Data and code from: Soil mediates the impact of functional traits on tree growth and survival across life stages in a temperate forest
Data files
Apr 09, 2026 version files 10.61 MB
-
code.R
10.35 KB
-
README.md
6 KB
-
RGR_adult.xlsx
2.06 MB
-
RGR_sapling.xlsx
2.76 MB
-
RGR_seedling.xlsx
120.74 KB
-
survival_adult.xlsx
2.36 MB
-
survival_sapling.xlsx
3.18 MB
-
survival_seedling.xlsx
115.14 KB
Abstract
Trait-based approaches are widely used to understand variation in plant demographic performance, yet their predictive power is often limited by context dependence. In particular, trait–demography relationships may shift across life stages and be modulated by local abiotic and biotic conditions. However, few studies have simultaneously examined how ontogeny, soil fertility, and neighborhood trait composition jointly shape the outcomes of trait–performance relationships. Here, we integrate long-term demographic data, functional traits, and soil variables from a 20-ha warm temperate forest dynamics plot to evaluate how three key traits, specific leaf area (SLA), leaf dry matter content (LDMC), and wood density (WD), affect tree growth and survival across seedling, sapling, and adult stages. We also examine how these effects, including neighborhood interactions for trait dissimilarity, are modified by soil fertility. Our study showed that soil fertility modulated trait effects on tree performance, which varied across life stages. Direct trait effects were detected primarily at the adult stage, where adult survival increased with WD but decreased with LDMC. In contrast, at the seedling and sapling stages, trait effects on survival were largely contingent on soil nutrients. Increasing soil organic matter, moisture, and available potassium intensified the negative effects of conservative traits (LDMC and WD) on seedling and sapling survival, while alleviating the negative effects of acquisitive traits (SLA). Seedlings exhibited lower survival when surrounded by heterospecific neighbors with dissimilar SLA. The negative effect of SLA dissimilarity among neighboring seedlings on seedling survival was amplified in soils rich in organic matter, moisture, and available potassium, whereas between seedlings and neighboring trees, it was alleviated under higher nitrogen and phosphorus availability. Our results demonstrate the importance of explicitly considering life stage, local abiotic conditions, and biotic neighborhood context. This multidimensional approach offers new insights into trait-based mechanisms underlying forest community dynamics across life stages.
This repository contains the datasets and R code used for growth- and survival-related analyses of seedlings, saplings, and adult trees.
Description of data
The repository includes six Excel data files and one R script:
Data files
- RGR_adult.xlsx: Dataset used for growth-related analyses of adults.
- RGR_sapling.xlsx: Dataset used for growth-related analyses of saplings.
- RGR_seedling.xlsx: Dataset used for growth-related analyses of seedlings.
- survival_adult.xlsx: Dataset used for survival-related analyses of adults.
- survival_sapling.xlsx: Dataset used for survival-related analyses of saplings.
- survival_seedling.xlsx: Dataset used for survival-related analyses of seedlings.
Code file
code.R: R script used to fit linear mixed-effects models and generalized linear mixed-effects models for growth and survival analyses across life stages.
Code description
The R script performs the following analyses:
- Reads the six Excel data files using read_xlsx().
- Fits growth models using linear mixed-effects models (lmer).
- Fits survival models using generalized linear mixed-effects models (glmer) with a binomial error distribution and logit link.
- Includes analyses for three life stages:
- seedling
- sapling
- adult
5. Includes two main groups of predictors: - FT: functional trait models using SLA, LDMC, and WD
- TD: trait dissimilarity models using SLA_TD, LDMC_TD, WD_TD and neighborhood competition variables
6. Uses species identity and quadrat as random effects.
Instructions for use
- Place the file code.R in the same directory as the six Excel data files.
- Set the R working directory to the folder containing these files.
Variable description for data files
RGR_sapling.xlsx, RGR_adult.xlsx: datasets used for growth-related analyses of saplings and adults
- sp: the 6-letter unique code for each species
- quadrat: the quadrat number to which the tree belongs
- dbh1: diameter at breast height of trees surveyed in 2010
- SLA: specific leaf area
- LDMC: leaf dry matter content
- WD: wood density
- SLA_TD: the SLA trait dissimilarity between the focal tree and neighboring trees
- LDMC_TD: the LDMC trait dissimilarity between the focal tree and neighboring trees
- WD_TD: the WD trait dissimilarity between the focal tree and neighboring trees
- PC1 and PC2: the first and second multivariate axes of principal component analysis (PCA) of soil fertility
- RGR: relative growth rate
- CON: the basal area at breast height of conspecific tree neighbors surrounding the focal tree
RGR_seedling.xlsx: dataset used for growth-related analyses of seedlings
- sp1: the unique code for each species
- quadrat: the quadrat number to which the seedling belongs
- height1: height of seedlings surveyed in 2014
- SLA: specific leaf area
- LDMC: leaf dry matter content
- SLA_TD_S: the SLA trait dissimilarity between the focal seedling and neighboring seedlings
- LDMC_TD_S: the LDMC trait dissimilarity between the focal seedling and neighboring seedlings
- SLA_TD_T: the SLA trait dissimilarity between the focal seedling and neighboring trees
- LDMC_TD_T: the LDMC trait dissimilarity between the focal seedling and neighboring trees
- PC1 and PC2: the first and second multivariate axes of principal component analysis (PCA) of soil fertility
- RGR: relative growth rate
- CON_S: the number of conspecific seedling neighbors surrounding the focal seedling
- CON_T: the basal area at breast height of conspecific tree neighbors surrounding the focal seedling
survival_sapling.xlsx, survival_adult.xlsx: dataset used for survival-related analyses of saplings and adults
- sp: the 6-letter unique code for each species
- quadrat: the quadrat number to which the tree belongs
- dbh1: diameter at breast height of trees surveyed in 2010
- SLA: specific leaf area
- LDMC: leaf dry matter content
- WD: wood density
- SLA_TD: the SLA trait dissimilarity between the focal tree and neighboring trees
- LDMC_TD: the LDMC trait dissimilarity between the focal tree and neighboring trees
- WD_TD: the WD trait dissimilarity between the focal tree and neighboring trees
- PC1 and PC2: the first and second multivariate axes of principal component analysis (PCA) of soil fertility
- status2: the survival status of the trees at the 2015 survey
- CON: the basal area at breast height of conspecific tree neighbors surrounding the focal tree
survival_seedling.xlsx: dataset used for survival-related analyses of seedlings
- sp1: the unique code for each species
- quadrat: the quadrat number to which the seedling belongs
- height1: height of seedlings surveyed in 2014
- SLA: specific leaf area
- LDMC: leaf dry matter content
- SLA_TD_S: the SLA trait dissimilarity between the focal seedling and neighboring seedlings
- LDMC_TD_S: the LDMC trait dissimilarity between focal seedling and neighboring seedlings
- SLA_TD_T: the SLA trait dissimilarity between focal seedling and neighboring trees
- LDMC_TD_T: the LDMC trait dissimilarity between focal seedling and neighboring trees
- PC1 and PC2: the first and second multivariate axes of principal component analysis (PCA) of soil fertility
- status: the survival status of the seedlings at the 2015 survey
- CON_S: the number of conspecific seedling neighbors surrounding the focal seedling
- CON_T: the basal area at breast height of conspecific tree neighbors surrounding the focal seedling
Code/Software:
Software used: The analyses were run in R version 4.4.1.
Packages used:
tidyr ‘1.3.2’
tidyverse ‘2.0.0’
dplyr ‘1.1.4’
readxl ‘1.4.3’
lme4 ‘1.1.35.1’
lmerTest ‘3.1.3’
performance ‘0.16.0’
stringr ‘1.6.0’
factoextra ‘1.0.7’
The script loads these packages at the beginning of the analysis. Users should install them before running the code.