Optimizing Forest Spatial Structure with Neighborhood-Based Indices: Four Case Studies from Northeast China

Abstract

The fine-scale spatial patterns of trees and their interactions are of paramount importance for controlling the structure and function of forest ecosystems; however, few management techniques can be employed to adjust the structural characteristics of uneven-aged mixed forests. This research provides an accurate, efficient, and impersonal comprehensive thinning index (P-index) for selecting candidate harvesting trees; the index was proposed by weighting the commonly used quantitative indices with respect to stand fine-scale structures, competition status, tree vigor, and tree stability. The applications of the proposed P-index in evaluating and simulating the process of thinning operations were examined using four 1-ha mapped plots with different forest types, namely, natural secondary forest, natural pine-broadleaved mixed forest, natural larch-birch mixed forest, and natural oak forest, which were widely distributed across the Heilongjiang Province in Northeast China. The results indicated that the proposed P-index could effectively affect the structural differentiations between different forest types and alternative thinning intensities. The marginal benefits of alternative thinning intensities on the integrated forest structure indicated that removing 10% of the trees from the plots might be the optimal thinning intensity from the perspective of optimizing stand structure, in which the P-index values could be increased by approximately 5%–11% for the four tested plots. The main conclusion from this paper was that the proposed P-index could be used as a quantitative tool to manage uneven-aged mixed forests.

Optimizing Stand Spatial Structure by Using Neighborhood-Based Quantitative Indicators: A Case of Boreal Forests

“…The developments of horizontal distribution patterns are usually shifted from aggregation to random with forest succession, maintaining the random pattern consistently if the forest community could reach its climax status. Our results indicated that the trees grown in NLF, SEF and KBF stands were all randomly distributed (Table 3), which were perfect in line with the results of Wan et al (2019) and Dong et al (2020); however, slight aggregated distribution was observed for NBF, mainly due to a certain percentage of trees were sprouting regimentation in the stands. Regardless of the distribution pattern, the tree-level harvest 20 optimization could produce significant randomly distributions immediately.…”

“…Equation 13 limits the proportion of assigned harvest trees to a range from 𝛼 𝑙 to 𝛼 𝑢 . According to recent published works (Macdonald et al, 2004;Li et al, 2014;Dong et al, 2020) and management practices (State Forestry Bureau, 2016), 𝛼 𝑙 and 𝛼 𝑢 are assumed to be 10% and 40%, respectively. Equation 14 requires that the decision variables are binary.…”

“…Therefore, the spatial aspects regarding the location of individual trees has often been recognized as more important than the non-spatial aspects (Dong et al, 2020).…”

“…A set of tree-level spatial structural measurements have been proposed as important and reflective of the nearest neighbor relations between a reference tree and its n nearest neighbor trees (called a "structural unit") (Figure 1; Hui, 2003;Davies and Pommerening, 2008;Li et al, 2014;Dong et al, 2020). These measurements include the uniform angle (W), mingling index (M), dominance index (D), and crowdedness index (C).…”

“…Thus, for the purposes of improving forest health and projected outcomes from the management of natural forests, integrating tree-level structural indices in the harvest selection process has drawn much attention. For instance, Pastorella and Paletto (2013), Ye et al (2018) and Dong et al (2020) all simulated selective thinning processes using a comprehensive index method that was based on frequently used spatial and non-spatial structure indices. Li et al (2017) optimized the structure of a stand using bivariate probability distributions of three spatial parameters (i.e., U-, M-, and D-indices).…”