Temperate forests on their warm and dry distribution limits are expected to be most vulnerable to reductions in water availability. This prediction is mostly based on studies assessing single forest functions, mainly growth. Water and nutrient cycling are functions that rely on tree roots and their symbiotic association with ectomycorrhizal (ECM) fungi. Trees can compensate for seasonal reductions in water availability by shifting root water-uptake (RWU) towards deeper soil layers, but ECM fungi mostly dwell in the topsoil, thus suffering from desiccation and compromising nutrient uptake. We hypothesised that drier sites should depict larger seasonal shifts in RWU, but at the expense of lower ECM fungal diversity and colonization of fine roots by ECM fungi. We selected three beech (Fagus sylvatica) forests in their warm distribution limit with contrasting geographic locations and mean annual precipitation: northern Atlantic (2500mm), intermediate transitional (1150mm), and southern Mediterranean (780mm). We collected soil, stem, and root samples in spring (wet) and summer (dry) to: (1) quantify fine-root density and colonization by ECM fungi, (2) infer RWU from isotopic composition of plant and soil water, and (3) characterize ECM communities through DNA-metabarcoding. Generalized and linear mixed models revealed that high topsoil moisture benefited ECM diversity, but higher diversity and ECM colonization did not imply larger contributions of the topsoil to RWU. The prevailing climate and abiotic conditions determined how ECM communities were structured, more than seasonal climatic variability. Across sites, communities differed in their functional diversity: ECM fungi with long hyphae, more vulnerable to water scarcity, dominated at the southernmost site, where water availability was the highest. Our results suggest that, in a climate change scenario, increasing drought might not compromise RWU, but it would still be detrimental for ECM communities, compromising key ecosystem services such as nutrient cycling and productivity.

Dataset DOI: 10.5061/dryad.0gb5mkmdg

Description of the data and file structure

This dataset includes ten datafiles and one script used in this publication:

Title: Ectomycorrhizal fungi and root water uptake respond independently to water availability

Authors list: Asun Rodríguez-Uña, David Moreno-Mateos, Silvia Matesanz, Lisa Wingate, Adrià Barbeta, Javier Porras-Gómez , Teresa E. Gimeno

Accepted for publication in Oikos on the 28th June 2025

Corresponding author: Asun Rodríguez-Uña (ar2190@cam.ac.uk, asun.rodez@gmail.com)

Data corresponds to soil, stem and root samples collected in spring (wet) and summer (dry) from three beech forests in their warm distribution limit with contrasting geographic locations and mean annual precipitation. Data analyses enable to (1) quantify fine-root density and colonization by ECM fungi, (2) infer RWU from isotopic composition of plant and soil water, and (3) characterize ECM communities through DNA-metabarcoding.

Files and variables

File: wi_sample_list.csv

This file contains the list of samples collected for analyses of water isotopic composition and volumetric water content. In each sampling campaign, three soil samples were collected at different depths in each selected tree, plus one lignified branch segment. One groundwater sample was also collected on each study site in each sampling campaign.

Sample -> character variable. Unique identifier of each sample
Site -> factor variable with three levels. Name of the site (Artikutza, Northern site; Iturrieta, Intermediate site; Diustes, Southern site)
Tree -> character variable. Unique identifier of the sampled tree for wood samples (see variable “Type”) or the tree next to the sampling point for soil samples (see variable “Type”).
Campaign -> factor variable with two levels (Jun and Aug). Sampling campaign (spring/June 2020 or summer/August 2020)
Date -> numerical variable. Sampling collection date in YYYYMMDD
Type -> factor variable with three levels (soil, wood and water). Type of sampled material.
Description -> character variable. Sampling depth for soil samples, type of material collected (branch vs. wood core) for wood samples and location of source for water samples.

File: water_extraction_data.csv

This file contains the list of samples collected for analysing relative water content. In each sampling campaign, three soil samples were collected at different depths in each selected tree, plus one lignified branch segment.

Sample -> character variable. Unique identifier of each sample
Type -> factor variable with two levels (SOIL TOP, STEM TISSUE). Source of each sample
Weight_sample_soil -> numerical variable. Weight of the soil sample transferred to the extraction vessel in g.
Fresh_sample_soil -> numerical variable. Fresh weight of the sample and the extraction vessel in g.
Water_volume_in_sample -> numerical variable. Weight difference between fresh and dry samples corresponding to the volume (or mass) of water in g.

File: soilProperties.csv

This file contains soil physical and chemical properties measured on three locations per study site. At each location, three samples were collected at different depths.