Genetic structure and diversity of the declining orchid Gymnadenia conopsea in Scandinavia: Implications for conservation and management
Data files
Mar 06, 2025 version files 14.70 MB
-
clean_raw_variants1he.txt
14.64 MB
-
populations.txt
58.23 KB
-
README.md
3.21 KB
Abstract
Understanding how historical versus contemporary processes shape population genetic structure and diversity is important to design effective management actions for threatened species. We genotyped 1,834 SNPs in 1,120 individuals in 110 Scandinavian populations of the declining orchid Gymnadenia conopsea, in three different habitat types, to examine if genetic structure was related to wind speed, terrain ruggedness, forest cover, and seasonality at the landscape scale, and if genetic diversity increases with census population size, and is higher in core habitats (fen and meadow) than in marginal, coastal habitats. We identified three genetic clusters and pronounced isolation by distance, consistent with two independent colonization routes after last glacial maximum, followed by admixture. Effective population size was highest in the admixed cluster. Estimates of effective migration indicated reduced gene flow along the Atlantic coast, between coastal and inland populations, and among southern meadow populations. High landscape resistance to gene flow was associated with complex topography and pronounced seasonality. Genetic diversity increased with population size, but did not vary among habitat types. Genetic diversity peaked in core habitats, i.e., southern meadows and inland fens along the Scandes mountains. Lowest genetic diversity was found on the Atlantic coast and in a few scattered populations. Current genetic structure suggests a strong legacy of historical events, and the high genetic diversity documented in the main Scandinavian range indicates that current viability and future adaptation potential is high. To maintain genetic diversity and connectivity between genetic groups, it is particularly important to preserve southern meadow populations, which currently are in strong decline. Generally, results illustrate how a declining species can help us understand the impact of historical and current processes, how landscape genetic data can inform proactive conservation, and how a slow genetic response to fragmentation can allow time to maintain genetic diversity through habitat restoration and management.
https://doi.org/10.5061/dryad.5x69p8dfp
Description of the data and file structure
The data contains genotypic and ecological data from 110 Scandinavian populations of Gymnadenia conopsea. In each population, we genotyped 1834 SNPs in 11 individuals. For each population, we collected data on population size (= number of flowering individuals) and habitat type (meadow, fen, coastal), as well as geographic coordinates.
Genotypic data are found in "clean_raw_variants1he.txt". Each row shows the genotype for all 1834 SNPs (first SNP = G1R1LP02p002_0002945, last SNP = G3R2YP08p503_040553) for one sample (= individual), where the sample is identified in the first column (first sample = 2AR1, last sample = ZA74).
Population data are found in "populations.txt". Each row corresponds to one individual. Column headers indicate:
- Individual = individual of Gymnadenia conopsea sampled in the field, n = 11 per population (fewer if population size was <11)
- Population = population name
- County = county in Sweden or Norway
- X = East geographic coordinate (WGS84 dec)
- Y = North geographic coordinate (WGS84 dec)
- hab0 = detailed habitat type (not used in analyses)
- hab1 = habitat type (M = fen, E = meadow, S = coastal)
- Part = geographic region (South or North)
Code/software
To analyse spatial genetic structure, we used Admixture v 1.3.0 (Alexander et al., 2009) and conStruct v 1.0.4 (Bradburd, 2019).
We used the following R packages:
- ade4 version 1.7.18 (Dray & Dufour, 2007)
- adegenet version 2.1.5 (Jombart & Ahmed, 2011)
- adespatial version 0.3.20 (Dray et al., 2006)
- PCAviz version 0.3.37 (Novembre et al., 2019)
- ResistanceGA version 4.2-10 (Peterman, 2018)
- raster version 3.6-31 (Hijmans, 2023)
- strataG version 1.0.6. (Archer et al. 2017)
The flow of the separate analyses are described in the uploaded R markdown files:
- conStruct_evaluation.Rmd - analyses of spatial genetic structure
- Effective_population_size.Rmd - analyses of effective population size
- Resistance.R - analyses of landscape resistance
- PCAviz_KS_v2.Rmd - PCA and visualization
Access information
Data was derived from the following sources:
- data on mean annual wind speed were downloaded from the Global Wind Atlas (https://globalwindatlas.info/en; Davis et al., 2023)
- elevation data were downloaded from SRTM 90-meter resolution data (https://bigdata.cgiar.org/srtm-90m-digital-elevation-database/)
- data on forest land cover were downloaded from the CORINE land cover 2018 database, 100-meter raster (https://doi.org/10.2909/960998c1-1870-4e82-8051-6485205ebbac; European Environment Agency 2019).
- climatic data were downloaded from worldclim (bioclimatic variable “precipitation seasonality” , BIO15, 30 s resolution, version 2, https://worldclim.org/, Fick & Hijmans, 2017)