A test of local adaptation to drought in germination and seedling traits in populations of two alpine forbs across a 2000 mm/year precipitation gradient

Abstract

Seed regeneration is a critical stage in the life histories of plants, affecting species' abilities to maintain local populations, evolve, and disperse to new sites. In this study, we test for local adaptations to drought in germination and seedling growth of two alpine forbs with contrasting habitat preferences: the alpine generalist Veronica alpina and the snowbed specialist Sibbaldia procumbens. We sampled seeds of each species from four populations spanning a precipitation gradient from 1200 to 3400 mm/year in western Norway. In a growth chamber experiment, we germinated seeds from each population at 10 different water potentials under controlled light and temperature conditions. Drought led to lower germination percentage in both species, and additionally, slower germination, and more investment in roots for V. alpina. These responses varied along the precipitation gradient. Seeds from the driest populations had higher germination percentage, shorter time to germination, and higher investments in the roots under drought conditions than the seeds from the wettest populations - suggesting local adaption to drought. The snowbed specialist, S. procumbens, had lower germination percentages under drought, but otherwise did not respond to drought in ways that indicate physiological or morphological adaptions to drought. S. procumbens germination also did not vary systematically with precipitation of the source site, but heavier-seeded populations germinated to higher rates and tolerated drought better. Our study is the first to test drought effects on seed regeneration in alpine plants populations from high-precipitation regions. We found evidence that germination and seedling traits may show adaptation to drought even in populations from wet habitats. Our results also indicate that alpine generalists might be more adapted to drought and show more local adaptations in drought responses than snowbed specialists.

Keywords: biomass allocation; polyethylene glycol; seed mass; time to 50% germination; time to max germination.

FIGURE 1

Geographical and climatic information about the location for seed collection of the two study species Veronica alpina and Sibbaldia procumbens. The sites are situated in southwestern Norway ranging from the wet coastal site with annual mean precipitation of 3402 mm/year to the drier inland site with 1226 mm/year. Precipitation is calculated based on daily means from 2009 to 2019 with data from the Norwegian Meteorological Institute (met.no). Soil moisture at the sites is obtained from the permanent loggers at the sites during the growing season from 2009 to 2019. Photos: Ragnhild Gya.

FIGURE 2

Germination percentage (number of seeds germinated/viable seeds in the Petri dish) for (a) Veronica alpina and (b) Sibbaldia procumbens across different water potentials (MPa). −0.25 MPa represents pure agar medium with no drought treatment (grey background color). Decreasing water potential yields increasing levels of drought. The colors are based on the average annual precipitation from 2009 to 2019 of the populations the seeds were sampled from, going from light blue in the driest to dark blue in the wettest population. Seeds were sampled at alpine sites in Western Norway. The lines represent predicted means from the model with 95% credible interval envelopes.

FIGURE 3

Time to 50% germination (in days) for (a) Veronica alpina and (b) Sibbaldia procumbens across different water potentials (MPa). Note the different scales on the y‐axis. −0.25 MPa represents pure agar medium with no drought treatment (grey background color). The decreasing water potential yields increasing levels of drought. The colors are based on the average annual precipitation from 2009 to 2019 of the populations the seeds were sampled from, going from light blue in the driest to dark blue in the wettest population. Seeds were sampled at alpine sites in Western Norway. The lines represent predicted means with 95% credible interval envelopes.

FIGURE 4

Root–shoot ratio for (a) Veronica alpina and (b) Sibbaldia procumbens across different water potentials (MPa). Note the different scales on the y‐axis. −0.25 MPa represents pure agar medium with no drought treatment/grey backgorund color). The inserts show all the data (the models are built on all the data), while the large panels zoom in on the patterns in the data excluding some outliers (showing 96% of the data for V. alpina and 98% for S. procumbens). The colors are based on the average annual precipitation from 2009 to 2019 of the populations the seeds were sampled from, going from light blue in the driest to dark blue in the wettest population. Seeds were sampled at alpine sites in Western Norway. The lines represent predicted means with 95% credible interval envelopes.

FIGURE 5

Seed mass per seed for (a) Veronica alpina and (b) Sibbaldia procumbens from populations across a precipitation gradient. The colors are based on the average annual precipitation from 2009 to 2019 of the populations the seeds were sampled from, going from light blue in the driest to dark blue in the wettest population. Seeds were sampled at alpine sites in Western Norway. Note the different scales on the y‐axis. Letters are based on significant differences between populations in a generalized linear model.

FIGURE 6

Germination percentage (number of seeds germinated/viable seeds in the Petri dish) for (a) Veronica alpina and (b) Sibbaldia procumbens across different water potentials (MPa). −0.25 MPa represents pure agar medium with no drought treatment (grey background color). The decreasing water potential yields increasing levels of drought. The colors are based on the seed mass of the different populations within each species, where dark brown represents the population with the heaviest seeds and beige the population with the lightest seeds (note the difference in seed size between species). Seeds were sampled at four alpine sites in Western Norway, across a large precipitation gradient. Matching the seed mass to precipitation for V. alpina, the population with 0.0632 mg seed mass comes from the population with an annual precipitation average between 2009 and 2019 of 3402 mm/year, 0.0644 mg from 1226 mm/year, 0.0688 mg from 2130 mm/year and 0.0692 mg from the population with 1561 mm/year. For S. procumbens, the population with seeds of 0.461 mg comes from the population at 1226 mm/year, 0.478 mg from 2130 mm/year, 0.534 mg from 3402 mm/year, and 0.569 mg from 1561 mm/year. The lines represent predicted means with 95% credible interval envelopes.