**4. Discussion**

The relationship between phylogenetic distance and ecological similarity is a key topic of community assembly [30]. Phylogenetic relationships influence the strength of species' interactions (competition or facilitation) [31]. Five of seven traits exhibit significant phylogenetic signals and the result showed phylogenetically conserved in Dinghushan plot. Previous studies have found that closely related species tend to more directly compete with one another [12]. However, examples are known of facilitation between congeneric species which are known to be closely related species [32]. In what circumstances facilitation or competition occurs among closely related species is largely unknown and in need of further study. To our knowledge, our study is the first to test, from the perspective of the competitor, whether the relationship of neighbor species to a dominant species relates its phylogenetic distance. Results demonstrated that across the 20 ha Dinghushan plot, more closely-related species and more distantly-related species were more likely to have individuals around the dominant species *C. chinensis* than expected by a null (random) model (Figure 4; Figure S1). Dominant species may provide facilitation for more closely-related relatives [32] resulting in more individuals from closely related species. At the same time, the distribution of 90% of the species (abundance ≥20) in the DHS plot was affected by topographic factors (e.g., slope, aspect, convexity, and elevation) [33]. It may promote competition in the same terrain, so more distantly-related species could make sufficient use of local resources that would differ from a dominant competitor thus supporting more individuals and species. Only the moderately-related species were neither facilitated by dominant species, nor differentiated enough to adapt to the competition surrounding our dominant competitors.

We also tested the role of dominant species in structuring neighbors at different successional stages in the same plot. Different successional stages led to different interactions among species (Figure 4) manifested by different community composition and distribution patterns. As illustrated in this study, the role of a dominant species on their surrounding species can change during community succession. The correlation between the DBH matrix of all *C. chinensis* and the geographical distance matrix from *C. chinensis* to other individuals were positively related in the whole plot by Mantel test (*p*-value = 0.032<0.05, Observation value = 0.029). That is, the larger DBH of *C. chinensis*, the more space it occupies, the greater the geographical distance between *C. chinensis* and other individuals. The mature patch of forest in the DHS plot experienced at least 400 years of succession after formal protection. Dominant species *C. chinensis* had large average sizes (average DBH 49 cm; largest tree in plot of 175 cm DBH) inferring a size advantage over other tree species. Large *C. chinensis* trees were primarily distributed along the ridge of the mountain (Figure 1) and dominated the top of the canopy, interferes with the relationship between the normal geographical distance matrix and the phylogenetic distance. This was the reason for the mantel test between geographic distance between *C. chinensis* and other trees, and phylogenetic distance between *C. chinensis* and other trees, give very low coefficients of correlations (observed value) and not very low *p* value. The spatial separation between plot with dominance of *C. chinensis* or *P. massoniana* can cause biased *p*-value as Guillot et Rousset's research [34]. And the mantel test revealed that, there were positive or negative correlations between geographic distance and phylogenetic distance in the successional forest or mature forest. Prior studies in this plot reported the distribution of 24 species were associated with niche differentiation [33], and with seedling mortality being related to patterns in the terrain [35]. Therefore, habitat selection (environmental filtering) can be suggested as another mechanism influencing the distribution of neighboring trees.

In the younger successional forest (60 years of age) derived from tree plantings of *P*. *massoniana*, *C. chinensis* is still the dominant species (abundance 2113, average DBH 25.2 cm). A null model test rejected the distribution of the most phylogenetically-related species (e.g., Group 1) around *C. chinensis.* Neighbors having greater phylogenetic distance were more likely to occur around *C. chinensis*(Figure 4a). A recent study suggested that intra-specific competitive exclusion and density-dependence appear to

play important roles in tree mortality in this subtropical forest [36]. The successional patch of forest had more individuals and higher densities than the mature patch of forest. Competitive exclusion and density-dependence here should be stronger resulting in larger percentages of neighbors occurring from more distant-related groups.

The effect of the dominant species on its neighbor species differed among successional and mature forests (Figures 3 and 4). This may explain why some studies found that more related species were aggregated [37,38], while others showed a repulsion of related species [16,39]. Target species may not have been common enough and/or phylogenetic relationships were not fully considered. We suggest that competitive exclusion or stable coexistence of neighboring species is determined partly by which successional phase that species occurs in. Spatial aggregation generally decreases with DBH, aggregation is weaker at larger diameter classes is largely due to self thinning [20], competitive associations were more frequently intraspecific than interspecific (Shen, 2013 [36]). In this study, individuals which DBH ≥40 cm are mainly distributed in mature forests, the frequency of neighbors around *C. chinesis* for the phylogenetically closely related species and most phylogenetically distantly related species (Group 6) were random. However, in the successional forests, young trees are most, the percentage of neighbor trees of *C. chinensis* was the most closely related species (i.e., Group 1), at all observed scales (Figure 4a; Figure S2).

In summary, we found that, as a dominant species, *Castanopsis chinensis* played an important role in structuring the species distributions and coexistence of neighbor species in a subtropical forest. Community successional stages and environmental filtering appeared to affect neighbor species relationships.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/1999-4907/11/3/352/s1, Figure S1: Distribution of neighbor trees within a 20 m radius to *Castanopsis chinensis* based on phylogenetic distance (six ordinal groups) in whole plot; Figure S2: Distribution of neighbor trees at six scales to *Castanopsis chinensis* based on phylogenetic distance (six ordinal groups) in whole plot; Figure S3: Distribution of neighbor trees at six scales to *Castanopsis chinensis* based on phylogenetic distance (six ordinal groups) in succession forest; Figure S4: Distribution of neighbor trees at six scales to *Castanopsis chinensis* based on phylogenetic distance (six ordinal groups) in mature forest; Table S1: Results from a test for phylogenetic signal in the functional trait data the 20 ha DHS plot, using the K statistic.

**Author Contributions:** W.Y., L.L. and J.L. designed the study, S.W. and L.L. performed analyses, S.W., H.C., Z.W., L.M. and X.O. collected data, S.W., L.L. and J.L. wrote the first draft of the manuscript. W.Y., S.E.N. and J.L. contributed substantially to revisions. All authors have read and agreed to the published version of the manuscript.

**Funding:** The study was funded by Strategic Priority Research Program of the Chinese Academy of Sciences (XDB31030000), the National Key R&D Program of China (grand No. 2017YFC0505802), National Natural Science Foundation of China (No. 41371078,No. 31870506, 31460155), Natural Science Foundation of Jiangsu Province (BK20181398), Key Laboratory of Ecology of Rare and Endangered Species and Environmental Protection (Guangxi Normal University), Ministry of Education, China and Chinese Forest Biodiversity Monitoring Network.

**Acknowledgments:** We appreciate Zhongliang Huang for the help of collecting data. We are grateful to Yue Bin for insightful suggestions on the revision of the MS. We thank many individuals who contributed to the field survey of the Dinghu plot. This plot is part of the Center for Tropical Forest Science, a global network of large-scale demographic tree plots. We would like to thank all of the reviewers and editors for reading the manuscript and providing useful feedback.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Review*
