Viewing fragmentation through the disease triangle: implications for pathogen pressure on seeds
Data files
Jun 05, 2026 version files 13.41 MB
-
1._VegPlot_Description_Dryad.csv
10.89 KB
-
10._Fungicide-Raw_Dryad.csv
130.84 KB
-
2._Fungi_Site_Description_Dryad.csv
537 B
-
3._Veg2014-SitebySpeciesMatrix_Dryad.csv
331.80 KB
-
4._Plant_Diversity_Summarized_Dryad.csv
16.40 KB
-
5._PCoA_Plant_AllPlots_FigureData_Dryad.csv
1.92 MB
-
6._Fungi_SitebySpeciesMatrix_Dryad.csv
3.95 KB
-
7._PCoA_Fungi_FigureData_Dryad.csv
1.99 KB
-
8._FungiDiversity_Dryad.csv
2.71 KB
-
9._ibutton_Dryad.csv
10.98 MB
-
README.md
5.20 KB
Abstract
Disease and landscape fragmentation each shape plant diversity, but how and why pathogen pressure varies across fragmented landscapes is not well understood. Landscape fragmentation likely influences all three components of the classic plant pathology “disease triangle”—hosts, pathogens, and abiotic environments. Thus, we would expect pathogen pressure to vary with fragment size. We focus on data from an experimentally fragmented landscape in Kansas, USA as a case study for ways in which fragmentation may impact soilborne plant disease. We surveyed plant communities in permanent plots stratified across three habitat types: small patches, large patch edges, and large patch interiors. We also surveyed fungal communities by isolating and identifying fungi from seeds buried for a year near survey plots. To assess the impact of pathogens on seed hosts, we buried seeds of six plant species; half of the seeds were treated with fungicide.
Habitat type influenced all vertices of the disease triangle. Plant communities were compositionally distinct among habitat types, with large patch interiors containing the most species. Fungal communities isolated from seeds varied according to both host species and habitat type. Mean daily temperature range was highest in small patches. More extreme temperature ranges decreased germination across the landscape for two plant species, and habitat type influenced germination success for two additional species. Fungicide increased seed germination success across habitat types in two species, but for four species, benefits of fungicide depended on where in the landscape seeds were buried. Altogether, our study demonstrates that patch size alters the community context for disease by impacting all three vertices of the “disease triangle,” with consequences for seed survival. Our study also shows that impacts of disease on seed germination are not uniform across habitat types. This study provides a foundation for future work to mechanistically explore how fungal-mediated seed survival contributes to biodiversity in fragmented landscapes.
Data from this study include temperature data from ibutton sensors buried across the fragmented landscape, plant composition data, fungal composition (fungi isolated from buried seeds), seed germination data.
Dataset DOI: 10.5061/dryad.j6q573nsm
Viewing fragmentation through the disease triangle: implications for pathogen pressure on seeds
Description of the data and file structure
We surveyed plant communities in permanent plots stratified across three habitat types: small patches (S), large patch edges (L-E), and large patch interiors (L-I). We also surveyed fungal communities by isolating and identifying fungi from three species of plant seeds buried for a year near survey plots. To assess the impact of pathogens on seed hosts, we buried seeds of six plant species; half of the seeds were treated with fungicide. We measured germination success of seeds with and without fungicide protection. We quantified temperature at each plot location using ibutton temperature sensors.
Files and variables
Data files below include site descriptors, plant data, fungal data, temperature data, and results from the germination experiment.
Site Descriptions
1.VegPlot_Description_Dryad: Data file with plot location information.
- Location = plot identifier
- Seedbank_ID = Site ID used for seedbank data collected from the site
- Patch_SZ_Text = patch size in which plot is embedded
- Cluster_ID = Cluster scale identifier—each cluster contains 30 plots
- HabType = L-E (Large Edge), L-I (Large-Interior), S (Small patch)
- New2014=N (No: permanent plot) or Y (Yes: added in 2014)
2.Fungi_Site_Description_Dryad: Data file with site location information for fungi analyses conducted at the cluster scale.
- Cluster_ID = Cluster scale identifier
- HabType = L-E (Large Edge), L-I (Large-Interior), S (Small patch)
- Landloc = Cluster ID_HabType
- Sz = Patch Size
- Host=Plant species from which fungi was cultured (AND = Andropogon, CAS = Chamaecrista, RUD = Rudbeckia)
Vegetation
3.Veg2014-SitebySpeciesMatrix_Dryad:
Plot Identifier is in Row 1. Column 1 contains the code for plant species in the dataset (first three letters of genus, first two letters of species). Cells contain percent cover values for each species a 1 sq-m scale. Used in PCoA analyses conducted with Primer software.
4.Plant_Diversity_Summarized_Dryad: Richness (S) and Evenness (J’) for each plot.
- Patch_SZ_Text = patch size in which plot is embedded
- Cluster_ID = Cluster scale identifier—each cluster contains 30 plots
- HabType = L-E (Large Edge), L-I (Large-Interior), S (Small patch)
- New2014=N (No: permanent plot) or Y (Yes: added in 2014)
- S = species richness (number of plant species in 1 sq meter)
- J’= Pielou’s J evenness (for plants detected in 1 sq meter)
5.PCoA_Plant_AllPlots_FigureData_Dryad: PCoA axes output from analyses conducted in Primer. Used to generate ordination figure.
Fungi
6.FungiSiteBySpeciesMatrix_Dryad: Plot Identifier is in Row 1. Column 1 contains the genetic sequence for each fungal species. Cells contain integers that reflect the number of plots in a cluster where that fungus was isolated from a seed. Used in PCoA analyses conducted with Primer software.
7.PCoA_Fungi_FigureData_Dryad: PCoA axes output from analyses conducted with Primer. Used to generate ordination figure.
8.FungiDiversity_Dryad: Fungal richness (S) calculated for each seed species in each burial location.
- HabType = L-E (Large Edge), L-I (Large-Interior), S (Small patch)
- Host=Plant species from which fungi was cultured (AND = Andropogon, CAS = Chamaecrista, RUD = Rudbeckia).
- S = species richness (number of fungal species in a given seed species from each habitat type)
Temperature
9.Ibutton_Dryad: Daily temperatures recorded (°C) by ibuttons buried with untreated seeds.
- ibuttonID = unique identifier for ibutton
- Season = season of year (Su= June, July, August; F= September, October, November; W= December, January, Febraury; Sp= March, April, May)
- Month
- Day
- DayNum (starting with first day recording occurred)
- Site (matches Seedbank_Site in other files)
- Temperature (°C)
- HabType = L-E (Large Edge), L-I (Large-Interior), S (Small patch); Cluster_ID, Date, and Julian Day.
Germination Experiment
10.Fungicide-Raw: Data file containing results from seed germination trials.
- Unique = unique identifier for each seed packet; Site (matches Seedbank_Site in other files) Species = plant species with six letter codes containing first three letters of genus and first three letters of species
- SiteSample = combination of Site and Species columns
- F.NF = experimental treatment (F = Fungicide; NF = No Fungicide)
- Cluster_ID, HabType = L-E (Large Edge), L-I (Large-Interior), S (Small patch)
- PatchSz = patch size in which seeds were buried.
- Denom=number of seeds exhumed from bags after 1 year
- Total.Germ=number of germinated seeds
- Fails.Germ=number of seeds that did not germinate
- PropGerm=proportion of seeds that germinated
- BROMUS.cov=percent cover Brome grass in the plot where seeds were buried
- BROMUS.pa=presence (1) or absence (0) of Brome grass
- RangeTemp=mean daily range (max-min) temperature (°C)
