Data from: Uncovering biodiversity deficits: experimental evidence for dispersal limitation of woody species in Mediterranean forest patches
Data files
Oct 16, 2025 version files 9.76 MB
-
code_1_sowing_demographic_analysis.R
89.82 KB
-
code_2_temporal_trends.R
19.10 KB
-
code_3_planting.R
9.12 KB
-
code_4_soil_analyses.R
19.13 KB
-
code_5_soil_moisture_analyses.R
19.33 KB
-
code_6_vegetation_cover.R
15.03 KB
-
planting_experiment.csv
19.77 KB
-
README.md
8.96 KB
-
soil_moisture_vwc_values.csv
211.46 KB
-
soil_physicochemical.csv
6.94 KB
-
sowing_experiment.csv
9.34 MB
-
vegetation_cover.csv
9.04 KB
Abstract
Deforestation and habitat disturbance are major drivers of local species extinctions within remnant forest patches. In plants, seed dispersal plays a key role in recolonising patches following disturbance; however, habitat fragmentation may hinder seed arrival to suitable but unoccupied patches, resulting in dispersal limitation and prolonged biodiversity deficits.
We assessed dispersal limitation in Mediterranean forest patches by experimentally introducing plant species from the regional pool. To do so, we sowed 61,000 seeds from ten woody species and compared their establishment between occupied and unoccupied patches, assessing whether absences were driven by habitat unsuitability or dispersal limitation. We included five dry- and five fleshy-fruited species to test whether dispersal mechanisms influenced dispersal limitation. Additionally, we evaluated major biotic (vertebrate granivores and herbivores) and abiotic (soil properties and moisture) factors that could act as ecological filters in unoccupied areas. Finally, we planted saplings of four species to assess survival at later life stages and test the effectiveness of planting as a restoration strategy.
On average, both dry- and fleshy-fruited species exhibited similar establishment rates in occupied and unoccupied patches, indicating that species absences were primarily due to dispersal limitation rather than to habitat unsuitability. Vertebrate antagonists had similar effects across patch occupancy levels, overall reducing establishment. Soil properties and moisture did not differ in ways that would limit establishment in unoccupied patches. Finally, planted saplings had similar or higher survival in unoccupied patches, indicating habitat suitability also in later life stages.
Our experiments uncover diversity deficits of woody plants in forest patches driven by dispersal limitation, irrespective of the species’ dispersal mechanisms. In general, neither biotic nor abiotic filters significantly constrained the establishment of introduced species in unoccupied patches.
Synthesis and applications: Mediterranean forest patches in highly altered landscapes may harbour greater woody plant diversity than currently observed, suggesting ongoing dispersal limitation. Assisted colonization through plant introductions can effectively restore this missing plant diversity.
Dataset DOI: 10.5061/dryad.t1g1jwtg3
Description of the data and file structure
Datasets
1. "sowing_experiment.csv"
2. "planting_experiment.csv"
3. "soil_physicochemical.csv"
4. "soil_moisture_vwc_values.csv"
5. "vegetation_cover.csv"
Note: In all datasets, values marked as "NA" indicate that the data is not available or applicable, or that a measurement was not taken, depending on the context of each variable.
Files and variables
File: sowing_experiment.csv
Description: Data on the sown seeds from the seed sowing experiment, used to evaluate the habitat suitability of occupied and unoccupied forest patches. Dataset contains 61,000 rows, each one containing the information of a sown seed (n = 61,000 sown seeds). The dataset also includes information on species identity, sowing site, protection treatment (protected or unprotected from vertebrate granivores and herbivores), and outcomes regarding germination, survival, and establishment.
Variables
- species: study plant species (n=10 species).
- fruit_type: the type of fruit of the study species, which represented two contrasting dispersal mechanisms (two levels: 'dry' and 'fleshy').
- site: study forest patch (ten sites).
- site_type: whether the study species was present or absent in the site (two levels: 'absent' and 'present').
- treatment: the level of protection against vertebrate granivores and herbivores (two levels: 'protected' and 'unprotected').
- year: the year when seeds were sown, since we replicated the sowing experiment in two years for a subset of species.
- microhabitat: microhabitat type where seeds were sowed (two levels: 'tree' and 'shrub'). Compared with the manuscript, 'tree' corresponds with 'open' microhabitat and 'shrub' corresponds with 'shrub' microhabitat.
- station: the experimental plot where seeds were sowed.
- source: the site from which sown seeds were collected.
- point: seeds were sowed uniformly distributed in 25 positions of a grid (5 rows × 5 columns) and this variable specifies the specific position in the grid where the seed was sown.
- seed: for four dry-fruited species, we sowed 4 seeds in each point.
- seedling_monitoring: for those species that we sowed more than one seed per point, we counted all germinated seeds but only monitored the survival of the first emerged seedling per position (the others were cut with scissors to avoid competition). This variable specifies whether the survival of the seedling was monitored (1) or not (0).
- germ: whether the seed germinated (1) or not (0).
- seedling_surv: whether the emerged plant survived to the first summer (1) or not (0).
- sapling_surv_sd: whether the seedling survived to the second summer (1) or not (0).
- germ_cohort: whether the seed germinated on the first monitoring year (1) or in the second monitoring year (2)
- station_id: unique code of the sowing station, conformed by the union of the following variables: 'species'-'site'-'station'-'treatment'-'year'
File: planting_experiment.csv
Description: Dataset with the data of the planting experiment. Dataset contains 299 rows, each one containing the information of a planted sapling (n = 299 planted sapling).
Variables
- species: study plant species (n=10 species).
- fruit_type: the type of fruit of the study species, which represented two contrasting dispersal mechanisms (two levels: 'dry' and 'fleshy') .
- site: study forest patch (ten sites).
- site_type: whether the study species was present or absent in the site (two levels: 'absent' and 'present').
- station: the experimental plot where saplings were planted.
- id: numeric code to identify each individual planted in the station (n=1-4 saplings per station).
- microhabitat: microhabitat type where seeds were sowed (two levels: 'open' and 'shrub').
- survival: whether the planted sapling survive to the first summer (1) or not (0).
File: soil_physicochemical.csv
Description: Dataset with the soil physicochemical variables for the 20 combinations of sites and microhabitats.
Variables
- site: study forest patch (ten sites).
- microhabitat: microhabitat type where seeds were sowed or saplings were planted (two levels: 'open' and 'shrub').
- PC1: first principal component score of a Principal Component Analysis (PCA).
- PC2: second principal component score of a PCA.
- clay: percentage (%) of the soil composed by clay.
- limousine: percentage (%) of the soil composed by limousine.
- sand: percentage (%) of the soil composed by sand.
- ph: pH.
- ce: specific conductivity or 'conductividad específica' in Spanish (µS/cm).
- nacl: NaCl (mg/L).
- c_org: percentage (%) of oxidisable organic carbon.
- om: percentage (%) of oxidisable organic matter.
- p: Phosphorus (mg/Kg).
- ca: Calcium (cmolc/Kg).
- mg: Magnesium (cmolc/Kg).
- na: Sodium (cmolc/Kg).
- k: Potassium (cmolc/Kg).
- fe: Iron (cmolc/Kg).
- mn: Manganese (cmolc/Kg).
- zn: Zinc (cmolc/Kg).
- cu: Copper (cmolc/Kg).
- no3-: nitrates (mg/Kg).
- mean_vwc: the mean value of soil moisture.
- cv_vwc: coefficient of variation of the values of vwc.
- min_vwc: minimum value of vwc.
- max_vwc: maximum value of vwc.
- summer_vwc: mean of vwc values recorded during the period Jun-September.
File: soil_moisture_vwc_values.csv
Description: Dataset with soil moisture information, measured as volumetric water content (vwc) with a time-domain reflectometer (TDR).
Variables
- species: study plant species (n=10 species).
- microhabitat: microhabitat type where seeds were sowed (two levels: 'open' and 'shrub').
- site: study forest patch (ten sites).
- site_type: whether the study species was present or absent in the site (two levels: 'absent' and 'present').
- station: the experimental plot where seeds were sowed.
- treatment: the level of protection against vertebrate granivores and herbivores (two levels: 'protected' and 'unprotected').
- year: the year when seeds were sown, since we replicated the sowing experiment in two years for a subset of species.
- station_id: unique code of the sowing station, conformed by the union of the following variables: 'species'-'site'-'station'-'treatment'-'year'
- mean_vwc: the mean value of soil moisture, measured as volumetric water content (vwc), at the station level.
- sd_vwc: standard deviation of the vwc values of the station.
- min_vwc: minimum value of vwc registred for the station.
- max_vwc: maximum value of vwc registred for the station.
- cv: coefficient of variation of the values of vwc registred for the station.
- summer_vwc: mean of vwc values recorded during the period Jun-September.
- winter_vwc: mean of vwc values recorded during the period Nov-Feb.
File: vegetation_cover.csv
Description: Dataset with data of vegetation cover in the ten study forest sites or patches. We characterised the vegetation structure and composition with 30-m linear intercept transects in each patch (n = 7–10).
Variables
- site: study forest patch (ten sites).
- species: woody plant species identified in the intercept transects (taxonomy is standardised according to World Flora Online).
- interval_cm: the length (cm) intersecting individual(s) on an specific species.
- total_sampled_cm: the total lenght sampled at the site level.
- cover: percentage of the total sampled lenght covered by an species at site level.
File: code_1_sowing_demographic_analysis.R
Description: Code employed to analyse germination, seedling survival, sapling survival and seedling establishment of the study functional groups (dry and fleshy-fruited species) and species (n=10) according to patch occupancy level.
File: code_2_temporal_trends.R
Description: For three species that were sown in two different years, this code explores the temporal consistency in the establishment patterns across patch occupancy levels.
File: code_3_planting.R
Description: Code employed to explore how the survival of planted saplings was affected by the patch occupance level.
File: code_4_soil_analyses.R
Description: Code employed to explore the relationship between soil dissimilarity indexes and species-level odds ratio for seedling establishment between unoccupied and occupied patches.
File: code_5_soil_moisture_analyses.R
Description: Code employed to explore the effect of soil moisture on the seedling establishment probability of sowed seeds.
File: code_6_vegetation_cover.R
Description: Code to characterise the vegetation composition and structure of the ten study sites or patches.
Code/software
All analyses were conducted in the R version 4.3.0 (2023-04-21).