ISSN: 0378-1127; 1872-7042
• DOI: 10.1016/j.foreco.2005.03.056

Forests canopies are dynamic, continuously varying, three-dimensional structures that display substantial heterogeneity in their spatial arrangement at many scales. At the stand-level, fine-scale spatial heterogeneity influences key canopy processes and contributes to the diversity of niche space and maintenance of forest biodiversity. We present a quantitative method that we developed based on a novel application of two well-established statistical techniques – lacunarity analysis and principal component analysis (PCA) – to determine the fine-scale (0.5–33 m) spatial heterogeneity found in the outer surface of a forest canopy. This method was specifically designed for the analysis of continuous canopy height data generated by airborne LiDAR systems or digital photogrammetry; however, in this study we demonstrate our method using a large, well-documented dataset composed of simulated canopy surfaces only. We found that the magnitude of the lacunarity statistic was strongly associated with canopy cover ( R 2 = 0.85) and gap volume ( R 2 = 0.84), while the pattern of decline in lacunarity across discrete measurement scales was related to many size- and density-related attributes of stand and canopy structure (0.27 ≤ R 2 ≤ 0.58) and their diverse vertical and horizontal spatial distributions. PCA uncovered two major gradients of spatial heterogeneity from the 10 dimensions of our original lacunarity dataset. The stronger of these two gradients reflected the continuous variation in canopy cover and gap volume, while a second, more subtle gradient was associated with the array of possible vertical and horizontal spatial configurations that might define any one measure of canopy cover. We expect that this quantitative method can be used to support a broad range of practical applications in sustainable forest management, long-term ecological monitoring, and forest science. Further research is required to understand how these statistical estimates and gradients of measured spatial heterogeneity relate to other ecologically relevant patterns of forest composition, structure, and function.