Contents lists available at ScienceDirect Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind Assessment of spatial stand structure of hemiboreal conifer dominated forests according to diferent levels of naturalness Eneli Põldveer ⁎ , Henn Korjus, Andres Kiviste, Ahto Kangur, Teele Paluots, Diana Laarmann Chair of Forest Management Planning and Wood Processing Technologies, Institute of Forestry and Rural Engineering, Estonian University of Life Sciences, Kreutzwaldi 5, Tartu 51006, Estonia ARTICLEINFO Keywords: Structural index Nearest neighbourhood relationship Structural pattern Forest management ABSTRACT Ecosystem legacies, refecting the management history traits, are visible in structural indicators of a forest stand. We assessed patterns of spatial forest structure by using individual tree indices based on the nearest-neigh- bourhood approach. Five diferent indices were quantifed – species mingling, deadwood mingling, deadwood dis- tribution, diameter differentiation and the uniform angle indices characterising the patterns that show the com- plexity and diversity of forests: the arrangement of tree dimensions, species and deadwood as well as tree positioning regularities. Managed Norway spruce (Picea abies) stands are remarkably more homogeneous in terms of occurrence and positioning of diferent tree species and dimensions when compared to unmanaged forests. Managed Scots pine (Pinus sylvestris) stands rarely show co-occurrence of living and dead trees and demonstrate lower dimensional variability than unmanaged stands. Trees are often quite randomly positioned; slightly regular positionings became evident in managed Norway spruce stands. Structural analysis confrm that large trees are particularly important in maintaining the structural diversity of forests. Forest management needs to integrate maintenance of important structural components and patterns into timber production for biodi- versity conservation and sustainable forestry. Natural forests are indispensable source for reference structures in forest ecosystems. 1. Introduction Forest ecosystems have signifcant heterogeneity in ecological quality, which depends on the level of anthropogenic infuence on them. For safeguarding ecological values of forest ecosystems, 25.6% of productive forestland is protected under diferent conservation regimes in Estonia (Estonian Environment Agency, 2017). However, all Esto- nian forests have been in some way shaped by long-term human impact (Reitalu et al., 2013) and the overall coherence and ecological quality of boreal forest ecosystems tend to decline (Kuuluvainen, 2002). The reasons are mainly related to common forest management practices such as invariable forest management regimes, homogenization of the forest structure and absence of natural forest remnants (Bengtsson et al., 2000; Jõgiste et al., 2017; Kuuluvainen, 2002; Lõhmus and Kraut, 2010; Tikkanen et al., 2006). The preservation of forest ecosystem complexity and diversity is often linked to forest naturalness (Eastaugh et al., 2013). Forest nat- uralness is a complex issue with many defnitions combining adaption to the changing environment, dynamics and disturbances, and in- cluding human infuence on forest ecosystems (Laarmann et al., 2009). The most common defnition is based on the comparison of a forest ecosystem’s current condition to its natural state (Winter, 2012). Forest stands in hemiboreal forests can span from highly artifcial to naturally vigorous forest ecosystems (Roberge et al., 2008, Ranius and Roberge, 2011). The degree of naturalness of forest ecosystems varies according to natural conditions and forest management practices used, e.g. Šaudyte et al. (2005) distinguish six classes of forest naturalness in Lithuania, ranging from the untouched virgin forest to an artifcial forest. For Estonian conditions, Korjus (2002) used four classes in Es- tonia according to European forest classifcation suggested by Wulf (1998): old-growth forests, natural forests, recovering forests and managed (semi-natural) forests. The indicators of naturalness include various tree ages, sizes and species, the presence of old and large (especially deciduous) trees and broadleaved species, the presence of deadwood and its volume and decay classes, recent and historical management practices as well as many other characteristics (Liira and Sepp, 2009; Lõhmus and Kraut, 2010). Proper understanding of the structural patterns of diferent forest ecosystems requires evaluating the spatial forest structure and quanti- fying tree composition (Maes et al., 2011). Classical forest inventory https://doi.org/10.1016/j.ecolind.2019.105944 Received 1 February 2019; Received in revised form 12 November 2019; Accepted 14 November 2019 ⁎ Corresponding author. E-mail address: eneli.poldveer@emu.ee (E. Põldveer). Ecological Indicators 110 (2020) 105944 1470-160X/ © 2019 Elsevier Ltd. All rights reserved. T