Data from: Warming leads to biomass increase, leaf nitrogen decline, and community turnover in Mediterranean Nardus stricta grasslands
Data files
May 09, 2026 version files 1.06 MB
Abstract
Nardus stricta L. grasslands are designated a priority habitat under the European Union Habitats Directive, reflecting their role in biodiversity conservation and soil and water regulation. However, these grasslands are listed as vulnerable in the European Red List of Habitats, primarily due to threats associated with global change. Despite their conservation importance, subalpine Nardus grasslands in Iberian mountains remain poorly studied, and responses to climate warming are largely unknown. We experimentally tested the effects of warming in five semi-natural N. stricta grasslands (1545–1875 m a.s.l.) in Serra da Estrela, Portugal, located at the transition between Mediterranean and Atlantic climates. In 2020, five paired open-top chambers (OTCs) and control plots were established at each site. Community composition and aboveground biomass were monitored annually from 2021 to 2023, and leaf nutrient content of N. stricta was analysed in 2022. Warming increased mean air temperature by ~1.9 °C on average, reduced freezing exposure, and enhanced thermal accumulation during growing season. These changes were associated with higher aboveground biomass and increased species richness, accompanied by moderate turnover. The dominant grass N. stricta increased in cover across site × year combinations, whereas the snow-associated endemic Festuca henriquesii, restricted to the high-elevation site COB, declined under warming. Warming also reduced leaf nitrogen concentration in N. stricta, altering the nutrient balance of the dominant grass and potentially leading to cascading ecological effects. Responses were site-dependent, with weaker biomass increases at drier, lower-elevation sites, suggesting baseline water availability modulated warming effects. Our results demonstrate a consistent thermal signal but context-dependent ecological responses. Although warming enhanced productivity and richness, it also altered species composition and reduced leaf nitrogen content, suggesting early restructuring of Iberian subalpine grasslands under climate change. These findings highlight the importance of incorporating site-specific environmental context, particularly soil moisture, into conservation strategies for Mediterranean mountain ecosystems.
Dataset DOI: 10.5061/dryad.sf7m0cgnt
Readme file - Data from
Correia M, Silva A and Rodríguez-Echeverría, S (2026) Warming leads to biomass increase, leaf nitrogen decline, and community turnover in Mediterranean Nardus stricta grasslands. Ecology and Evolution.
General Metadata description
This dataset supports a warming experiment conducted in five semi-natural Nardus stricta grasslands in Serra da Estrela, Portugal, spanning elevational (1546–1875 m a.s.l.) and edaphic gradients (Table 1, main document).
Sampling design: 5 Control and 5 Warming plots per site per year.
Years covered: 2021, 2022, and 2023 (separate sheets).
Note: CUM 2023 is excluded from all datasets due to extensive damage to OTCs and monitoring equipment.
Common Variables (all datasets):
-
Site — study site identifier: ALX, COB, CUM, LAC, NSA
-
Treatment — Warming (open-top chamber) or Control (fenced ambient plot)
-
Year — sampling year ( 2021, 2022, 2023)
-
Date — calendar date (YYYY-MM-DD), where applicable
-
Plot — unique plot code combining site and treatment (e.g., ALX_cont1, ALX_OTC1), where applicable
Dataset A — Microclimate and moisture (2020–2023)
Dataset_A_microclimate_moisture.xlsx
Continuous air temperature and soil moisture recorded by TMS-4 dataloggers (TOMST) at 15-minute intervals, aggregated to daily values, across five sites and two treatments. Records before July 2021 are from HOBO MX2301 dataloggers (2-hour intervals). One datalogger per site × treatment combination.
Sheet A1 — Daily Air Temperature
Air temperature recorded at 15 cm height.
-
MeanT — daily mean air temperature (°C) (**used in the paper)
-
MinT — daily minimum air temperature (°C)
-
MaxT — daily maximum air temperature (°C)
Sheet A2 — Soil moisture (Volumetric Water Content)
Soil moisture recorded at 6 cm depth, daily, calculated using the standard peat soil calibration curve (Wild et al. 2019). Available from July 2021 onwards.
- daily mean volumetric water content (%)
Sheet A3 — GDD_FDD
Seasonal thermal accumulation indices calculated from daily mean air temperature for the meteorological growing season (March–July).
One aggregated value per site × treatment × year combination (n = 28).
-
Growing_Degree_Days_GDD — cumulative daily mean temperatures exceeding 5 °C (°C·days)
-
Freezing_Degree_Days_FDD — cumulative daily mean temperatures below 0 °C (°C·days)
-
Number_of_days(n) — number of days contributing to the seasonal calculation
Dataset B — Plant Community, cover, diversity and Biomass (2021–2023)
Dataset_B_Plant_community_cover_diversity_Biomass_(2021–2023).xlsx
Plot-level species percent cover and aboveground biomass measured annually at peak growing season (early July).
Sheets B1, B2, B3 — Species Cover Matrices (2021, 2022, 2023)
Each sheet contains one row per experimental plot and one column per species, with values representing percent cover (%). Used to calculate alpha and beta diversity metrics.
Data type: species*plot_treatments_site matrix; Quantitative (% cover per species per plot).
Sampling design: 5 Control and 5 Warming plots per site per year.
Years covered: 2021, 2022, and 2023 (separate sheets).
Species identification: to species level wherever possible
Standardization: Percent cover estimated using the same method and observers across years for consistency.
Sheet B4 — Diversity Indices
The indices were calculated based on previous species cover matrices (2021, 2022, 2023)
One row per plot per year containing derived community metrics.
-
Plant_N_species — species richness (number of species per plot)
-
Plant_Shannon — Shannon diversity index
-
Plant_Simpson — Simpson diversity index
-
Plant_InvSimpson — Inverse Simpson index
-
Plant_Evenness — Pielou's evenness
Note: one cell contains an NA (not applicable) value for Plant_evenness because this metric could not be calculated for the specific year × site × plot × treatment combination (2022 × NSA × control1), where only one species was recorded. One cell is blank indicating inapplicable or missing data.
The Species Cover Matrices (2021, 2022, 2023) were also used to calculate:
-
Beta diversity and species turnover (Suplementary Info Table S11): Quantification of beta diversity (variation in species composition across time (2021–2023) and experimental treatments, and sites.
-
Species-level responses to warming (Suplementary Info Table S12): Calculation of Δ cover (difference in mean percent cover between Warming and Control) per species, per site across years; Standardized effect sizes (Hedges’ g) were estimated to quantify the magnitude and direction of species responses to experimental warming across sites and years when possible (sampling size).
Sheet B5 — Plant Biomass
Biomass was estimated using the Robel pole method and converted to g/m2 via site-specific height–biomass regressions.
One row per plot per year containing aboveground biomass estimates.
- Plant_biomass — aboveground plant biomass (g/m2)
Dataset C — Leaf Nutrient Analysis (2022)
Dataset_C_Leaf_nutrients_WarmingExp_data_2026.xlsx
Nardus stricta green leaves collected in July 2022 from five randomly selected plants per plot, pooled into composite samples, oven-dried (60 °C, 48h), and finely ground. Carbon and nitrogen were measured by elemental analysis (CN 802, VELP Scientifica); phosphorus by ICP-OES after aqua regia digestion.
Leaves were collected in spring 2022 from plants within each plot.
Each row corresponds to one composite sample per plot (n = 50; 5 sites × 2 treatments × 5 replicates).
Each sample was analyzed in triplicate.
-
C — leaf carbon concentration (% dry weight)
-
N — leaf nitrogen concentration (% dry weight)
-
C_N — carbon to nitrogen mass ratio (dimensionless)
-
P — leaf phosphorus concentration (mg kg -1 dry weight)
Study Design & Site Characterization
Five semi-natural Nardus stricta grasslands in Serra da Estrela, Portugal (1546–1875 m a.s.l.) were monitored from October 2020 to September 2023. All sites are characterized by Nardus stricta as the dominant species and are subject to traditional extensive summer grazing by sheep, with no mowing, fertilization, or other management practices applied. Sites spanned two partially independent environmental gradients: an elevational-thermal gradient separating three high-altitude plateau sites (> 1800 m; ALX, COB, CUM; MAT ≈ 7.5 °C) from two lower, warmer sites (NSA and LAC; MAT 8.2 °C and 9.3 °C, respectively), and an edaphic gradient of soil water content and organic matter, grouping sites into high (ALX, NSA), intermediate (COB, CUM), and low (LAC) moisture and organic content. Soils across all sites are classified as Umbrisols, characterized by a coarse texture, low pH (4.3–4.9), and high organic matter content. Site-level soil properties — pH, gravimetric soil water content, soil organic matter, nitrogen content, and texture — were measured from samples collected in November 2020 and used to characterize the edaphic gradient across sites. ALX and COB present wet, organic, sandy-loam soils; CUM a moderately moist loam; NSA a waterlogged depression with high organic content; and LAC the warmest and driest conditions with the lowest soil organic matter.
At each site, five paired open-top chambers (OTCs; warming) and fenced control plots were established in October 2020; 50 experimental units total.
Microclimate Monitoring
Air temperature (15 cm height), soil temperature, and volumetric water content (VWC %, 6 cm depth) were recorded continuously at 15-minute intervals using TMS-4 dataloggers (TOMST). Data were aggregated to daily means. Growing Degree Days (GDD) and Freezing Degree Days (FDD) were calculated for the meteorological growing season (March–July; base threshold 5 °C).
Plant Sampling
Species cover (%), plant biomass (Robel pole method, g/m2), and species richness were assessed annually at peak growing season (July 2021–2023).
Leaf Traits
Nardus stricta leaves were collected in July 2022, dried (60 °C, 48 h), and ground. Carbon and nitrogen were measured by elemental analysis (CN 802, VELP Scientifica); phosphorus by ICP-OES following aqua regia digestion.
Data Analysis
All models included site and treatment (warming vs. control) as fixed effects. Interaction terms were retained when they improved model fit based on AIC. For variables measured repeatedly across years — microclimate, plant community metrics, and biomass — plot identity nested within site was included as a random intercept to account for the hierarchical structure of the experiment and repeated measurements over time. Species-specific responses to warming were quantified as the difference in mean cover between warming and control plots (Δ % cover = warming − control) within each site × year combination, and classified as increased, decreased, gained, or lost under warming.