Background/Question/Methods Structure is a fundamental attribute of forest ecosystems reflecting patterns of establishment, mortality, and disturbance. LiDAR (Light Detection and Ranging) data can extend conventional plot-level structural assessments to the stand and landscape scales. This study addressed two questions related to using LiDAR data to map patterns of canopy structure: (1) How do canopy structural classes relate to forest type and stand age? (2) How does canopy structural complexity vary between and within locations at the plot to multiple patch scales? The study area was in the Tsuga heterophylla (western hemlock) and Abies amabilis (Pacific silver fir) forest zones of the Pacific Northwest, USA. A mixture of secondary (<96 years old) and primary (220 – 350 years old) study sites were used.
Results/Conclusions
Three LiDAR metrics of structural attributes (95th percentile height, rumple, and canopy density) best captured the range of statistical variation in the LiDAR data. This study first classified canopy structure at the plot (0.09 ha) scale at 94 study sites using cluster analysis without making a priori assumptions about patterns of stand development. When class memberships were compared to site ages, general trends of stand development predicted from theory were supported, but considerable variation within and between age classes were observed. A number of factors including forest type, the presence or absence of chronic partial disturbance, conditions of stand establishment, and site productivity could explain the overlapping ranges of canopy structural complexity in secondary and primary forest stands. In a second phase of the study, forest structure at the multiple patch scale (9 ha) was studied using wavelet analysis for a separate set of 91 study sites in the same study area. Three patterns of variance were observed: peak variance at scales ~< 30 m with rapid decline in variance at larger scales, high levels of variance at all scales measured, and low variance ~< 30 m but increasing variance at larger scales. In on-going analyses, associations have been observed between these patterns and stand age but with considerable variation within age classes, suggesting that differences in establishment or partial disturbance may help explain forest structure at multiple scales.
