Plot-Scale Spatial Structure and Variability of Forest Floor Properties in an Old Growth Stand of Coastal British Columbia

The spatial structure of physical and chemical forest floor properties can be difficult to quantify, because a large number of biotic and abiotic environmental variables contribute to their distribution. In this study, I investigated the spatial structure of selected forest floor properties at the plot scale (range of meters) in a coastal old-growth stand in southwestern British Columbia. I also explored methods of quantifying the amount of variation according to different environmental variables, with a focus on the understory hardwood species vine maple (Acer circinatum Pursh.). I used spatial autocorrelation analysis to quantify the patterns of distribution of specific forest floor properties. In addition, I applied a variance partitioning method based on redundancy analysis to determine whether selected spatial and local environmental variables contributed to the observed variation in forest floor properties. I found that eight out of ten properties exhibited some degree of spatial autocorrelation over a four to ten meter range, with pH, total N, available Ca and available K showing the strongest spatial dependence. All variables that exhibited spatial autocorrelation were anisotropically distributed. Variance partitioning results showed that environmental variables explained 3 1.7% of the total variation. Of this variation, 11.9% could be assessed by the spatial descriptors alone, indicating that these descriptors act as surrogates for unmeasured underlying processes. Local variables contributed 11.2% to the total explained variation, with forest floor type, presence of CWD, bulk density and the presence of slope at sampling locations contributing to the greatest extent. A large amount of variation (68.2%) remained unexplained. Ordination results indicate that while canopy extent of vine maple had little influence on the distribution of forest floor properties, moder forest floor, as determined by the presence of vine maple, was one of the primary local factors that affected variability. These results indicate that vine maple can impact forest floor variability, but that it may need to be present for longer time periods before its influence on forest floor properties becomes obvious. Keywords: forest floor properties, spatial autocorrelation, variance partitioning, vine maple