Quantification of small-scale spatial patterns in alpine–treeline ecotones

Abstract

Alpine treeline ecotones, when viewed up close, display considerable variation in spatial patterns, which have been associated with different responses to climate change. Two important dimensions of treeline-ecotone spatial patterns are the abruptness of the change in tree height (“abrupt” vs. “gradual”) and the change in canopy cover (“discrete” vs. “diffuse”) when moving from closed forest to treeless alpine vegetation. These dimensions are suited to classify treeline ecotones into different types of patterns, but this is typically done intuitively without explicitly stated criteria, and patterns are not quantified. Consistent, robust metrics allowing comparisons between sites are lacking. We suggest several metrics to quantify abruptness and discreteness of treeline ecotones and describe how to derive these metrics from point-pattern data of tree positions and sizes, and from high-resolution treecover data. We developed these based on field data from the Spanish Pyrenees and an extensive dataset of treeline patterns created by the individual-based Spatial Treeline-Ecotone Model (STEM). We quantified the abruptness of a treeline by the largest change in canopy height, determined in 5-m bands, between the top of the ecotone (i.e., alpine vegetation) and the first band where canopy height exceeds 3 m. We quantified the discreteness by the steepness of a logistic function fitted to tree cover. Band widths and cut-off values were optimised for our data. Although they can be flexibly adjusted to specific case studies, standard settings are recommended to assure comparability. Our results indicate that the “discreteness” metric provides a satisfactory quantification of this pattern dimension within the dataset used here, whereas the “abruptness” pattern dimension turned out to be more difficult to capture. The metrics developed here may provide field researchers with a tool to compare their field sites in a standardised way, and potentially promote synthesis on treeline data and dynamics on a global scale.