Interactions between maternal provisioning and natural selection on seed mass fluctuate across heat waves
Data files
Abstract
This repository contains two datasets from Gaspard et al. 2026. FileA corresponds to seed and germination properties collected from crosses in a resurrection experiment, while FileB corresponds to seed and germination properties from a growth chamber experiment manipulating drought stress to simulate natural heat waves. Below is the abstract from the corresponding manuscript.
Premise of study: Changing climates are leading to more frequent and severe heatwaves, potentially threatening plant populations. Both acclimation to stress and selection for heat-resistant phenotypes occur during heatwaves. However, plastic responses and selection do not necessarily interact cohesively – even producing responses in opposite directions, a phenomenon termed counter gradient variation – that could cause maladaptation in future generations.
Methods: Here we examine patterns of maternal provisioning and natural selection on seed mass and germination characteristics following two natural heatwaves across twelve annual Mimulus guttatus populations. We then recreate heatwave events occurring early and late within a growing season using a manipulative experiment to determine the reproducibility of our results.
Key Results: There was selection for greater seed mass during the year of the early season heat wave and seed mass remained high through the late-season heatwave. However, field observations of seed mass only reflect this pattern during the early season heatwave, and plants that experienced a late season heatwave had small seeds that germinated later. This incongruency is indicative of counter gradient variation with decreased maternal provisioning coinciding with selection for larger seeds.
Conclusions: Our results demonstrate that the relationship between selection and maternal provisioning fluctuates between years. Such counter gradient variation may be adaptive if smaller seeds are more likely to persist in the seed bank or if climatic conditions exhibit negative temporal autocorrelation. However, counter gradient variation could also constrain the potential of natural selection and lead to high mortality in heatwaves during successive years.
Two data files are contained within this Dryad entry. FileA corresponds to seed and germination properties collected from crosses in a resurrection experiment, while FileB corresponds to seed and germination properties from a growth chamber experiment manipulating drought stress to simulate natural heat waves.
FileA. This is a comma delimited file (fileA.csv) which contains seed mass and seed number data from 550 different Mimulus guttatus crosses from a resurrection experiment conducted in the University of Louisiana, Lafayette greenhouse as well as germination data for the progeny of each cross during a germination trial conducted within custom-walk-in growth chambers at UC-Berkeley. These experiments are described in detail in Gaspard et al. 2026. Briefly, seeds were collected from 12 different populations each year between 2018 and 2022. Field seeds were grown in greenhouse refresher generations each fall after collection and each line was selfed and saved. Seeds grown in all refresher generation were grown in a greenhouse common garden (a resurrection experiment) in 2023. Each plant was outcrossed with another plant from the same population and year to produce the crosses in this data file. Seeds from these crosses were weighted and then used within a germination trial at UC-Berkeley within growth chambers. We attempted to germinate five seeds from each line and recorded germination date for each seed that germinated.
FileB. This is a comma delimited file (fileB.csv) which contains phenotype data for 73 Mimulus guttatus individuals from a manipulative growth chamber experiment conducted at the University of Louisiana, Lafayette. This experiment is described in full detail in Gaspard et al. 2026. The experiment had three treatments including a early stress treatment, a late stress treatment and a control treatment. For the early stress treatment, plants were subjected to dry down between day 14-29 of the experiment. The late stress treatment was subjected to dry down between day 30 and 39, while the control treatment was well watered throughout the experiment. At the end of the dry down treatments, pots typically had ~20% volumetric water content. Following day 39, all plants were well watered. In each treatment, lines were grown from two populations. These lines were derived from field collections that had been grown in a greenhouse refresher generation and selfed. We collected morphological and phenological phenotypes for each plant and crossed plants with another plant from within the same population and treatment. We weighed seeds for each cross and conducted a germination trial to determine the germination ability and timing for each cross.
Questions? Please contact Nic Kooyers (nkooyers@gmail.com).
Description of the Data and file structure
FileA.
Each row of this spreadsheet represents a cross except for the first row header. Missing data is coded as '9999'
Format (Column -- Variable -- Description)
Column A -- Population -- The population where the individuals from the cross were originally collected.
Column B -- Year -- The year the lines in the cross were collected from the field.
Column C -- Mom -- The greenhouse ID within the resurrection experiment associated with the mother of a cross.
Column D -- Dad -- The greenhouse ID within the resurrection experiment associated with the father of a cross.
Column E -- SeedMass -- The mean mass of a seed from a particular cross (milligrams)
Column F -- SeedN -- Number of seeds weighted to determine mean seeds mass for a cross.
Column G -- MeanGermDate -- The mean number of days following planting to germination for seeds for a particular cross.
Column H -- GermPerc -- The percentage of seeds that germinated within the germinated trial for a individual cross. Five seeds were planted per cross.
Column I -- HW -- Whether the year of collection had a heat wave during the growing season (No heat wave: nonhw, heat wave: hw).
Column J -- CombYear -- Whether the year of collection had an early season or late season heat wave during the growing season (No heat wave: nonhw, early season heat wave: hw19, late season heat wave: hw21).
FileB.
Each row of this spreadsheet represents an individual plant except for the first row header. Missing data is coded as '9999'
Individual Line Population Year Treatment FlowerTime PlantHeight LeafNumFlowering LeafCountES LeafCountLS CrossPartner FemCorollaWidth MaleCorollaWidth CrossCount SeedCount SeedWeight PercGerm DayGerm
Format (Column -- Variable -- Description)
Column A -- Individual -- A number identifying each unique individual.
Column B -- Line -- The maternal line of the individual. Individuals from the same maternal line come from the same selfed cross in the refresher generation.
Column C -- Population -- The population where the line was originally collected.
Column D -- Year -- The year when then the line was originally collected
Column E -- Treatment -- The treatment that the individual experienced (early season dry down: Early Stress, late season dry down: Late Stress or control: Control)
Column F -- FlowerTime -- The time between germination and the first flower for an individual (days)
Column G -- PlantHeight -- The height from ground to top of the apical meristem for an individual at time of first flower (mm)
Column H -- LeafNumFlowering -- The number of leaves or an individual at time of first flower
Column I -- LeafCountES -- The number of leaves or an individual prior to the start of the early stress treatment.
Column J -- LeafCountLS -- The number of leaves or an individual prior to the start of the late stress treatment.
Column K -- CrossPartner -- The unique identifier (column A) for the father of a cross (sire) for an individual.
Column L -- CrossCount -- The number of flowers crossed between partners.
Column M -- SeedCount -- The number of total seeds resulting from all crosses from column L
Column N -- SeedMass -- The mean mass of a seed from the crosses (grams)
Column O -- PercGerm -- The percentage of seeds that germinated within a subsequent germinated trial. Five seeds were planted per cross.
Column P -- DayGerm -- The mean day of germination following removal from seed stratification for seeds that germinated from column O.
Sharing/access Information
Links to other publicly accessible locations of the data:
None available
Was data derived from another source?
No