Bet hedging is not sufficient to explain germination patterns of a winter annual plant

Bet hedging consists of life history strategies that buffer against environmental variability by trading off immediate and long-term fitness. Delayed germination in annual plants is a classic example of bet hedging and is often invoked to explain low germination fractions. We examined whether bet hedging explains low and variable germination fractions among 20 populations of the winter annual plant Clarkia xantiana ssp. xantiana that experience substantial variation in reproductive success among years. Leveraging 15 years of demographic monitoring and 3 years of field germination experiments, we assessed the fitness consequences of seed banks and compared optimal germination fractions from a density-independent bet-hedging model to observed germination fractions. We did not find consistent evidence of bet hedging or the expected trade-off between arithmetic and geometric mean fitness, though delayed germination increased long-term fitness in 7 of 20 populations. Optimal germination fractions were 2 to 5 times higher than observed germination fractions, and among-population variation in germination fractions was not correlated with risks across the life cycle. Our comprehensive test suggests that bet hedging is insufficient to explain the observed germination patterns. Understanding variation in germination strategies will likely require integrating bet hedging with complementary forces shaping the evolution of delayed germination.

Methods
A full description of the methods for data collection are provided in the manuscript associated with this Dryad submission. Briefly, we used field surveys and experiments to observe components of above- and below-ground demography for 20 populations across the range of Clarkia xantiana ssp. xantiana. To collect data on seedling survival, fruit production, and seed set, we used field surveys. In each population, these surveys included observations in both permanent plots as well as additional, haphazardly sampled plots arrayed across the population. To observe emergence of seedlings and seeds remaining intact in the soil seed bank, we conducted field experiments, which were complemented with lab experiments in order to assay viability of seeds. Brief details for each dataset are provided below, and the manuscript associated with these datasets describes the survey and experimental methods in further detail. We also estimated cumulative precipitation from February-June at each of the 20 populations, following methods described in a manuscript previously published on this study system (Eckhart et al. 2011).

The README.md file describes the contents of each file in this Dryad repository. Two metadata files provide additional information about the files in this Dryad submission: (1) attributes.csv is a comma-separated value file that lists and defines each variable in all the data files, and (2) creators.csv is a comma-separated value file that lists the co-authors on the manuscript associated with this Dryad submission. The metadata files were created using the R package dataspice (Boettiger et al. 2021). 

countFruitsPerPlantAllPlants.csv
Observations of total fruit equivalents per plant from field surveys in 2006-2012. Data come from up to 15 plants per permanent plot, plus plants found across the site. The counts are of “undamaged fruit equivalents” per plant, and were made by counting the number of undamaged fruits, and then counting the damaged fruits and estimating how many undamaged fruits these damaged fruits amounted to.

variableName
description

site
Site acronym

countFruitNumberPerPlant  
Count of fruits on a plant; the count combined undamaged and damaged fruits into a single value when fruits were surveyed in the field

permanentPlot
Binary variable indicating whether the plant on which fruits were counted was growing in one of 30 permanent plots at the site (1=true) or located in either an additional, haphazardly located plot or found by surveying the site (0=false)

damage
Binary variable indicating whether the fruit that was used to count seeds was undamaged (0=false) or damaged (1=true); all observations in this dataset have NA here because the count is of total fruit equivalents

year
Year in which observations were made

countFruitsPerPlantFromPermanentPlots.csv
Observations of total fruit equivalents per plant from field surveys in 2007-2012. Data come from up to 15 plants per permanent plot. The counts are of “undamaged fruit equivalents” per plant, and were made by counting the number of undamaged fruits, and then counting the damaged fruits and estimating how many undamaged fruits these damaged fruits amounted to.

variableName
description

site
Site acronym

transect
Transect number

position
Plot number

plantNumber
Variable indexing the plants on which fruits per plant was recorded; the index starts at 1 in each permanent plot in each year at each site

countFruitsPerPlant  
Count of fruits on a plant; the count combined undamaged and damaged fruits into a single value when fruits were surveyed in the field

year
Year in which observations were made

countSeedPerFruit.csv
Observations of seeds per fruit from seeds collected in the field in 2006-2020. From 2006-2020, data are for counts of seeds per undamaged fruit for up to roughly 30 fruits per site. From 2013-2020, the data also include the counts of seeds per damaged fruit.

variableName  
description

site
Site acronym

year
Year in which observations were made

damaged
Binary variable indicating whether seeds were counted in a fruit that was undamaged (0=false) or damaged (1=true)

sdno
Count of seeds

countUndamagedDamagedFruitsPerPlantAllPlants.csv
Observations of undamaged and damaged fruits per plant from field surveys in 2013-2020. Data come from up to 15 plants per permanent plot, plus plants found across the site. The counts are of “undamaged fruits” and "damaged fruits" on each plant.

variableName
description

site
Site acronym

countUndamagedFruitNumberPerPlant  
Count of undamaged fruits on a plant

countDamagedFruitNumberPerPlant
Count of damaged fruits on a plant

permanentPlot
Binary variable indicating whether the plant on which fruits were counted was growing in one of 30 permanent plots at the site (1=true, 0=false)

year
Year in which observations were made

countUndamagedDamagedFruitsPerPlantFromPermanentPlots.csv
Observations of undamaged and damaged fruits per plant from field surveys in 2013-2020. Data come from up to 15 plants per permanent plot. The counts are of “undamaged fruits” and "damaged fruits" on each plant.

variableName
description

site
Site acronym

transect
Transect number

position
Plot number

plantNumber
Variable indexing the plants on which fruits per plant was recorded; the index starts at 1 in each permanent plot in each year at each site

countUndamagedFruitsPerPlant  
Count of undamaged fruits on a plant

year
Year in which observations were made

countDamagedFruitsPerPlant
Count of damaged fruits on a plant

seedBagsData.csv
Observations of intact seeds and germinants from the seed bag burial experiment, conducted from October 2005-October 2008. The experiment consisted of 3 experimental rounds, starting in October 2005, 2006, and 2007. The data in this file includes counts of intact seeds and germinants. The data for the lab component of the seed bag burial experiment is found in viabilityData.csv.

variableName  
description

site
Site acronym

transect
Transect number

position
Plot number

bagNo
Bag number

round
Experimental round in which the bag was buried

yearStart
Year in which the experimental round started; bags were buried In October of this year

age
Number of years the bag had been buried when it was collected from the field in October

yearData
Year in which observations were made for this bag; the bag was dug up in January and October of this year

seedlingJan
Count of seedlings in the bag when it was dug up in January

intactJan
Count of intact, ungerminated seeds in the bag when it was dug up in January

totalJan
Sum of the count of seedlings and intact, ungerminated seeds in the bag when it was dug up in January

intactOct
Count of intact seeds in the bag when it was dug up in October

seedlingFruitingPlantCountsPermanentPlots.csv
Observations of seedlings and fruiting plants in permanent plots from 2006-2020. Data are counts of seedlings in permanent plots in January/February censuses, and counts of fruiting plants in permanent plots in June/July censuses.

variableName
description

site
Site acronym

transect
Transect number

position
Plot number

year
Year in which observations were made

seedlingNumber  
Count of seedlings in 0.5 m x 1 m permanent plot; NA if not recorded

fruitplNumber
Count of fruiting plants in 0.5 m x 1 m permanent plot; NA if not recorded

siteAbioticData.csv
Summary of abiotic variables associated with each study site in the long-term study of Clarkia xantiana ssp. xantiana demography.

variableName  
description

site
Site acronym

siteName
Full name for site

easting
Geographic position reported as eastward measured distance, reported for UTM zone 11, NAD 1927 (units of 100 meters)

northing
Geographic position reported as northward measured distance, reported for UTM zone 11, NAD 1927 (units of 100 meters)

elevation
Height above sea level (meters)

area
Size of the Clarkia xantiana ssp. xantiana population at the study site (hectares)

surfaceRock
Dominant soil parent material

springPrecipitationData.csv
Cumulative spring precipitation at each of the study sites and years in the long-term study of Clarkia xantiana ssp. xantiana demography. A full description of the methods used to estimate cumulative precipitation at the study sites was previously published (Eckhart et al. 2011).

variableName  
description

site
Site acronym

year
Year in which observations were made

springPrecipitation
Cumulative precipitation from February to June, estimated from daily rainfall measurements at weather stations in the field and interpolated to the location of study populations (units of millimeters) 

viabilityData.csv
Observations from lab germination assays and viability assays that are associated with the seed bag burial experiment. The lab experiments were conducted with seeds from seed bags that were recovered from the field in October 2006, 2007, and 2008. The data for the field component of the seed bag burial experiment is found in seedBagsData.csv.

variableName  
description

site
Site acronym

bagNo
Bag number

round
Experimental round in which the bag was buried

age
Number of years the bag had been buried when it was collected from the field in October

block
Block number for the Petri dishes used in germination trials

germStart
Count of seeds starting the germination trials

germCount
Count of seeds germinating over the course of the germination trials

viabStart
Count of seeds tested in the viability trials

viabStain
Count of seeds staining red in the viability trials

Works cited

Boettiger, C., Chamberlain, S., Fournier, A., Hondula, K., Krystalli, A., Mecum, B., Salmon, M., Webbink, K., and Woo, K. (2021). dataspice: Create Lightweight Schema.org Descriptions of Data. R package version 1.1.0. https://CRAN.R-project.org/package=dataspice

Eckhart, V. M., M. A. Geber, W. F. Morris, E. S. Fabio, P. Tiffin, and D. A. Moeller. (2011). The Geography of Demography: Long-Term Demographic Studies and Species Distribution Models Reveal a Species Border Limited by Adaptation. The American Naturalist 178:S26–S43. https://doi.org/10.1086/661782