# Data from: Spatiotemporal interaction of risk-spreading strategies for a seed-dimorphic plant

## Cite this dataset

Gan, Qian; Liao, Huixuan; Liu, Jingyu; Peng, Shaolin (2024). Data from: Spatiotemporal interaction of risk-spreading strategies for a seed-dimorphic plant [Dataset]. Dryad. https://doi.org/10.5061/dryad.612jm64bj

## Abstract

For plants, the risk of stress-induced reproductive failure can be spread temporally and spatially by increasing the variation in stress tolerance and dispersibility, respectively. Yet, we have limited understanding on how intraspecific stress tolerance can be interactively adjusted by these two risk-spreading strategies. Seed-dimorphic plants, which are frequently found in stressful and fluctuating environments, can produce dimorphic offspring with different stress tolerances and dispersibilities. Using a seed-dimorphic Asteraceae,* Synedrella nodiflora*, we illustrated how maternal-drought-induced changes in drought tolerance and dispersibility may interactively increase intraspecific drought tolerance variation, and thus potentially improve adaptation to various water conditions. Maternal drought stress increased the mean or variation of drought tolerance of the less-dispersible (R-type) offspring, whereas it had variable effects on the drought tolerance of the more-dispersible (D-type) offspring depending on maternal seed morph (R-type vs. D-type) and habitat (dry vs. wet). Because the relative abundances of these two types of offspring remained largely unchanged by maternal drought, there was an overall increase in the intraspecific mean and variation of drought tolerance under maternal drought stress, driven by the changes in the less-dispersible offspring.

Synthesis. Our results indicate a strong interaction between stress tolerance variation and seed dispersibility in seed-dimorphic plant species. Using drought tolerance as an indicator, we demonstrate a potential pathway for the adaptive evolution of seed-dimorphic plants that may contribute to their wide distribution. More importantly, our findings highlight the ecological significance of seed dimorphism and the variation of stress tolerance to avoid reproductive failure under various environmental conditions.

## README: Spatiotemporal interaction of risk-spreading strategies for a seed-dimorphic plant

https://doi.org/10.5061/dryad.612jm64bj

### Description of the data and file structure

**1. Seed mass of maternal plants.csv**

Dataset for the dry mass of seeds collected from field.

Population = population origin

Seed morph = seed morph

Maternal family ID = ID of the maternal family

Seed mass per grain (mg) = seed mass per grain (mg)

**2. Offspring drought tolerance.csv**

Dataset for the drought tolerance of F2 generation.

Population = population origin

Maternal seed morph = seed morph of the maternal individual

Maternal family ID = ID of the maternal family

Maternal environment = the water treatment imposed on the maternal individual

Offspring seed morph = seed morph of the offspring individual

Total biomass under drought-stressed treatment (mg) = offspring total biomass under drought-stressed treatment (mg)

Total biomass under well-watered treatment (mg) = offspring total biomass under well-watered treatment (mg)

Mean total biomass under well-watered treatment (mg) = mean total biomass across offspring of each combination of F0 maternal plant habitat (dry vs. wet) x F0 maternal plant ID (1-4) x F1 maternal environment (drought vs. well-watered) x F1 seed morph (R- vs. D-type) x F2 seed morph (R- vs. D-type) under well-watered treatment (mg) Drought tolerance = offspring drought tolerance

Missing data code- NA

**3. Offspring drought tolerance and seed number for data simulation**

The mean, standard deviation, and seed number for each combination of F0 maternal plant habitat (dry vs. wet) x F0 maternal plant ID (1-4) x F1 maternal environment (drought vs. well-watered) x F1 seed morph (R- vs. D-type) x F2 seed morph (R- vs. D-type).

Population = population origin

MS = maternal seed morph (seed morph of the maternal individual)

ME = maternal environment (the water treatment imposed on the maternal individual)

OS = offspring seed morph (seed morph of the offspring individual)

Family = ID of the maternal family

Mean = mean drought tolerance

SD = standard deviation of drought tolerance

SN = seed number

**4. Species-level drought tolerance simulation**

Datasets for species-level drought tolerance simulation.

This file contains seven tabs corresponding to seven presumed proportions of maternal plants that would be exposed to drought stress (i.e., 20%, 25%, 33%, 50%, 67%, 75%, 80%). File structures are the same for all tabs.

Population = population origin

MS = maternal seed morph (seed morph of the maternal individual)

ME = maternal environment (the water treatment imposed on the maternal individual)

OS = offspring seed morph (seed morph of the offspring individual)

Family = ID of the maternal family

Simulation #1-#5 = the five replicates of data simulations

## Methods

**1. Experimental data**

(1) Seed collection in field

The mature seeds of* **Synedrella* from two distinct natural habitats in Guangzhou and Zhuhai, China, were collected in January and February 2018 (see details in supporting information Table S1). While both habitats were beneath trees along roadsides, one had a relatively dry environment with a *Synedrella* monoculture (hereafter the dry habitat), while the other was close to a stream hosting a species-diverse plant community (hereafter the wet habitat). In each habitat, we collected seeds from four individuals (F0) of similar size. The collected seeds were carefully separated by seed morphs, which were then air-dried and stored in paper bags at room temperature (approximately 20 ℃) for about four months prior to the greenhouse experiment. In total, we collected 16 replicated seed sources (2 seed morphs × 4 F0 individuals × 2 habitats).

(2) Greenhouse experiment

A greenhouse experiment was conducted at Sun Yat-sen University, Guangzhou, China (23°03ʹN, 113°11′E). The mature *Synedrella* seeds collected in the field of maternal plants of generation F0 were surface sterilized by soaking in 1% sodium hypochlorite solution for 15 min and then incubated on moist double-layer filter paper in Petri dishes in a growth chamber (25 ℃ 12 h day / 20 ℃ 12 h night). Germinated seeds were transplanted into seedlings trays. In June 2018, four individuals of similar size were selected and transplanted into 3 L pots with one plant per pot (i.e., F1 generation). These 3 L pots were filled with a 1:1 mixture of horticultural sand: peat with 8 g of slow-release fertilizer per pot (Osmocote 5#, Scotts, Marysville, OH, United States, with a 14N: 13P: 13K elemental ratio, duration 5-6 months). Before soil water manipulations, all seedlings received adequate watering for one week with 45% soil moisture by volume (equivalent to 100% field capacity). Then, two of the four individuals from each original source (2 seed morphs × 4 F0 individuals × 2 habitats) were randomly assigned to well-watered and drought-stressed treatments (Fig. S1). For well-watered treatment, the pots were regularly watered twice a day to keep soil moisture saturated, when soil moisture by volume was maintained at 45%. The drought-stressed treatment was manipulated by not watering the plants for about one week until their leaves appeared wilt, when the soil moisture by volume dropped to about 9% (equivalent to soil relative water content of 20%). The soil moisture for the drought-treated pots was then maintained by weighing the pots and supplementing the water losses every day. As a result, a total of 64 parental plants (2 replicates × 2 maternal seed morphs × 4 F0 individuals × 2 habitats × 2 water treatments; Fig. S1) were grown to reproductive maturity. The positions of the plants were shuffled weekly. Manual pollination was carried out between the two replicated parental plants to ensure a high fertilization percentage. The mature seeds detached from withered seed heads were collected in October 2018 (120 days after germination), when the plants began to senesce. All seeds were air-dried and stored at room temperature (approximately 20 ℃) until further experiment. The dimorphic seeds were separated, weighted and counted for each F1-generation plant.

In October 2020, mature seeds of the F1 generation were germinated using the same protocol as those of the F0 generation. Three days after germination, 5 pairs of these F2-generation seedlings of similar sizes, a total of 640 seedlings, were transplanted into seedling trays filled with a 1:1 mixture of horticultural sand: peat. Seedlings received adequate watering with 45% soil moisture by volume (equivalent to 100% field capacity) for 48 h. Then half of the F2-generation seedlings were subjected to well-watered treatment, while the other half were subjected to drought treatment. Both soil water treatments were manipulated using the same protocol as that for the F1 generation. For these F2-generation plants, they were grown for 21 days (Sultan et al., 2009). Due to mortality from drought stress, only 591 seedlings were harvested at the end of experiment. The individuals were oven-dried at 72 ℃ for 3 days and weighed.

(3) Calculation of drought tolerance

For each drought-stressed seedling from each combination of maternal seed morph, maternal individual, and habitat, its drought tolerance was calculated following the equation by Valliere et al. (2019):

Drought Tolerance = (Total Biomass _{drought} – Mean Total Biomass _{well-watered}) / Mean Total Biomass _{well-watered}

Here, Total Biomass _{drought} is the total biomass of focal drought-stressed seedling, and Mean Total Biomass _{well-watered} is the averaged biomass of all well-watered seedlings from the same combination of maternal seed morph, maternal plant, and habitat. A high negative value corresponds with a weak drought tolerance, while a low negative value corresponds with a strong drought tolerance. Sometimes, a positive value occurred, indicating that the seedling performs better under drought than in well-watered condition.

For each seed morph of each maternal plant, its mean and variation of drought tolerance were calculated as the mean and standard deviation (SD) of the five replicated drought-stressed seedlings.

**2. Simulation data**

Given the ubiquitous spatial heterogeneity of water conditions, some *Synedrella* maternal plants may experience drought stress while the rest may live in well-watered conditions. Because different maternal environments may result in differential reproductive patterns in terms of dimorphic offspring production, the intraspecific mean and variation of drought tolerance are likely to change in response to the altered spatial frequency of maternal drought stress. Thus, we generated a simulated dataset of offspring drought tolerance to determine the intraspecific mean and variation of drought tolerance along a gradient where different proportions of maternal plants were exposed to drought stress. Under the assumption that all seeds will successfully germinate, we simply manipulated the abundance of dimorphic seeds based on: (1) the numbers of the D- and R-type seeds produced under well-watered and drought-stressed treatments by each maternal plant from the focal population; and (2) the presumed proportions of maternal plants that would be exposed to drought stress (i.e., 80%, 75%, 67%, 50%, 33%, 25%, 20%).

Step by step simulation of intraspecific drought tolerance along a gradient of spatial frequency of drought stress is as follows:

(1) Determine the data distribution of offspring drought tolerance for simulation

Basing on the experimental data of dimorphic offspring drought tolerance, we first calculated the mean and standard deviation of offspring drought tolerance. Since we have a total of 32 combinations between F0 habitat (dry vs. wet) x F0 plant ID (1-4) x F1 environment (drought vs. well-watered) x F1 seed morph (R- vs. D-type), we calculated the mean and standard deviation of offspring drought tolerance for each combination. For each combination, we assumed that the drought tolerance of all D- and R-type offspring produced by this combination followed a normal distribution defined by the mean and standard deviation quantified using the above-mentioned equations.

(2) Determine the numbers of dimorphic offspring used for each simulation scenario

Because the numbers of dimorphic offspring produced by each combination differed under different simulation scenarios, we need to determine the numbers of dimorphic offspring used for each simulation scenario. Basing on the experimental data of dimorphic offspring production by each maternal plant under drought-stressed and well-watered treatment, we first determined the numbers of dimorphic offspring used for each simulation scenario. For instance, for the scenario of 50% probability of maternal drought stress, we assigned 50% of maternal plants to drought treatment. For the scenario of 20% probability of maternal drought stress, we assigned 20% of maternal plants to drought treatment. For convenience, we assign an ID to each combination: the combinations associated with maternal drought-stressed treatment were assigned as combinations 1-16, while those associated with maternal well-watered treatment were assigned as combinations 17-32. When *x*% of F1 maternal plants were assigned to drought treatment, the relative proportion of each type of dimorphic offspring produced by the *i ^{t}*

^{h}combination of can be calculated using the following equation:

*Seed Number _{D-type}* =

*x%*

*×*

*Seed Number*+ (1-

_{D-type, drought-stressed}*x%*)×

*Seed Number*(Eq. 1)

_{D-type, well-watered}*Seed Number _{R-type}* =

*x%*

*×*

*Seed Number*+ (1-

_{R-type, drought-stressed}*x%*)×

*Seed Number*(Eq. 2)

_{R-type, well-watered }Total Seed Number = *Seed Number _{D-type}* +

*Seed Number*(Eq. 3)

_{R-type}For i between 1 and 16:

*Proportion _{D-type,i} *=

*x%*

*×*

*Seed Number*/

_{D-type, i}*Total Seed Number*(Eq. 4)

*Proportion _{R-type,i} *=

*x%*

*×*

*Seed Number*/

_{R-type, i}*Total Seed Number*(Eq. 5)

For i between 17 and 32:

*Proportion _{D-type,i} *= (1-

*x%*)

*×*

*Seed Number*/

_{D-type, i}*Total Seed Number*(Eq. 6)

*Proportion _{R-type,i} *= (1-

*x%*)

*×*

*Seed Number*/

_{R-type, i}*Total Seed Number*(Eq. 7)

where *Seed Number _{D-type}*,

*Seed Number*and

_{R-type}*Total Seed Number*represent the total numbers of D-type, R-type and both types of seeds produced, respectively.

*Seed Number*and

_{D-type,i}*Seed Number*represent the numbers of D- and R-type seeds produced by the

_{R-type,i}*i*

^{th}combination.

*Proportion*and

_{D-type,i}*Proportion*represent the relative proportion of D- and R-type seeds produced by the

_{R-type,i}*i*

^{th}combination, respectively. Then, under the assumption that a total of 1,000,000 offspring will be produced in each scenario, the numbers of dimorphic offspring used for the scenario when

*x*% of F1 maternal plants were exposed to drought stress can be calculated as follows:

*Offspring Number _{D-type,i} *=

*Proportion*×1,000,000 (Eq. 8)

_{D-type,i}*Offspring Number _{R-type,i} *=

*Proportion*×1,000,000 (Eq. 9)

_{R-type,i}where *Offspring Number _{D-type,i}* and

*Offspring Number*represent the numbers of D- and R-type offspring for the

_{R-type,i}*i*

^{th}combination that would be used for data simulation, respectively.

(3) Data simulation

For the *i*^{th} combinations, we used Python version 3.0 (van Rossum & Drake, 2009) to randomly draw a drought tolerance value from the corresponding normal distribution defined by step 1 and repeated the process for the corresponding number of times defined by step 2. This process was repeated for all 32 combinations to get the drought tolerance data for all 1,000,000 offspring individuals. Since there were seven different scenarios (i.e., meternal drought probabilities of 20%, 25%, 33%, 50%, 67%, 75% and 80%), we assigned x to 20, 25, 33, 50, 67, 75 and 80, and repeated the simulation process for each x value. As a result, we obstained seven datasets, each containing 1,000,000 offspring drought tolerance values.

## Funding

National Natural Science Foundation of China, Award: NSFC 32071530

Vegegraphy of China, Award: 2015FY210200-13

Zhang Hongda Science Foundation

Fu Jia-Rui Scholarship