Using novel metrics to assess biogeographic patterns of abrupt treelines in relation to abiotic influences

© The Author(s) 2015 Alpine treeline ecotones display spatial variability in a range of features that often reflect underlying abiotic variation and its control on the processes that form and maintain treelines. In this study, we explore treeline pattern–environment relationships for continuous, abrupt Nothofagus treelines in New Zealand across a seven degree latitudinal gradient (circa 39–46° S). Our main aims were: 1) to develop a set of metrics for characterising spatial variation in abrupt treelines at a landscape scale; 2) to relate these metrics to underlying abiotic variation in order to determine the relative roles of climate, substrate, topography and disturbance in structuring Nothofagus treeline patterns; and 3) to develop a possible biogeographic typology of Nothofagus treelines. A GIS-based approach was used to develop seven metrics describing different facets of landscape-scale treeline pattern. Regression modelling and variance partitioning were used to explore relationships among treeline pattern metrics and abiotic variation. Cluster analysis was used to characterise emergent treeline types and GIS was used to map their biogeographic distributions. The individual treeline metrics characterised unique patterns of treeline variation across New Zealand and, upon clustering, resulted in seven distinctive treeline types. Nothofagus treeline patterns in New Zealand are strongly structured by environmental gradients, with about half the landscape-scale variation in treelines being structured by patterns of abiotic variation. Gradients of climatic and disturbance were most influential in explaining landscape-scale variation of individual treeline metrics and of multivariate treeline patterns. The presented, metric-based approach offers a means to develop a comprehensive picture of continuous, landscape-scale treeline variation, bridging an existing research gap between studies at site and global scales. Our approach can enable the development biogeographic treeline typologies that could facilitate the comparison of treeline patterns across large areas and provide a basis for the generation of new hypotheses regarding treeline formation and dynamics