Data from: Habitat fragmentation drives pest termite risk in humid but not arid biomes
Data files
Nov 25, 2024 version files 81.47 MB
-
Checklist_of_Isoptera_in_local_haitats.xlsx
1.72 MB
-
Checklist_of_Isoptera_on_global_islands.xlsx
498.27 KB
-
Data_for_analysis_and_visualization.xlsx
846 KB
-
README.md
29.82 KB
-
Species_list_and_traits.xlsx
507.67 KB
-
termites_tacted.tre
77.87 MB
Abstract
Predicting global change effects poses significant challenges due to the intricate interplay between climate change and anthropogenic stressors in shaping ecological communities and their function, such as pest outbreak risk. Termites are ecosystem engineers, yet some pest species are causing worldwide economic losses. While habitat fragmentation seems to drive pest-dominated termite communities, its interaction with climate change effect remains unknown. We test if climate and habitat fragmentation interactively alter interspecific competition that may limit pest termite risk. Leveraging global termite cooccurrence including 280 pest species, we found that competitively superior termite species (e.g. large-bodied) increased in large and continuous habitats solely at high precipitation. While competitive species suppressed pest species globally, habitat fragmentation drove pest termite risk only in humid biomes. Unfortunately, humid tropics have experienced vast forest fragmentation and rainfall reduction over the past decades. These stressors, if not stopped, may drive pest termite risk potentially via competitive release.
README: Habitat fragmentation drives pest termite risk in humid but not arid biomes
https://doi.org/10.5061/dryad.k98sf7mdt
Description of the data and file structure
We provided the checklist of coocurring termite species at local scale ('Checklist of Isoptera in local haitats.xlsx') and at oceanic-island scale ('Checklist of Isoptera on global islands.xlsx'), which included the original records of species distribution per site, the environmental attributes and spatial coordinates per site, and the full reference list from which the distribution records were extracted.
We also provided the functional trait information for each studied species ('Species list and traits.xlsx'). For morpho-species, we calculated the mean trait values of intrageneric species as the estimate. Trait description and reference source were also reported. For body size, we plotted the correlation of body size between soldier and imago caste, which was significantly positive (R2 = 0.62). Therefore, we used the body size of soldier caste for analyses.
We compiled the response and predictor variables per island (or local site) into a single file ('Data for analysis and visualization.xlsx'). This file assembled all environmental predictors, community-weighted means of pest species risk and other functional traits, and the standardized effect sizes (SES) of mean pairwise functional distance (MFD) and mean pairwise phylogenetic distance (MPD). Besides, we used the proportional data among cooccurring species per local community to calculate the proportion-weighted mean of pest species risk.
Species pool was defined in two different ways. First, we treated all species occur on islands (or local sites) as the full species pool. Second, dispersion-field species pool was defined for each island (or local site), by including any species within the genera that had geographic extents overlapping the focal assemblage. We calculated the functional and phylogenetic SES values based on the full or dispersion-field species pool. For the former case, we labelled the SES values as 'SES.MFD' and 'SES.MPD'. For the latter case, we labelled the SES values as 'SES.MFD.df' and 'SES.MPD.df'.
We provided 1000 pseudo-posterior phylogeney tree ('termites_tacted.tre') of more than 3000 termite species that were constructed by Dr. Fernanda S. Caron and Prof. Marcio R. Pie (see their orginal research for more details; 10.1111/zsc.12502). This tree included all species covered in this study, and the morpho-species was inserted into the tree topology within their corresponding genera, subfamilies or families, by taking into account branch lengths determined by local diversification rates. The final phylogeny tree was generated through reserving species occurring in this study.
Description of the variables
Checklist of Isoptera on global islands
Sheet 1: Occurrence data
- scientificName: Currently valid scientific name
- Subspecies: Subspecies name
- family: Family name
- subfamily: Subfamily name
- Genus_final: Curated genus name
- genus: Genus name reported by the source
- species: Species name reported by the source
- Island name: The name of the island where the species occurs
- Island.No: The ID of island
- Latitude: Latitude of the island centroid
- Longitude: Longitude of the island centroid
- Country: The country name where the species occurs
- Reference: The data source that reports the specific island where the species occurs
- DOI: DOI or the accessing link of the reference
- Comment: Other information
- Single: Is it the only termite species that occurs on the specific island? 1 - Yes; 0 - No
Sheet 2: Island metric
- Island.No: The ID of island
- Latitude: Latitude of the island centroid
- Longitude: Longitude of the island centroid
- Country: The country name where the species occurs
- Archip: The name of the archipelago where the island occurs
- Island: The name of the island where the species occurs
- Area: Island area (km2)
- Dist: Distance to the nearest mainland (km)
- SLMP: The log10‐transformed sum of the proportions of landmass within buffer distances of 100, 1,000 and 10,000 km around the island perimeter
- Temp: The maximum values per island polygon of mean annual temperature (°C)
- Prec: The maximum values per island polygon of mean annual precipitation (mm)
- CCVT: Climate change velocity, calculated as the ratio between the temporal change in temperature per year (°C/year) and the contemporary spatial change in temperature (°C/m) and expressed in distance units per time (m/year)
- Ecoregion: The smallest-scale units in the Marine Ecoregions of the World (MEOW) system
- Province: The meidum-scale units in the Marine Ecoregions of the World (MEOW) system
- Realm: The largest-scale units in the Marine Ecoregions of the World (MEOW) system
- Single: Does the island have only one termite species? 1 - Yes; 0 - No
Sheet 3: Reference list
- Title: The data source that reports the distribution or checklist of termite species on the oceanic island(s)
- DOI: DOI or the accessing link of the reference
Checklist of Isoptera in local haitats
Sheet 1: Occurrence data
- Habitat type: Is the local site located on the mainland or oceanic island?
- StudyID: The ID of the reference
- scientificName: Currently valid scientific name
- Original record: The full species name reported in the data source
- Family: Family name
- subfamily: Subfamily name
- Genus_final: Curated genus name
- Genus: Genus name reported by the source
- Species: Species name reported by the source
- Abundance type: The method of counting the numeric advantage of each species used in the specific study. "Occurrence frequency" is the number of transects or quadrats where the species is found. "Total individual number" is the total counts of termite workers and soldiers found in the site. "Total individual number (OTU reads)" is the total counts of the OTU reads sampled from the site.
- Abundance: The abundance of the species
- Percent ratio: The relative abundance of the species found in the site
- Pest.risk: The level of pest risk of the species
- Diet information: The dite of the species
- Sampling method: The method of sampling termites in the specific study: transect or quadrat
- Number of sites: The total number of local sites of the same area that are pooled to calculate the abundance
- Number of sampling units: The total number of transects or quadrats of the same area that are pooled to calculate the abundance
- Width of sampling unit: The width of the transect or quadrat
- Length of sampling unit: The length of the transect or quadrat
- Total sampling area: Total sampling area
- Site name: The full name of local site
- SiteID: The ID of local site
- Taxonomic-level: The taxonomic level for how the species is identified
- Comment: Other information
- Diet reference1: The first reference that reports the diet information of the species
- Diet reference2: The second reference that reports the diet information of the species
- Single: Is it the only termite species that occurs in the local site? 1 - Yes; 0 - No
Sheet 2: Habitat metric
- SiteID: The ID of local site
- Latitude: Latitude of the local site
- Longitude: Longitude of the local site
- Treecover: Percentage of tree cover in 2010 at 30-m resolution
- Treecover.z: The Z score of percentage tree cover by comparing the observed tree cover to the mean values of all grids (30 × 30 m) within the 5 × 5 km. Specifically, the difference between observed and mean values was divided by the standard deviation of percent tree cover among all grids. This metric was independent of the cross-regional difference in natural tree cover and, thus, could be used to assessed the relative level of habitat area across regions
- Landuse: The land use proportion (%) within a 5,000 × 5,000 m area that was covered by human infrastructure, cropland, and water, based on the 2019 global land cover map at 30-m resolution
- Temp: Mean annual temperature (°C)
- Prec: Mean annual precipitation (mm)
- CCVT: Climate change velocity, calculated as the ratio between the temporal change in temperature per year (°C/year) and the contemporary spatial change in temperature (°C/m) and expressed in distance units per time (m/year)
- Land use: The land use reported in the source
- Land use type: One of the seven types of land use categories that the specific site belongs to, including the abandoned, agriculture, disturbed, forestry, grazing, protected, and restored
- Native vegetation: One of the four types of native vegetation categories that the specific site belongs to, including the forest, woodland, grassland, and savanna
- Ecoregion: The smallest-scale units in the Terrestrial Ecoregions of the World (TEOW) system
- Biome: The meidum-scale units in the Terrestrial Ecoregions of the World (TEOW) system
- Realm: The largest-scale units in the Terrestrial Ecoregions of the World (TEOW) system
- Continent: The continent where the local site distributes
- Habitat.type: Is the local site located on the mainland or oceanic island?
- Single: Does the local site have only one termite species? 1 - Yes; 0 - No
Sheet 3: Reference list
- Study: The ID of the reference
- Title: The data source that reports the community composition and/or abundance of termites in the local site(s)
- DOI: DOI or the accessing link of the reference
Species list and traits
Sheet 1: Simplified
- Species: The scientific name of the termite species
- Island: Does the termite species occur on any of the documented islands?
- Local: Does the termite species occur in any of the local sites?
- Pest.risk: The classification of pest damage level (0 = non-pest; 1 = pest; 2 = major pest)
- Size.soldier: The head width (mm) of largest soldier caste; for soldierless species, the head width of largest worker caste is used
- Morph.soldier: The number of size morph for soldier caste (0 = soldierless; 1 = monomorphic; 2 = dimorphic or trimorphic
- Diet.dominant: The most common feeding diet per genus (1 = wood-feeding lower termites; 2 = wood-feeding Termitidae; 3 = Humus-feeding Termitidae; 4 = Mineral-soil-feeding Termitidae)
- Nesting.complexity: The complexity of nesting strategy (1 = one-piece nester; 2 = multiple-piece nester; 3 = central-piece nester)
- Soil.access: Whether or not have access to soil (0/1)
- Diet.differentiation: The number of distinct diet types per genus
- Forager: Whether or not foraging outside the nest
- Diet.soil: Whether or not ≥ 1 species per genus show soil-feeding habit (or feeding in wood/soil interface)
- Taxonomy: The taxonomic level for how the species is identified
Sheet 2: Full data
- scientificName: The scientific name of the termite species
- Used_name: The adapted format of species name that is used in the data analyses
- Island (0/1): Does the termite species occur on any of the documented islands?
- Local (0/1): Does the termite species occur in any of the local sites?
- Unique species in one-species site: Does the termite species only occur in one local site or island? 1 - Yes; 0 - No
- Taxonomy: The taxonomic level for how the species is identified
- Family: Family name
- Subfamily: Subfamily name
- Genus: Genus name
- Pest.risk: The classification of pest damage level (0 = non-pest; 1 = pest; 2 = major pest)
- Size: Maximum head width of soldier caste (worker caste for soldierless species)
- Size.soldier: The head width (mm) of largest soldier caste; for soldierless species, the head width of largest worker caste is used
- Morph.soldier: The number of size morph for soldier caste (0 = soldierless; 1 = monomorphic; 2 = dimorphic or trimorphic
- Size_imago: The head width (mm) of imago alate
- Forager: Whether or not foraging outside the nest
- Diet.dominant: The most common feeding diet per genus (1 = wood-feeding lower termites; 2 = wood-feeding Termitidae; 3 = Humus-feeding Termitidae; 4 = Mineral-soil-feeding Termitidae)
- Diet.differentiation: The number of distinct diet types per genus
- Diet.soil: Whether or not ≥ 1 species per genus show soil-feeding habit (or feeding in wood/soil interface)
- Nesting.complexity: The complexity of nesting strategy (1 = one-piece nester; 2 = multiple-piece nester; 3 = central-piece nester)
- Soil.access: Whether or not have access to soil (0/1)
Sheet 3: Trait description (defining the variables presented in the previous two spreadsheets)
Sheet 4: Size literature
- Species: The scientific name of the termite species
- Family: Family name
- Caste: Soldier caste for species with soldier caste, or worker caste for soldierless species
- Head width: Head width (mm) of largest soldier caste
- Head width_large: Head width (mm) of largest soldier caste
- Head width_small: Head width (mm) of smallest soldier caste
- Reference1: The first reference reporting the head width
- DOI1: The DOI or accessing link of the first reference
- Reference2: The second reference reporting the head width
- DOI2: The DOI or accessing link of the second reference
- Reference3: The third reference reporting the head width
- DOI3: The DOI or accessing link of the third reference
- Range1: The range of head width reported from the first reference
- Range2: The range of head width reported from the second reference
- Range3: The range of head width reported from the third reference
- Comment: Other information
Sheet 5: Size correlation
- scientificName: The scientific name of the termite species
- Family: Family name
- Subfamily: Subfamily name
- Genus: Genus name
- Size_soldier: The head width of soldier caste
- Size_imago: The head width of imago caste
Sheet 5: Diet literature
- Genus: Genus name
- Family: Family name
- Subfamily: Subfamily name
- Diet.dominant: The most common feeding diet per genus (1 = wood-feeding lower termites; 2 = wood-feeding Termitidae; 3 = Humus-feeding Termitidae; 4 = Mineral-soil-feeding Termitidae)
- Diet.differentiation: The number of distinct diet types per genus
- Wood: Does any species of the genus eat wood? 1 - Yes; 0 - No
- Litter: Does any species of the genus eat litter? 1 - Yes; 0 - No
- Grass: Does any species of the genus eat grass? 1 - Yes; 0 - No
- Epiphyte: Does any species of the genus eat epiphyte? 1 - Yes; 0 - No
- Fungus: Does any species of the genus eat fungus? 1 - Yes; 0 - No
- Dung: Does any species of the genus eat dung? 1 - Yes; 0 - No
- Soil: Does any species of the genus eat soil? 1 - Yes; 0 - No
- Reference for dominant diet: The reference reporting the dominant diet information
- Reference for soil-feeding (or wood/soil interface) species within non-soil feeding genus (i.e. Dite_dominant = 2): The reference reporting the soil-feeding behaviors for species under the genus whose dominant diet is non-soil
Data for analysis and visualization
Sheet 1: Island
- Island.No: The ID of island
- SES.MFD: The standardized effect size (SES) for functional dissimilarity, with the regional species pool being species occurring on all islands
- SES.MPD: The standardized effect size (SES) for phylogenetic dissimilarity, with the regional species pool being species occurring on all islands
- SES.MFD.df: The standardized effect size (SES) for functional dissimilarity, with the regional species pool being species with geographic extents overlapping the focal assemblage
SES.MPD.df: The standardized effect size (SES) for phylogenetic dissimilarity, with the regional species pool being species with geographic extents overlapping the focal assemblage
Pest.risk: Community-weighted mean of pest risk level
Size.soldier: Community-weighted mean of the soldier head width
Morph.soldier: Community-weighted mean of the number of size morph for soldier caste
Diet.dominant: Community-weighted mean of the dominant diet
Nesting.complexity: Community-weighted mean of nesting strategy complexity
Soil.access: Community-weighted mean of soil access
Diet.differentiation: Community-weighted mean of diet differentiation
Forager: Community-weighted mean of forager trait
Diet.soil: Community-weighted mean of soil-feeding trait
PC1: The first principal component of termite functional traits
PC2: The second principal component of termite functional traits
PC3: The third principal component of termite functional traits
PC4: The fourth principal component of termite functional traits
Latitude: Latitude of the island centroid
Longitude: Longitude of the island centroid
Country: The country name where the island occurs
Archip: The name of the archipelago where the island occurs
Island: Island name
Area: Island area (km2)
Dist: Distance to the nearest mainland (km)
SLMP: The log10‐transformed sum of the proportions of landmass within buffer distances of 100, 1,000 and 10,000 km around the island perimeter
Temp: The maximum values per island polygon of mean annual temperature (°C)
Prec: The maximum values per island polygon of mean annual precipitation (mm)
CCVT: Climate change velocity, calculated as the ratio between the temporal change in temperature per year (°C/year) and the contemporary spatial change in temperature (°C/m) and expressed in distance units per time (m/year)
Ecoregion: The smallest-scale units in the Marine Ecoregions of the World (MEOW) system
Province: The meidum-scale units in the Marine Ecoregions of the World (MEOW) system
Realm: The largest-scale units in the Marine Ecoregions of the World (MEOW) system
Single: Does the island have only one termite species? 1 - Yes; 0 - No
SR: The total number of termite species on the island
PrecS: The scaled and centered values of mean annual precipitation
AreaS: The scaled and centered values of island area
DistS: The scaled and centered values of distance to the nearest mainland
SLMPS: The scaled and centered values of SLMP
Sheet 2: Local
- SiteID: The ID of local site
- SES.MFD: The standardized effect size (SES) for functional dissimilarity, with the regional species pool being species occurring in all local sites
- SES.MPD: The standardized effect size (SES) for phylogenetic dissimilarity, with the regional species pool being species occurring in all local sites
- SES.MFD.df: The standardized effect size (SES) for functional dissimilarity, with the regional species pool being species with geographic extents overlapping the focal assemblage
SES.MPD.df: The standardized effect size (SES) for phylogenetic dissimilarity, with the regional species pool being species with geographic extents overlapping the focal assemblage
Pest.risk: Community-weighted mean of pest risk level
Size.soldier: Community-weighted mean of the soldier head width
Morph.soldier: Community-weighted mean of the number of size morph for soldier caste
Diet.dominant: Community-weighted mean of the dominant diet
Nesting.complexity: Community-weighted mean of nesting strategy complexity
Soil.access: Community-weighted mean of soil access
Diet.differentiation: Community-weighted mean of diet differentiation
Forager: Community-weighted mean of forager trait
Diet.soil: Community-weighted mean of soil-feeding trait
PC1: The first principal component of termite functional traits
PC2: The second principal component of termite functional traits
PC3: The third principal component of termite functional traits
PC4: The fourth principal component of termite functional traits
Latitude: Latitude of the local site
Longitude: Longitude of the local site
Treecover: Percentage of tree cover in 2010 at 30-m resolution
Treecover.z: The Z score of percentage tree cover by comparing the observed tree cover to the mean values of all grids (30 × 30 m) within the 5 × 5 km. Specifically, the difference between observed and mean values was divided by the standard deviation of percent tree cover among all grids. This metric was independent of the cross-regional difference in natural tree cover and, thus, could be used to assessed the relative level of habitat area across regions
Landuse: The land use proportion (%) within a 5,000 × 5,000 m area that was covered by human infrastructure, cropland, and water, based on the 2019 global land cover map at 30-m resolution
Temp: Mean annual temperature (°C)
Prec: Mean annual precipitation (mm)
CCVT: Climate change velocity, calculated as the ratio between the temporal change in temperature per year (°C/year) and the contemporary spatial change in temperature (°C/m) and expressed in distance units per time (m/year)
Land use: The land use reported in the source
Land use type: One of the seven types of land use categories that the specific site belongs to, including the abandoned, agriculture, disturbed, forestry, grazing, protected, and restored
Native vegetation: One of the four types of native vegetation categories that the specific site belongs to, including the forest, woodland, grassland, and savanna
Ecoregion: The smallest-scale units in the Terrestrial Ecoregions of the World (TEOW) system
Biome: The meidum-scale units in the Terrestrial Ecoregions of the World (TEOW) system
Realm: The largest-scale units in the Terrestrial Ecoregions of the World (TEOW) system
Continent: The continent where the local site distributes
Habitat.type: Is the local site located on the mainland or oceanic island?
Single: Does the local site have only one termite species? 1 - Yes; 0 - No
SR: The total number of termite species on the island
PrecS: The scaled and centered values of mean annual precipitation
TreecoverS: The scaled and centered values of percentage tree cover
Treecover.zS: The scaled and centered values of the Z score of percentage tree cover
LanduseS: The scaled and centered values of land use proportion
Sheet 3: SES.island
- Island.No: The ID of island
- SES.MFD: The standardized effect size (SES) for functional dissimilarity, with the regional species pool being species occurring on all islands
- SES.MPD: The standardized effect size (SES) for phylogenetic dissimilarity, with the regional species pool being species occurring on all islands
- SES.MFD.df: The standardized effect size (SES) for functional dissimilarity, with the regional species pool being species with geographic extents overlapping the focal assemblage
- SES.MPD.df: The standardized effect size (SES) for phylogenetic dissimilarity, with the regional species pool being species with geographic extents overlapping the focal assemblage
Sheet 4: SES.local
- SiteID: The ID of local site
- SES.MFD: The standardized effect size (SES) for functional dissimilarity, with the regional species pool being species occurring in all local sites
- SES.MPD: The standardized effect size (SES) for phylogenetic dissimilarity, with the regional species pool being species occurring in all local sites
- SES.MFD.df: The standardized effect size (SES) for functional dissimilarity, with the regional species pool being species with geographic extents overlapping the focal assemblage
- SES.MPD.df: The standardized effect size (SES) for phylogenetic dissimilarity, with the regional species pool being species with geographic extents overlapping the focal assemblage
Sheet 5: Pest.risk.global
- SiteID: The ID of local site
- Scale: The spatial scale of the species assemblages: island or local site
- Latitude: Latitude of the local site
- Longitude: Longitude of the local site
- Pest.risk: Community-weighted mean of pest risk level
Sheet 6: Pest.risk.weighted
- SiteID: The ID of local site
- SES.MFD: The standardized effect size (SES) for functional dissimilarity, with the regional species pool being species occurring in all local sites
- SES.MPD: The standardized effect size (SES) for phylogenetic dissimilarity, with the regional species pool being species occurring in all local sites
- SES.MFD.df: The standardized effect size (SES) for functional dissimilarity, with the regional species pool being species with geographic extents overlapping the focal assemblage
SES.MPD.df: The standardized effect size (SES) for phylogenetic dissimilarity, with the regional species pool being species with geographic extents overlapping the focal assemblage
Pest.risk: Community-weighted mean of pest risk level
Size.soldier: Community-weighted mean of the soldier head width
Morph.soldier: Community-weighted mean of the number of size morph for soldier caste
Diet.dominant: Community-weighted mean of the dominant diet
Nesting.complexity: Community-weighted mean of nesting strategy complexity
Soil.access: Community-weighted mean of soil access
Diet.differentiation: Community-weighted mean of diet differentiation
Forager: Community-weighted mean of forager trait
Diet.soil: Community-weighted mean of soil-feeding trait
PC1: The first principal component of termite functional traits
PC2: The second principal component of termite functional traits
PC3: The third principal component of termite functional traits
PC4: The fourth principal component of termite functional traits
Latitude: Latitude of the local site
Longitude: Longitude of the local site
Treecover: Percentage of tree cover in 2010 at 30-m resolution
Treecover.z: The Z score of percentage tree cover by comparing the observed tree cover to the mean values of all grids (30 × 30 m) within the 5 × 5 km. Specifically, the difference between observed and mean values was divided by the standard deviation of percent tree cover among all grids. This metric was independent of the cross-regional difference in natural tree cover and, thus, could be used to assessed the relative level of habitat area across regions
Landuse: The land use proportion (%) within a 5,000 × 5,000 m area that was covered by human infrastructure, cropland, and water, based on the 2019 global land cover map at 30-m resolution
Temp: Mean annual temperature (°C)
Prec: Mean annual precipitation (mm)
CCVT: Climate change velocity, calculated as the ratio between the temporal change in temperature per year (°C/year) and the contemporary spatial change in temperature (°C/m) and expressed in distance units per time (m/year)
Land use: The land use reported in the source
Land use type: One of the seven types of land use categories that the specific site belongs to, including the abandoned, agriculture, disturbed, forestry, grazing, protected, and restored
Native vegetation: One of the four types of native vegetation categories that the specific site belongs to, including the forest, woodland, grassland, and savanna
Ecoregion: The smallest-scale units in the Terrestrial Ecoregions of the World (TEOW) system
Biome: The meidum-scale units in the Terrestrial Ecoregions of the World (TEOW) system
Realm: The largest-scale units in the Terrestrial Ecoregions of the World (TEOW) system
Continent: The continent where the local site distributes
Habitat.type: Is the local site located on the mainland or oceanic island?
Single: Does the local site have only one termite species? 1 - Yes; 0 - No
SR: The total number of termite species on the island
PrecS: The scaled and centered values of mean annual precipitation
TreecoverS: The scaled and centered values of percentage tree cover
Treecover.zS: The scaled and centered values of the Z score of percentage tree cover
LanduseS: The scaled and centered values of land use proportion
Pest.risk.weighted: Abundance-weighted mean of pest risk level
Code/Software
All statistical analyses were conducted in R 4.2.3. We provided four R scripts to generate all results reported in the main text and supplementary materials.
'SES-full species pool.R' was used to calculate the SES.MPD among cooccurring species for 1000 pseudo-posterior trees. It may take 1~2 weeks to finish the whole process. Meanwhile, SES.MFD could be fastly computed and therefore we inserted the respective code in the next script.
'SES-dispersion field.R' was used to calculate the SES.MFD.df and SES.MPD.df among cooccurring species. Given that dispersion-field species pool was defined differently for each island or local site, the running time was considerably higher (i.e. times the number of islands or local sites) than that of 'SES-full species pool.R'.
'Analysis code.R' was used to generate the dataset for analysis and visualization ('Data for analysis and visualization.xlsx'). Furthermore, multiple linear regression after stepwise selection was used to determine the important predictors and interaction effects. Finally, global Moran's I analysis was conducted to evaluate the spatial autocorrelation in the residuals of each response variable after accounting for the environmental effects.
'Figure code.R' was used to visualize the distribution of data records (Fig. 1a) and interaction effects between climate and habitat fragmentation (Figs. 3-5 and Figs. S4-S7). In addition, we also provided the code for plotting the Pearson's correlation matrix among functional traits (Fig. S2) and environmental predictors (Fig. S3).