Data and code for: The evolution of plasticity in drought-related traits in central and edge populations of Mimulus cardinalis
Data files
Aug 05, 2026 version files 298.26 MB
-
1_Figure1_mapCV.R
12.46 KB
-
2_calculate_plasticity_TableS2.R
34.74 KB
-
3_Q1_Q2_brms_FigureS1-S8.R
44.15 KB
-
4_Figure2.R
14.97 KB
-
5_Q3_selection_analyses_Figure3_S9-15.R
89.54 KB
-
north_garden_scans.zip
122.90 MB
-
README.md
3 KB
-
south_garden_scans.zip
175.16 MB
Abstract
Phenotypic plasticity can influence population persistence, range dynamics, and evolutionary potential in heterogeneous landscapes. If adaptive, plasticity may buffer populations against environmental variability, climate change, and extreme events. Yet, few studies have examined how plasticity and its evolution vary across species’ ranges. This is especially important for edge populations, which are often most vulnerable to novel conditions. The Climate Variability Hypothesis postulates that greater environmental variability selects for genotypes that exhibit greater phenotypic plasticity. However, field tests of this hypothesis across species’ ranges remain rare. Using scarlet monkeyflower (Mimulus cardinalis), we combined a resurrection approach with common gardens to test whether (1) central and edge populations differ in plasticity in drought-associated traits, (2) there were evolutionary shifts in plasticity following an extreme drought, and (3) plasticity was adaptive. We measured plasticity in first flower date, specific leaf area, and leaf dry matter content across two leading-edge, two range-center, and two trailing-edge populations, comparing pre-drought ancestors and post-drought descendants grown in three gardens spanning the latitudinal range following a historic drought event. We found significant plasticity in all traits, and plasticity in the date of first flower was greatest in trailing-edge populations that have experienced greater interannual variation in precipitation. Evolutionary changes in plasticity occurred in some populations but did not consistently increase first-year fitness. The adaptive value of plasticity was trait-specific. These results show that plasticity is shaped by both spatial and temporal environmental variation and highlight the need to examine trait- and population-level responses to understand selection on plasticity at range edges.
Dataset DOI: 10.5061/dryad.kd51c5bp1
Description of the data and file structure
Using scarlet monkeyflower (Mimulus cardinalis), we combined a resurrection approach with common gardens to test whether traits associated with drought adaptation have evolved in populations across a species’ range in response to extreme drought and whether central and edge populations differ in plasticity in drought-associated traits.
Files and variables
File: north_garden_scans.zip
Description: This file includes raw images of leaf scans in the north garden in 2023 and their associated row-positions.
File: south_garden_scans.zip
Description: This file includes raw images of leaf scans in the south garden in 2023 and their associated row-positions.
File: 1_Figure1_mapCV.R
Description: Plot map and climate variables for M. cardinalis study populations (Fig. 1).
File: 2_calculate_plasticity_TableS2.R
Description: Calculate phenotypic plasticity for first flower date with and without central garden, as well as specific leaf area (SLA), and leaf dry matter content (LDMC) plasticity. Get sample size of families used to calculate plasticity for each trait (Supplemental Table S2). Output plasticity dataframes for subsequent analyses: doy_pp.csv, doy_pp_NS.csv, sla_pp.csv, ldmc_pp.csv.
File: 3_Q1_Q2_brms_FigureS1-S8.R
Description: Using brms, estimate uncertainty for differences among populations (Q1), and cohorts (Q2) in trait plasticity (first flower date plasticity (with and without central garden), SLA plasticity, and LDMC plasticity). Make Supplemental Figure 1-8, Supplemental Tables 3 - 5.
File: 4_Figure2.R
Description: Make Figure 2, plasticity in first flower date, sla, and ldmc across populations and cohorts.
File: 5_Q3_selection_analyses_Figure3_S9-15.R
Description: Model selection (linear vs quadratic models) with loo compare and across garden selection analyses. Plot Supplemental Table S6 (loo compare output) and S7 (selection coefficients using fitness from all gardens) and S8 (selection coefficients using fitness in north and south gardens only). Plot Figure 3 using raw data and estimated relationship with fitness and plasticity using linear models. Figure 3 uses fitness averaged across gardens, Supplemental Figure 13 uses fitness averaged across north and south gardens only and plot Supplemental Figures S9 - S15 visualizing estimated effect of plasticity on average (across garden) fitness.
Code/software
Analyses were done using R (Version 4.4.2) using the 'brms' package (Version 2.22.0; Burkner, 2017). Model support was evaluated with the 'LOO' package (Vehtari et al., 2017), and posterior distributions of plasticity slopes were extracted using the 'emmeans' package (Lenth, 2018).