Skip to main content
Dryad logo

What doesn’t kill you can make you stronger: variation in plasticity in response to early temporally heterogeneous hydrological experience


王, 姝 (2022), What doesn’t kill you can make you stronger: variation in plasticity in response to early temporally heterogeneous hydrological experience, Dryad, Dataset,


Temporally heterogeneous environments may drive rapid and continuous plastic responses, leading to highly variable plasticity in traits. However, direct experimental evidence for such meta-plasticity due to environmental heterogeneity is rare.

Our objective was to investigate the effects of early experience with temporally heterogeneous water availability on the subsequent plasticity of plant species in response to water conditions.

We subjected eight plant species from three habitats, four exotic and four native to North America, to initial exposure to either a first round of alternating drought and inundation treatment (Ehet, temporally heterogeneous experience) or a consistently moderate water supply (Ehom, homogeneous experience), and to a second round of drought, moderate watering or inundation treatments. Afterwards the performance in a series of traits of these species, after the first and second rounds of treatments, was measured.

Compared to Ehom, Ehet increased final mean total mass of all species considered together, but did not affect mean mortality. Ehet relative to Ehom, decreased the initial total mass of native species as a group, but increased the mass of exotic species or species from hydric habitats; Ehet also increased the late growth of natives, but did not for exotics, and increased the late growth of mesic species more than xeric and hydric species.

Our results suggest that previous exposure to temporal heterogeneity in water supply may be not beneficial immediately, but can be beneficial for plant growth and response to water stress later in a plant’s lifetime. Heterogeneous experiences may not necessarily enhance the degree of plasticity, but may improve the expression of plasticity in terms of better performance later, effects of which differ for different groups of species, suggesting species-specific strategies for dealing with fluctuating abiotic environments. 

Synthesis. Previous temporally heterogeneous experience can benefits plant growth later in life though modulating the expression of plasticity, leading to adaptive meta-plasticity. Studies of meta-plasticity may not only improve our understanding of the importance of variable plasticity in relation how plants cope with environmental challenges, but also the costs versus benefits of plastic responses and its limits over the long term.


The species we used in this study were four exotic invasive species - Leucanthemum vulgare Lam., Centaurea stoebe L. ssp. micranthos (Gugler) Hayek, Leonurus cardiaca L. and Potentilla recta L., and four native species - Heterotheca villosa (Pursh) Shinners, Gaillardia aristata Pursh, Agastache urticifolia (Benth.) Kuntze  and Potentilla arguta Pursh (Table S1; also see Wang et al. 2017). Thus, three pairs of exotics and natives shared families and one pair were congeners. All seeds were collected from natural grasslands in western Montana, USA. These species overlap substantially in distributions, but the selected groups of target species generally occur in habitats that share positions on a soil moisture gradient (for details see Wang et al. 2017).

Mortality rates and total mass for all treatment combinations were calculated. Overall, 720 individuals survived to the end of the experiment, and were used for analyses. Traits of total mass, shoot mass and root mass and root to shoot ratio were used to assess the early performance of species, with these traits and late growth in shoot mass, root mass and total mass used to assess their late performance. Late growth (LG) in a trait for each species in all early - late treatment combinations was calculated with the formula (1) (Wang et al. 2017):

LG = (YX) / X                                                              (1)

where X is the mean trait value in one of the two early treatments (the 1st round), and Y is the mean trait value in one of the three late treatments (the 2nd round) with the same early treatment.

Mean values of shoot mass, root mass and total mass were log-transformed, and mortality rates were square root- and arcsine-transformed (by calculating the arcsine of the square root of mortality rate), to minimize variance heterogeneity.

Plasticity in response to late inundation or drought conditions in traits was calculated using the Simplified Relative Distance Plasticity Index (RDPIs, (Valladares, Sanchez-Gomez & Zavala 2006), abbreviated as “PI”. We used the formula (2-1) for mass traits and formula (2-2) for late growth of mass traits to calculate plasticity in a given trait of a species:

PI = (Y2Y1) / Y1                                                             (2-1)

PI = Y2Y1                                                                      (2-2)

where Y2 was the mean trait value in inundation or drought (the 2nd round) after an early  treatment (the 1st round), and Y1 was the mean trait values in moderate water conditions after the same early treatment. The PI value was regarded as significant when mean values in two late water treatments differed at the 0.05 level (LSD method in one-way ANCOVA).

To compare effects of early heterogeneous vs. homogeneous experience among species or late conditions, we defined “difference variables (Diff-value)” as the difference between individuals with the early heterogeneous and homogeneous experiences in a given trait of a  species (Wang et al. 2017), which was calculated with formula (3) as:

Diff-Y = Yhet Yhom                                                        (3)

where Diff-Y was the difference between individuals with early heterogeneous (Ehet) and homogeneous (Ehom) experiences in the performance of a trait in a late condition, Yhom was the mean trait value in the late condition after Ehom (control), and Yhet was the mean trait value in the same late condition after Ehet. For a given trait, Diff-values due to early treatments were regarded as significant when its mean values (in the same late conditions) differed between two early treatments at 0.10 level (LSD method in one-way ANOVA). 

Usage Notes

Microsoft or WPS.


National Science Foundation EPSCoR Cooperative Agreement, Award: 1757351

National Natural Science Foundation of China, Award: 31800335, 32171511