Species richness mitigates but does not eliminate functional vulnerability in natural forests

Published Jul 24, 2026 on Dryad. https://doi.org/10.5061/dryad.jdfn2z3m4

Abstract

The dataset contains species abundance data from 6-hectare plots within five forest types, along with the corresponding species functional trait information. The functional trait data cover six key ecological functional attributes: specific leaf area (SLA), leaf dry matter content (LDMC), wood density (WD), leaf nitrogen content (LNC), leaf phosphorus content (LPC), and maximum potential height (H). Using these six functional traits, the functional entity framework was employed to assess functional redundancy, functional over-redundancy, and functional vulnerability for each forest type.

https://doi.org/10.5061/dryad.jdfn2z3m4

Description of the data and file structure

This project provides a dataset for assessing functional vulnerability across different forest types using the functional entity framework. The dataset includes raw functional trait data, species abundances for each forest type, and a discretized trait classification dataset. These data are used in R to calculate the relevant functional metrics.

This repository includes one R script:

Code.R: this script is organized into seven main sections:

1. Classification of Functional Traits (crude classification): The main purpose is to classify continuous functional traits (LDMC, SLA, WD, LNC, LPC, H) into several discrete levels based on predefined thresholds, which will be used for calculating relevant indices of functional entities (FEs) based on combinations of trait categories.

2. Calculate functional indices (FR, FOR, FV) based on trait classification: In different forest plots (JFL, JH, XLS, KNS, MLZ), the functional redundancy, functional over-redundancy, and functional vulnerability of communities are quantified based on functional entities (FEs).

3. Functional space: Plot the functional space based on functional entities.

4. Convex hull volume: Calculate the functional richness for each forest type.

5. Null model test: Determine whether the observed patterns of functional redundancy and vulnerability deviate from random expectations.

6. Quality assessment of the functional space: To assess whether differences among distinct FEs can be robustly captured within a multidimensional functional space.

7 Calculation of functional uniqueness: Calculate the functional uniqueness of species in each forest type based on functional trait values.

8. Functional traits space: Plot the functional trait space based on functional trait values.

9. The correlation between categorized-trait Gower distances and raw-trait Euclidean distances: Compute the correlation between Gower distances (computed on categorised traits) and euclidean distances (computed on raw traits).

The dataset mainly comprises the following components:

Files: baskets_species_weights_JFL.csv , baskets_species_weights_JH.csv, baskets_species_weights_KNS.csv, baskets_species_weights_MLZ.csv, baskets_species_weights_XLS.csv

Description: These datasets represent species abundance matrices for each 6-hectare plot.

File: classified_data.csv

Description: This dataset represents functional trait combination types generated based on the functional trait classification criteria.

File: Functional_trait_cat.xls

Description: The functional trait classification types are standardized as “ordered (O).” This dataset has explicitly encoded all selected functional traits as ordered variables (abbreviated as 'O') and stored them in this format. The data structure can be directly imported into the R environment and is compatible with analytical codes and functions designed for ordered variables.

File: Species_per_FE.xls

Description: Based on the functional entity framework, the distribution of species within functional entities.