Abstract:Treeline as one of the most important indicators of climate change has been extensively researched for the past years. Elucidation of past and present treeline formation can help us to understand how rising temperatures influence the treeline pattern and to predict treeline response to future climate change. On the other hand, global treeline positions have been considered to advance or retreat because of human disturbance and/or changes in local environmental factors combined with climate warming. Every natural treeline has a common convergence characteristic, namely, a limitation of tree height growth because of low temperature or other stress conditions, and this causes a gradual change to elfin or shrubby trees along an elevation gradient. Thus, elucidation of the functional differences between trees and alpine shrubs will facilitate an understanding of the alpine treeline. The spatial distribution pattern is an indicator of the selective adaption of a population to a specific environment, and it is considered to be a crucial character for describing the relative spatial location of the population. The extent to which plant individuals are aggregated can reflect the dispersal strategies of a species and determine how this species utilizes resources. The point process theory provides several statistics containing Riply's K-function and the paired correlation function [pcf or g(r)] to analyze point patterns at different scales. Investigation of the spatial point pattern of a species at the treeline ecotone will provide a valuable insight into ecological processes, for example, the way in which individuals occupy this habitat and the strategies whereby these individuals utilize resources. In the present study, we evaluated the spatial distribution of a Betula ermanii population growing in a 0.64 hm2 permanent sample plot at the treeline ecotone on the northern slope of Changbai Mountain, northeastern China. On the basis of the number of branches and tree height, we defined three life stages (adult, medium-sized, and sapling) in single-stemmed B. ermanii individuals and one shrubby shape in multi-stemmed B. ermanii individuals. We subsequently analyzed correlations between these different life stages and tree shape to verify whether life stage and tree shape were related to the spatial pattern. Results showed that 7.5% of the B. ermanii individuals exhibited a tree shape, whereas 32.0% exhibited a shrubby shape except the rest medium-sized and sapling ones. The critical limitation of tree height growth for B. ermanii is 1.5-3.0 m. Larger trees were rarely observed in the study area, indicating that this species is markedly constrained by height growth limitation at the treeline ecotone. We observed that adult and medium-sized trees were clustered in a small area with appropriate micro-topography and/or soil conditions for tree survival, for example, high-quality soil and concave grounds with seasonal snow drift. On the other hand, saplings and shrubby trees showed a weaker tendency for clustering and exerted weaker selective effects on the spatial pattern. Our results imply that the most suitable tree architecture for B. ermanii at the treeline ecotone is a shrubby shape and that an important period of life stage division may exist during the process of tree establishment and development. We further showed a higher degree of clustering among shrubby-shaped individuals than among tree-shaped individuals, implying that a multi-stemmed life shape with low stature is advantageous for survival of B. ermanii in the treeline ecotone. Taken together, our findings provide a valuable insight into the survival strategies of B. ermanii under local environmental conditions.