Ongoing climate change poses an increasing threat to biodiversity. To avoid decline or extinction, species need to either adjust or adapt to new environmental conditions or track their climatic niches across space. In sessile organisms such as plants, phenotypic plasticity can help maintain fitness in variable and even novel environmental conditions and is therefore likely to play an important role in allowing them to survive climate change, particularly in the short term. Understanding a species’ response to rising temperature is crucial for planning well-targeted and cost-effective conservation measures. We sampled seeds of three Hypericum species (H. maculatum, H. montanum, and H. perforatum), from a total of 23 populations originating from different parts of their native distribution areas in Europe. We grew them under four different temperature regimes in a greenhouse to simulate current and predicted future climatic conditions in the distribution areas. We measured flowering start, flower count, and subsequent seed weight, allowing us to study variations in the thermal plasticity of flowering phenology and its relation to fitness. Our results show that individuals flowered earlier with increasing temperature, while the degree of phenological plasticity varied among species. More specifically, the plasticity of H. maculatum varied depending on population origin, with individuals from the leading range edge being less plastic. Importantly, we show a positive relationship between higher plasticity and increased flower production, indicating adaptive phenological plasticity. The observed connection between plasticity and fitness supports the idea that plasticity itself may be adaptive. This study underlines the need for information on plasticity for predicting species' potential to thrive under global change and the need for studies on whether higher phenotypic plasticity is currently being selected for as natural populations experience a rapidly changing climate.

Study species and populations

Hypericum maculatum (Crantz), H. montanum (L.) and H. perforatum (L.) are perennial herbs native to Europe. The study species share many characteristics, such as yellow flowers and leaf arrangement. The native ranges of H. maculatum and H. perforatum extend from southern to Northern Europe and they are commonly found in grassland habitats (GBIF Secretariat, 2022; GBIF Secretariat, 2022). Compared to H. maculatum and H. perforatum, H. montanum has a more limited range, especially at higher latitudes, and is a habitat specialist occurring more scarcely and mainly in woodlands (GBIF Secretariat, 2022). Out of the three species, H. montanum occurs within the narrowest range (range = 2.65ᵒC, mean = 0.89ᵒC) while having the highest average of annual temperatures (based on average calculated over the years 1970-2000).

We obtained seeds, subject to availability, from across the study species´ European distribution areas to represent the trailing (southern distributional edge), core, and leading (northern distributional edge) parts of the species distributions. We chose seed accessions, i.e., seeds collected from the same location at the same time, among the accessions available in managed seed banks (e.g., Millennium Seed Bank, The European Native Seed Conservation Network ENSCONET partners (Eastwood & Rivière, 2009)) and augmented them with populations collected afresh. In total, we included 23 populations in the experiment: six H. maculatum, six H. montanum, and 11 H. perforatum populations. 

Temperature treatments and greenhouse cultivation

We grew the plants under common garden conditions in greenhouses of the Viikki Plant Growth Facilities, University of Helsinki, from December 2021 to May 2022. In May, we transferred the plants outside for seed maturation. We collected seeds from the experimental individuals during summer and weighed them from August to October 2022.

In the greenhouses, we grew the plants under four different temperature treatments with two replicates for each, totaling eight greenhouse compartments, and each species on a separate table within each compartment. The daytime (16h) temperatures were set to 16°C (“Cold”), 20°C (“Medium”), 24°C (“Warm”), and 28°C (“Hot”). The choice of temperature treatments was loosely based on data on average summer temperatures at the trailing-, core-, and leading distributional areas of the study species from the climatic information service WorldClim (https://www.worldclim.org). The night-time temperatures (8h) were set at 8°C and 10°C below the daytime temperature at the germination and vegetative stages, respectively. Photoperiod in all treatments was 16/8h light/dark. In addition to the automated temperature settings, we monitored realized temperature conditions at the plant level using temperature loggers (Lascar EL-USB-2-LCD+). The realized temperatures were somewhat higher than the set temperatures, particularly as solar radiation increased with the advancing spring, but the differences between treatments remained approximately equal.

On the 1st of December 2021, we sowed 25 seeds per population and replicate into trays filled with the sowing mixture (Kekkilä kylvöseos W HS R8017; KEK31116) with seeds of two populations sown on each tray separated by a border of cardboard and sand. We then placed the trays on a water retaining rug on greenhouse growing tables and its soil topped with coarse sand after sowing the seeds. In the beginning of February 2022, we randomly chose up to ten (depending on availability) of the germinated plants per population, treatment, and replicate to be included in the experiment and transferred them from trays to individual 1L pots filled with soil (Kekkilä Professional Karkea Ruukutusseos; KEK33933) and we then placed the pots on a water retaining rug on greenhouse growing tables. At the same time, we propped up the plants on a support stick if the plant was large enough. We marked all plant individuals with QR-coded ID tags. We separated the individuals of each population into replicates A and B, each with up to ten plants. To avoid microclimatic biases, we periodically rotated both the germination trays and later the plant pots. We regularly fertilized the plants, approximately every 3 weeks with fertilizer solution.   

Watering was implemented by an automated watering system with treatment-specific schedules to keep all experimental plants equally moist. The watering system of H. perforatum in the “Hot”-treatment, replicate A, broke at the end of March leaving the plants dry for some days, which we take into account in the interpretation of the results.

Data collection

Starting two months after sowing, we collected flowering phenology data by monitoring the plants two times per week and recording the date of the first observation of an open flower for each plant individual.

During a four-week data collection phase at peak flowering (April-May), we counted flowers at different stages (i.e., bud, flowering, flowered, seed capsule), except for H. maculatum and H. montanum populations which could not be counted in the “Medium” and “Cold” treatments due to limited time. For each study individual, up to five seed capsules or withered flowers were marked to ensure a matching collection of seeds after ripening. After the flowering counts, we transferred the plants to outdoor conditions at the Kumpula Botanic Garden to allow the seeds to ripen over the summer, whereafter we collected them. We dried the collected seeds for a minimum of five days in RH 15% and cleaned them using sieves (800µm and 250µm), after which we counted and weighed them. We used average seed mass (total seed mass / total seed number) as a measure of reproductive output, as some of the seeds could have dispersed before the capsules were collected, rendering the total seed number an unreliable proxy measure of fitness.

References

Eastwood, R., & Rivière, S. (2009). The ENSCONET Virtual Seed Bank. ENSCONEWS, 5.