Metadata – Soil nutrient availability rather than spatial nutrient heterogeneity shapes the intraspecific response of root architectural, morphological, and mycorrhizal traits in Vaccinium myrtillus

This submission includes one primary dataset associated with the manuscript:

• File: dataset_root_traits_Vmyrtillus.csv
• Description: Root traits for all samples collected for the submitted manuscript
• Dimensions: 179 rows × 19 columns

---

Dataset Structure

Rows

Each row represents a root sample of Vaccinium myrtillus collected in the field according to the methods described in the manuscript.

Columns

• row_num: row number from 1 to 179

Plot Information

• Plot: Identifier for one of the ten plots (CFB036, CFB053, CFB102, CFB110, CFB124, CFB137, CFB140, CFB162, CFB164, CFB173)
• Subplot: Numbered 1 to 6; each plot has 3 subplots (not fixed – based on Vaccinium cover)
• Sample number: Sample identifiers from 1–180, with some additional samples (e.g., 22b, 24b, 24c) and four samples were missing (23, 24, 87, and 73).

Treatment Variables

• treatment_patch: Patch-level treatments (Con, Ca, K, N, P) in heterogeneous subplots; NA for control or homogeneous subplots
• treatment_subplot: Subplot-level treatment category (heterogen, control, homogen)

---

Trait Categories

Morphological Traits

• Average.diameter (mm) – Mean root diameter based on root scans
• Median.diameter, Maximum.diameter (mm) – Derived from root scan measurements
• Dry.weight (mg) – Dry weight of root samples
• Specific.root.length (mm/mg) – Total root length divided by dry mass
• Root.tissue.density (g/cm³) – Dry weight over root volume

Note on NA: Dry mass could not be determined for two samples (sample number 138 and 24c) due to loss during the weighing procedure. Consequently, traits requiring dry mass—specific root length (SRL) and root tissue density (RTD)—are unavailable for these samples and recorded as NA.

Mycorrhizal Traits

• Hypha.occurrence (0–1) – Proportion of root intersections colonized by intracellular hyphae
• Coiling.occurrence (0–1) – Proportion colonized by ericoid coils
• DSE.occurrence (0–1) – Proportion colonized by dark septate endophytes (DSE)

• ErM.occurrence (0–1) – Combined presence of coils/hyphae

Note on NA: Concerning mycorrhizal traits, three samples (sample numbers 20, 43, 179) were not measured due to a loss during the staining process. These are recorded as NA. 

• Hyphal.length (m/g) – Length of extraradical hyphae per gram of dry soil

  Note on NA: For Hyphal length, only half of the samples were measured due to time constrain (except CFB137 which was fully recorded). For all plots, the samples with most material collected were selected per subplot and patch (2 samples in control, 2 in homogeneous and 5 in heterogeneous). 

Physiological Trait

• Phosphatase.activity (µmol mg⁻¹ h⁻¹) – Enzymatic activity based on 4-MUB-phosphate cleavage.
  Note: Results should be reported as arbitrary units (a.u.) rather than in µmol mg⁻¹ h⁻¹, as the data are not suitable for comparisons among species or locations. However, they are valid for comparing treatments within sites.

Note on NA: Two samples could not be measured during the process due to a labelling issue (Sample numbers 75 and 96), these are reported as NA in the dataset. 

Chemical Trait

• RNC (Root Nitrogen Concentration) (mg/g) – Pooled per subplot/patch (70 values total), assigned to the first sample in each pooled group
  Averaged values used for multivariate trait analyses

Note on NA: To ensure a sufficient amount of material (minimum 4 mg), the samples were pooled within subplots for homogeneous and control subplots (resulting in one sample per control subplot and one sample per homogeneous subplot) and pooled per patch within the heterogeneous subplot (resulting in five samples per heterogeneous subplot). A total of 70 samples were analyzed for RNC.

Data layout note: The dataset may list multiple rows per subplot (control/homogeneous) or per patch (heterogeneous). Within each such group, the first row carries the pooled RNC value that represents that row and all subsequent rows in the same group. All subsequent rows in that group have RNC recorded as NA (i.e., they do not contain independent RNC measurements). Thus:

• Control/homogeneous: the first row for a given subplot holds the pooled RNC; following rows with the same subplot have RNC = NA.

• Heterogeneous: the first row for a given subplot + treatment_patch holds the pooled RNC; following rows with the same subplot + treatment_patch have RNC = NA.

For the multivariate analysis including all traits (PCA), only heterogeneous subplots were used. To align trait resolution with the RNC measurements (patch level), all other traits were averaged to the patch level prior to analysis.

Architectural Trait

• Branching.Intensity (cm⁻¹) – Number of first-order root tips per length of second-order root

Notes on NA: This method requires very clear scans, for this a selection of samples were analyzed for each treatment at the subplot and patch level. 49 samples were not used for this analysis and are recorded NA. 

Structure - R codes analysis

The code use to perform the analysis and the figures is presented in the following different scripts:

0_preparation_datasets_for_analysis.Rmd

• This code prepares the following datasets :
• dataset_all_root_2024.Rdata (see attached README text file with all columns and units)
• cv_trait_all.Rdata
• dataset_all_root_2024_grouped.Rdata
• Groot_dataset.Rdata

1_a_lmer_models_Figure_2.Rmd

• Test the effect of the fertilization treatments (control, homogeneous and heterogeneous) on all traits.
• The models used are lmer models : traits~ treatment_subplot + (1|Plot)
• The statistics will be used for the significance level in Figure 2
• Create table S4 and plots found in Annex Figure S4-6

1_b_response_ratio_Figure_2.Rmd

• Calculate log10 response ratio for all traits (homogeneous over control and heterogeneous over control)
• Figure 2 - with the response ratio for all traits and both treatments with the statistical significance calculated based on the models found in : 1_b_response_ratio_Figure_2

2_a_lmer_models_Figure_3.Rmd

• Test the effect of the treatments (heterogeneous versus homogeneous) on trait variation (CV).
• CV per treatment and per plot was calculated for all traits in : 0_preparation_datasets_for_analysis
• In particular, this script aims to :
  • extract the model estimates and significance to be used on Figure 3 in the article.
  • extract CV variation per trait for Table S5 and CV over all traits- table S6

2_b_response_ratio_Figure_3.Rmd

• Calculate log10 response ratio for CV of all traits (CV heterogeneous over CV homogeneous)
• Figure 3 - with the response ratio for all traits with the statistical significance calculated based on the models found in : 2_b_response_ratio_Figure_3

3_PCA_analysis_Correlation_Figure_4.Rmd

• Create a correlation table - Table S6
• PCA analysis with only RES root traits : Figure 4A
• PCA analysis with all traits measured - 1_2 axis : Figure 4B
• PCA analysis with all traits measured - 1_3 axis : Figure 9A
• PCA analysis with data at the interspecific level from - Figure 9B :

This last script uses data from GRoot :

Guerrero-Ramirez N, Mommer L, Freschet GT, Iversen CM, McCormack ML, Kattge J, Poorter H, van der Plas F, Bergmann J, Kuyper TW, York LM, Bruelheide H, Laughlin DC, Meier IC, Roumet C, Semchenko M, Sweeney CJ, van Ruijven J, Valverde-Barrantes OJ, Aubin I, Catford JA, Manning P, Martin A, Milla R, Minden V, Pausas JG, Smith SW, Soudzilovskaia NA, Ammer C, Butterfield B, Craine J, Cornelissen JHC, de Vries FT, Isaac ME, Kramer K, König C, Lamb EG, Onipchenko VG, Peñuelas J, Reich PB, Rillig MC, Sack L, Shipley B, Tedersoo L, Valladares F, van Bodegom P, Weigelt P, Wright JP, Weigelt A. (2021).
Global Root Traits (GRooT) Database. Global Ecology and Biogeography, 30(1), 25–37.
https://doi.org/10.1111/geb.13179

This data is not included in the dataset published here and to replicate the code, users need to download the GRoot dataset from the related repository and cite accordingly : https://groot-database.github.io/GRooT/