The effect of neighbor species’ phylogenetic and trait difference on tree growth in subtropical forests

# Manuscript effect of neighbor on individual tree growth This study aimed to investigate the relative importance of biotic interactions on tree growth by examining several metrics, including hierarchical and absolute trait differences of focal trees to neighbors, neighborhood crowding index, functional community structure, and phylogenetic distance. ## Description of the data and file structure The dataset includes information related to the **site** and the **plot** where the tree individuals were collected, as well as their **code** in the permanent plots. Additionally, the **species'** individual trees were identified, as well as their growth **class** and annual growth rate (**AGR**). We utilized plant functional traits, which are indicators of plant functional strategies and are expected to be related to individual tree growth and performance. These traits included height (**H**, m), leaf area (**LA**, cm²), specific leaf area (**SLA**, cm².g-1), leaf dry matter content (**LDMC**, mg.g-1), and wood density (**WD**, g.cm³). The neighborhood variables considered in the study were: a) Neighborhood crowding index (**NCI**): based on the steam size and spatial distance of neighboring trees within a fixed radius of 5 m; b) neighborhood hierarchical mean trait difference (**LAh, SLAh, LDMCh, WDh, Hh**) and neighborhood absolute mean trait difference (**LAs, LDMCs, WDs, Hs**) : dissimilarity of functional traits between focal trees and their neighbors. These indices are associated with competitive abilities' trait-mediated ranks and niche differentiation, respectively; c) community weighted mean (**CWM; CWM_LA, CWM_SLA, CWM_LDMC, CWM_WD, CWM_H**): mean trait values of each community of tree neighbors accounting for species relative abundance in each community; d) functional diversity (**FD**): Rao weighted by species relative abundance in each community; e) mean phylogenetic difference (**mean_PD**): the mean phylogenetic distance between the focal individual to all neighbor trees within the radius To assess the effects of neighborhood interactions and community structure on tree growth, we performed quantile regression models (QR) with AGR logarithmic as the response variable and non-correlated neighborhood variables as predictors, using the 'rq' function in the R package quantreg. We ran the QR using the equation log AGR~ mean_PD+ SLAh+ WDs+ LAh+ LDMCh+ LDMCs+ Hs+ NCI+ CWM_LA+ CWM_SLA+ CWM_LDMC+ CWM_H+ FD and using three quantiles in line with our sampling design: the 25th percentile (QR 25%) for focal trees with slower growth, the 50th percentile (QR 50%) for trees with intermediary growth, and the 75th percentile (QR 75%) for focal trees with faster growth. The analysis is described in the script. ## Sharing/Access information none ## Code/Software

# 邻体对单木生长的影响研究手稿
本研究旨在通过多项指标探究生物交互作用对树木生长的相对重要性，涵盖目标木与邻体的层级性状差异、绝对性状差异、邻体拥挤度指数、功能群落结构以及系统发育距离。

## 数据与文件结构说明
本数据集包含树木个体采集的**样地（site）**与**样方（plot）**信息，以及其在永久样方中的**编号（code）**。此外，还完成了目标木的物种鉴定，并记录了其生长**等级（class）**与年生长率（annual growth rate, AGR）。
我们采用了植物功能性状——这类指标可反映植物功能策略，且被证实与单木生长及个体表现密切相关——包括：树高（H，单位：m）、叶面积（LA，单位：cm²）、比叶面积（SLA，单位：cm²·g⁻¹）、叶干物质含量（LDMC，单位：mg·g⁻¹）与木材密度（WD，单位：g·cm⁻³）。
本研究考量的邻体变量如下：
a) 邻体拥挤度指数（Neighborhood Crowding Index, NCI）：基于固定半径5米范围内邻体树木的茎干大小与空间距离计算得到；
b) 邻体层级平均性状差异（LAh、SLAh、LDMCh、WDh、Hh）与邻体绝对平均性状差异（LAs、LDMCs、WDs、Hs）：分别表征目标木与邻体间功能性状的相异程度，前者与基于性状等级的竞争能力相关，后者则与生境生态位分化相关；
c) 群落加权均值（Community Weighted Mean, CWM；CWM_LA、CWM_SLA、CWM_LDMC、CWM_WD、CWM_H）：以各群落中邻体树木的物种相对丰度为权重计算得到的平均性状值；
d) 功能多样性（Functional Diversity, FD）：以各群落中物种相对丰度为权重的Rao指数；
e) 平均系统发育差异（mean_PD）：目标个体与半径范围内所有邻体树木间的平均系统发育距离。

为评估邻体交互与群落结构对树木生长的影响，我们以对数转换后的年生长率为响应变量，以无多重共线性的邻体变量为预测变量，采用R包`quantreg`中的`rq`函数构建分位数回归模型（Quantile Regression, QR）。本研究的分位数回归基于以下公式：
`log(AGR) ~ mean_PD + SLAh + WDs + LAh + LDMCh + LDMCs + Hs + NCI + CWM_LA + CWM_SLA + CWM_LDMC + CWM_H + FD`
并依据采样设计设置了三个分位数：25%分位数（QR 25%）对应生长较慢的目标木，50%分位数（QR 50%）对应生长中等的目标木，75%分位数（QR 75%）对应生长较快的目标木。具体分析流程详见脚本文件。

## 共享与获取信息
无

## 代码与软件