Spatio‐temporal structure of multi‐storied, multi‐aged interior Douglas fir (Pseudotsuga menziesii var. glauca) stands

Abstract

Summary 1. We used replicated, repeated‐measures data to examine the spatio‐temporal structure of multi‐storied, multi‐aged interior Douglas fir (Pseudotsuga menziesii var glauca (Mirb.) Franco) stands growing on dry sites under more than 50 years of fire protection. Along with the univariate and bivariate Ripley’s K and related functions, we used a random coefficients mixed model to investigate the variation in these functions over replicates and time. 2. The spatio‐temporal analyses revealed that trees greater than 1.3 m in height were clustered over space, and clustering was more evident for small trees (i.e. diameter at breast height (d.b.h.) ≤ 7.0 cm). 3. Bivariate functions indicated that small trees were spatially aggregated with large trees, indicating higher germination success and early survival near large trees. For these dry sites, moisture is more limiting than light and large trees provide moister microsites. 4. Dead trees were clustered, more commonly smaller in size and aggregated near large trees, indicating competition for moisture. 5. For the 16‐year period of the study, there was very little evidence that the spatio‐temporal patterns changed from a clustered to a more regular arrangement over time, unlike evidence from studies in other forest types. 6. Using the random coefficients mixed model approach, the majority of spatio‐temporal variation was due to differences among replicates, with little variation over time. 7. Under fire protection, interior Douglas fir stands on dry sites might be expected to have lower mortality but any increase in density will be limited by moisture availability. Since interior Douglas fir is moderately shade‐tolerant, this may result in an approach to a steady state of regeneration and death over the long term. 8. Synthesis. Spatio‐temporal analysis using replicated, repeated‐measures data, including a random coeffcieints mixed model approach, gave useful insights into mortality and recruitment in multi‐storied and multi‐aged stands. Similar patterns might be expected in other naturally occurring multi‐storied and multi‐aged stands. The use of replicated, repeated‐measures data rather than chronosequence data allows for the examination of true changes in spatial patterns over time.