Skip to main content
Dryad logo

Data from: Translocation of an arctic seashore plant reveals signs of maladaptation to altered climatic conditions

Citation

Hällfors, Maria et al. (2020), Data from: Translocation of an arctic seashore plant reveals signs of maladaptation to altered climatic conditions, Dryad, Dataset, https://doi.org/10.5061/dryad.3n5tb2rfk

Abstract

Ongoing anthropogenic climate change alters the local climatic conditions to which species may be adapted. Information on species’ climatic requirements and their intraspecific variation is necessary for predicting the effects of climate change on biodiversity. We used a climatic gradient to test whether populations of two allopatric varieties of an arctic seashore herb (Primula nutans ssp. finmarchica) show adaptation to their local climates and how a future warmer climate may affect them. Our experimental set-up combined a reciprocal translocation within the distribution range of the species with an experiment testing the performance of the sampled populations in warmer climatic conditions south of their range. We monitored survival, size, and flowering over four growing seasons as measures of performance and, thus, proxies of fitness. We found that both varieties performed better in experimental gardens towards the north. Interestingly, highest up in the north, the southern variety outperformed the northern one. Supported by weather data, this suggests that the climatic optima of both varieties have moved at least partly outside their current range. Further warming would make the current environments of both varieties even less suitable. We conclude that Primula nutans ssp. finmarchica is already suffering from adaptational lag due to climate change, and that further warming may increase this maladaptation, especially for the northern variety. The study also highlights that it is not sufficient to run only reciprocal translocation experiments. Climate change is already shifting the optimum conditions for many species and adaptation needs also to be tested outside the current range of the focal taxon in order to include both historic conditions and future conditions.

Methods

Study species and seed sampling

Siberian primrose (Primula nutans) is a small-statured perennial herb with a discontinuous, circumpolar distribution. The Fennoscandian subspecies P. nutans ssp. finmarchica (Jacq.) Á. Löve & D. Löve is a red-listed species (VU in Norway and NT in Finland with a disjunct distribution (Fig. 1). It comprises two morpho-ecological varieties (Mäkinen & Mäkinen, 1964): var. finmarchica (the northern variety) grows on the shores of the Arctic Sea, while var. jokelae (the southern variety) occurs by the Bothnian Bay and the White Sea. It is a habitat specialist that mainly grows in seashore and riverside meadows (Kreivi, Aspi, & Leskinen, 2011; Mäkinen & Mäkinen, 1964). The habitat preference of the Siberian primrose is believed not to be affected by specific habitat requirements - it is neither a halophyte nor does it require regular flooding (Mäkinen & Mäkinen 1964). Rather, its occurrence in these habitats is likely due to its poor ability to compete in combination with a tolerance to both salty conditions and flooding. It propagates sexually by seeds and vegetatively by stolons growing from the wintering buds and the rosettes (Mäkinen & Mäkinen, 1964).  On the basis of matrix models published by Björnström et al. (2011) individuals especially sterile rosettes can be relatively persistent and hence live for several years. The species is insect-pollinated and produces copious seeds that spread by gravitation, via water flows, and possibly also with birds (Ulvinen 1997).

Due to the geographic distance (c. 400-550 km) between its two varieties var. jokelae and var. finmarchica, and the lack of obvious adaptations for efficient gene flow between the areas, the varieties are presumably genetically isolated, which is also reflected in their genetic differences (Kreivi et al., 2011). The main areas where the varieties occur are climatically different (Hällfors et al., 2016), and there is a clear difference in the requirement of colder night conditions for flower induction in the northern variety compared to southern variety (Mäkinen & Mäkinen, 1964).

We sampled seeds from wild populations of both varieties of Primula nutans ssp. finmarchica from August to September 2012. Seeds of the southern variety were collected in five sites in Haukipudas and Ii in Finland, and seeds of the northern variety from six sites in Sør-Varanger in Norway. Seed sampling in Finland was approved by the Centre for Economic Development, Transport and the Environment in North Ostrobothnia, Finland, on July 7th, 2012 (approval number: POPELY/346/07.01/2012). For sampling seeds in Norway, no permit was needed as the species is not protected in Norway. Voucher specimens from each sampling site are deposited at the herbarium of the Finnish Museum of Natural History (H sensu Thiers 2016). From each site we collected seeds from 10 - 53 individuals (as available, following seed collecting guidelines; ENSCONET, 2009; Table S1) to obtain a representative sample. The number of seeds obtained from the collected capsules of each sampled individual varied from 0 to c. 200. We stored the seed lot from each maternal individual separately and left them to dry and ripen at room temperature for a minimum of two weeks after which we placed them in a freezer (-18 to -20°C) for cold stratification to break dormancy. The seeds were kept in these conditions for about six months, until used to produce plant material for the translocation experiment (in spring 2013).

Experimental design and plant material

We set up a common garden experiment in 2013 in five botanic gardens located in Estonia and Finland, and on research station grounds in Norway. We chose to use botanic gardens as testing grounds instead of natural sites to avoid genetic contamination of natural populations by introducing alien genotypes, as well as for logistical reasons and legal restrictions of introducing species outside their natural range.

We chose the seeds for producing experimental plants for the trial through a hierarchical randomized method. For each experimental garden, we selected material from three seed sampling sites per variety. From each site, we randomly chose five maternal individuals. We anticipated that not all seeds would germinate and therefore we sowed eight seeds per maternal individual (120 seeds per variety for each site) although we needed only three seedlings per maternal individual (45 seedlings per variety for each site).

We used the F1 offspring for our tests for two main reasons. Producing a ‘refresher’ generation between collecting the seeds and growing the experimental plants would have been difficult or impossible in uniform conditions as the two varieties flower in different temperatures conditions (Mäkinen & Mäkinen 1964). Second, even in the best case, this would have delayed the experiment by at least one, possibly 2–3 years, due to the life history of the focal species. Nevertheless, a recent meta-analysis (Yin et al. 2019) concluded that perennial plants show hardly any transgenerational responses (i.e., effects on the offspring of the ancestor environmental conditions), whereby it is not likely that the use of F1 offspring significantly affected our results.

We sowed each seed in a 6 × 6 cm pot with commercial sowing soil (Kekkilä) mixed with sand, vermiculite and perlite (5:2:2:0.5 litres, respectively), and with a thin layer of sand on top. The seeds were sown in six cohorts, five weeks before intended planting at each experimental garden. They were left to germinate in a greenhouse in Kumpula Botanic Garden in Helsinki, first in a small greenhouse inside the actual greenhouse to enable suitable temperatures during early spring. Temperature conditions varied between 18 and 24 ⁰C and lighting was set to 12 h of extra light (400 W) during daytime, in addition to the day light reaching the plants through the greenhouse walls.

Due to poor germination and growth, possibly caused by supraoptimal temperature or a lack of nutrients or light, the seedlings were moved out of the small greenhouse into the actual one to allow cooler temperatures and more light (10–18 ⁰C). The moving followed the same order and time intervals as the sowing so that each cohort spent the same amount of time in the small greenhouse. Five days after the move, the pots were given a one-time nutrient addition with a general commercial fertilizer (Kekkilä Taimi-Superex).  This is a common procedure when using nutrient poor germination soil. Without fertilization of the soil after seed germination the growth usually comes to a halt.

For each experimental garden we selected 45 seedlings of each variety, so that each variety was represented by three seed sampling sites, each of which contributed five seedlings from three different maternal individuals. If there were not enough seedlings from a certain maternal individual, we complemented the design with seedlings from another maternal individual of the same seed sampling population (in 69 cases). If there was no other such maternal individual, we used seedlings from another maternal individual from another seed sampling population but of the same variety (in 21 cases).

Because of a delayed spring in 2013, establishing the experimental plots had to be postponed by three weeks to avoid frozen soil and frost that could kill the seedlings after planting. The approximately two-month-old seedlings were transported and planted during May–June 2013, starting from the southernmost garden in mid-May (Tartu, Estonia) and reaching north (Svanvik, Norway) in late June. The intention was to follow the advance of the season so as to plant as early as possible while avoiding frost damage. In each experimental garden, we planted the seedlings in three 165 × 145 × 20 cm experimental plots with a minimum distance of 30 m between the plots, following a randomized block design. A filter cloth was placed at the bottom to prevent the ground soil from affecting the growing conditions and to help retain moisture. The plots were filled with fine sand and peat mixture (75 and 25 volume-%, respectively) with some added dolomite lime. The intention was to roughly mimic the seashore meadow soils where the species grows in the wild while using commercially available soils that could be obtained in large quantities by a commercial soil provider and transferred to the all experimental gardens. The aim with using the same substrate was to homogenise (to certain extent) the growth conditions and thus to be able to separate the thermal effect from other effects.

In each plot, we planted 30 seedlings (i.e., 90 seedlings per garden) in a grid, c. 21 cm from each other and 30 cm from the wooden frame. In each plot, 15 seedlings were of the southern and 15 of the northern variety, placed in a random order. We placed light metal cages on top of each plot to prevent large animals from interfering with the experiment. Plots were irrigated with c. 20 litres of water once a week during the growing season to keep the soil sufficiently moist. Although precipitation is part of climate, the effects of which we attempt to measure, we considered basic watering of the trials necessary as this is a sea-shore plant naturally exposed to high water tables. This watering regime was not intended to directly resemble the seashore meadows in nature, where the water level can vary substantially and occasionally even flood the plants. Instead, we applied irrigation to avoid large effects of drought, which this seashore species growing on occasionally inundated meadows, would not endure. Thus, watering was done only to the amount to compensate for the fact that our test sites were not at the shoreline, i.e. to mimic the natural sites of the primrose up to a certain baseline so that the plants would not die because of an unnatural unfavourable moisture regime. Plots were also weeded when deemed necessary to avoid effects of competition from other species, as our focus was not on the effect of competition on plant performance.

Since almost 50% of the plants died during the first summer (139 individuals of the northern variety and 57 of the southern; likely due to small seedling size and too little irrigation), in the autumn of 2013, the experimental plots were supplemented with left-over plants that had been growing outdoors in Kumpula botanic garden in Helsinki during the summer. We chose new plants repeating the plant selection process described above for situation where there were not enough representatives of the preselected maternal individual for the experimental garden. We recorded the planting time, to enable differentiating between original and new plants in subsequent analyses. Altogether, 614 individuals were planted in the five gardens during the early and late summer visits.

To describe the performance of individuals, we measured their survival, size, and flowering (whether they flowered, how many flowers they produced, and when) from year 2014 to 2016.  The fitness of genotypes can be described as the relative success with which they transmit their genes to the next generation (Silvertown & Charlesworth, 2001). Because of the limited temporal extent of our study and because we do not measure the individual's fitness through its ability to produce viable progeny, we focus on key plant performance measures likely to correlate with fitness. Survival is a definite measure of fitness as a dead individual cannot transmit genes, but survival can also be stochastic. Also, the physical size of an individual and reproductive output, such as abundance of flowering, tend to correlate with plant fitness (Silvertown & Charlesworth, 2001). Each spring (2014, 2015, and 2016) we recorded flowering presence and abundance. Flowering was inventoried approximately every second day for 14 days after the first flower appeared in that garden in the specific year. In the autumn of the same years, we recorded survival and photographed each surviving plant from above. We assessed plant size from photographs, through digitally cutting out and measuring the area (cm2) of the visible vegetative parts. We used the size at the time of planting as a measurement of original size, but 24 individuals lack data on that due to missing photographs or insufficient resolution. These individuals were therefore not included in the analyses, and our total N is reduced from 614 to 590.

Weather and climatic data

For describing the climatic conditions at the seed sampling sites and experimental gardens, we obtained data on historic climatic conditions (1970–2000; 10 min resolution) and future projections (CMIP5 for 2050, 10 minutes resolution, HADGEM2-ES model) through the Worldclim database (Fick & Hijmans, 2017). These climatic data were downloaded in R using the getData function in the raster package (Hijmans, 2019). Additionally, we modelled plant performance as a function of mean annual temperature in the experimental gardens during 2013­-2016. The weather data were obtained from the Gridded Agro-Meteorological Data in Europe (Joint Research Centre, 2014), which contains meteorological parameters from European weather stations interpolated on a 25x25 km grid. We used the function biovar in the package dismo (Hijmans et al., 2017) to calculate bioclimatic variables of the weather data in R.

Occurrence data

To produce Figure 1a in the original manuscript, we obtained occurrence data of the varieties. Var. finmarchica occurs by the Arctic Sea in N-Norway and var. jokelae by the Bothnian bay in Finland and Sweden and the White Sea in Russia. The source for occurrence points are: Global Biodiversity Information Facility (GBIF 2013); Kastikka (Finnish plant distribution database; Lampinen, Lahti, & Heikkinen 2012); unpublished records from the of Finnish Environment Institute; occurrences in Russia based on information from herbarium specimens (from collections in the herbarium of the Finnish Museum of Natural History [H sensu Thiers 2016] and the herbarium of the University of Turku [TUR sensu Thiers 2016] and manually included occurrence points based on visually inspecting the distribution map by Hultén and Fries [1986]).

References:

Björnström, T., Hyvärinen, M., Siikamäki, P., Strengell, H., Väre, H., & Tuomi, J. (2011). Demographic variability and population viability of an endemic variety of the Siberian primrose. Annales Botanici Fennici, 48(2), 107–119. Retrieved from JSTOR.

ENSCONET 2009: ENSCONET Seed Collecting Manual for Wild Species. Available at: http://ensconet.maich.gr/PDF/Collecting_protocol_English.pdf

Fick, S. E., & Hijmans, R. J. (2017). WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas. International Journal of Climatology, 37(12), 4302–4315. doi: 10.1002/joc.5086

GBIF (2013). Global Biodiversity Information Facility Data Portal.

Hällfors, M. H., Liao, J., Dzurisin, J., Grundel, R., Hyvärinen, M., Towle, K., Wu., G., Hellmann, J. J. (2016). Addressing potential local adaptation in species distribution models: implications for conservation under climate change. Ecological Applications, 26(4), 1154–1169. doi: 10.1890/15-0926

Hultén, E., & Fries, M. (1986). Atlas of North European vascular plants: north of the Tropic of Cancer, vol. 1-3. Königstein, Koeltz.

Hijmans, R.J. (2019). raster: Geographic Data Analysis and Modeling. R package version

            2.8-19. https://CRAN.R-project.org/package=raster

Hijmans, R.J., Phillips, S., Leathwick, J. & Jane Elith, J. (2017). dismo: Species

     Distribution Modeling. R package version 1.1-4. https://CRAN.R-project.org/package=dismo

Hultén E., and Fries M. (1986): Atlas of North European vascular plants: north of the Tropic of Cancer, vol. 1-3. Königstein, Koeltz.

Hyvärinen, E., Juslén, A., Kemppainen, E., Uddström, A. & Liukko, U.-M. (eds.) (2019): The 2019 Red List of Finnish Species. (Ministry of the Environment & Finnish Environment Institute, 2019).

Joint Research Centre 2014: Gridded Agro-Meteorological Data in Europe. URI: http://data.europa.eu/88u/dataset/jrc-marsop4-7-weather_obs_grid_2015

Kreivi, M., Aspi, J., & Leskinen, E. (2011). Regional and local spatial genetic structure of Siberian primrose populations in Northern Europe. Conservation Genetics, 12(6), 1551–1563. doi: 10.1007/s10592-011-0252-7

Kreivi, M., Huttunen, S., & Aspi, J. (2006). Isolation and characterization of polymorphic microsatellite markers from Primula nutans (Primulaceae). Molecular Ecology Notes,

Lampinen, R., Lahti, T., Heikkinen, M. 2012. Kasviatlas 2011. -- Helsingin Yliopisto, Luonnontieteellinen keskusmuseo, Helsinki. Distribution maps available at: http://www.luomus.fi/kasviatlas.

Lehtimäki, I. (2016). Uhanalaisen ruijanesikon suojelu. Etäsuojelun ja avustetun leviämisen näkökulma. Master’s thesis [In Finnish]. Department of Biosciences, University of Helsinki.

Mäkinen, L., & Mäkinen, Y. (1964). The distribution, ecology, morphology and taxonomy of Primula nutans Georgi ssp. finmarchica (Jacq.) Löve & Löve. Annales Botanici Fennici, 1(4), 273–291. Retrieved from JSTOR.

R Core Team 2019: R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. <http://www.R-project.org/>.

Silvertown, J., & Charlesworth, D. (2001). Introduction to Plant Population Biology. 4th edition. Blackwell Science.

Thiers, B. (2016) Index Herbariorum: A Global Directory of Public Herbaria and Associated Staff. New York Botanical Garden’s Virtual Herbarium. http://sweetgum.nybg.org/science/ih/

Ulvinen, T. (1997). Rujanesikko. In: Uhanalaiset kasvimme. Ryttäri T. and Kettunen T. (eds.). [In Finnish]. Suomen ympäristökeskus, Helsinki.

Funding

Academy of Finland, Award: 126915

Societas pro Fauna et Flora Fennica

University of Helsinki Research Fund, LUOVA – Doctoral Programme in Wildlife Biology Research

Jane and Aatos Erkko Foundation through the Research Centre for Ecological Change