Summary

R code associated with Martin et al.'s "Measuring demographic stochasticity and selection due to attrition in ex situ collections for plant conservation"

These scripts are part of a project on provenance selection and opportunity for selection in ex situ collections for plant conservation. The project uses as a study system a living collection of Quercus arkansana (Fagaceae) maintained in the Oertli Hardy Plant Nursery of the Missouri Botanical Garden.

The "viridis" R package was used for color palettes. https://cran.r-project.org/web/packages/viridis/index.html

Data preparation

Provenance data preparation (ProvenanceDataPreparation_2024August05.R)

• Obtain occurrence and provenance data for Q. arkansana (GBIF).
• Map known occurrences and collection localities of the unique accessions in the common garden experiment.
• Assign accessions to geographic regions, following Thomas et al. (2023, Biological Conservation 283: 110052).
• Save provenance data as a text file.

Survival data preparation (SurvivalDataPreparation_2024July03.R)

• Examine survival survey dates.
• Compare survival data between the two surveys.
• Examine the distribution of plantings among accessions.
• Examine the distribution of planting deaths among accessions.
• Save the data as a text file.

Null model for provenance selection

(NullModelDemographicStochasticity.R)

• Using the null model, representing demographic stochasticity, gauge provenance selection and opportunity for selection
• Read and examine distribution of initial plants per maternal line per provenance (i.e., geographic region), and surviving plants per maternal line per provenance.
• Calculate a demographic stochasticity null model.

Provenance selection and opportunity for selection

(CalculateProvSelection_OpportunitySelection.R)

• Calculate selection coefficients for provenances (i.e., geographic regions).
• Examine the relationship between the number of maternal lines and the mean survival rates across provenances.
• Calculate opportunity for selection within each provenance.
• Compare the null and observed opportunity for selection.

Q. arkansana provenance and occurrence map

(Q.arkansanaProvOverOccurrenceMap.R)

• Map latitudes and longitudes obtained from GBIF data and collected provenance records. Plot created with ggplot2 https://cran.r-project.org/web/packages/ggplot2/index.html