Niche construction mediates climate effects on recovery of tundra heathlands after extreme event

Seed rain study Seed rain samples were collected annually from all five sites, between 2010 and 2016, by using one large green plastic doormat (FinnTurf™, 45x55 cm) per block (total of 10 doormats per site), stuck to the ground with 10 cm nails [40]. The mats were put out from early June and to mid-September to sample the current year seed rain. When collected, the mats were bagged and taken inside to dry at about 30 ˚C. Each mat was then turned up-side down and, with the help of a plastic hammer, the contents were removed and each sample was cleaned for E. nigrum leaves using a magnifying microscope, due to their phytotoxicity on seedling growth [16, 41]. The samples were then placed on petri dishes with a Whatman no. 4 filter paper for moisture and put in growing chambers with optimal conditions for seed germination. Emerg- ing seedlings were identified as dicotyledon or monocotyledon and removed from the petri dish. After 12 weeks, the samples were given a stratification period in a dark chamber at 2˚C for four weeks. Thereafter, they were placed again in growing chambers with optimal condi- tions and emerging seedlings were again recorded for 12 more weeks. Seed addition study in the field In autumn 2011, seeds of 7 vascular plant species were sown in all 144 plots, with help of the wooden frame used for vegetation sampling (16 subplots, 0.0125 m2 each). Seeds were sown in 7 of the 16 subplots, with one species per subplot (Fig 1). The species sown were (1) Avenella flexuosa (ca.100 seeds), (2) Solidago virgaurea (ca.100 seeds), (3) Calamagrostis phragmitoides

(ca.100 seeds), (4) Chamerion angustifolium (ca.100 seeds), (5) Bistorta vivipara (30 seeds), (6) Rhinantus minor (ca.50 seeds), and (7) Melampyrun pratense (5 seeds). Thus, we sowed a total of 485 seeds per plot and over 80.000 seeds across all sites. We chose faster growing species than E. nigrum, some that are known to grow together with E. nigrum (A. flexuosa, S. vir- gaurea, B. vivipara and C. phragmitodes) and some species that are not usually found in E. nigrum heathlands (C. angustifolium, R. minor, M. pratense). Seeds had been collected during autumn 2010 and stored in paper bags at 4 ˚C until the experiment took place. Seed germina- tion viability was tested before they were sown out in the field and was for all species between 85% and 95%. Seedlings and established individuals were recorded each year at the end of August between 2011 and 2016. Bioassay study in the phytotron Eight years after simulating the extreme event we evaluated the soil quality in the plots using seedling growth of four species as a bioassay. In autumn 2016, the top 5 cm of soil was sampled with a soil core (2.5cm diameter) from all plots. The soil cores were kept in plastic bags at -18˚C until the experiment took place. We chose half of the soil cores, from blocks 1,3,5,7 and 9, to have a representative sample per site. Soil was thawed and placed in plastic trays with 48 square pots of 5.5 cm x 5.5 cm. First, each pot was filled halfway with commercially available sand. Then the field collected soil was homogenized and put on top. We had 25 control sam-

ples where peat of commercially available quality was placed on top of the sand. Each hole was divided in three sections and three seedlings of A. flexuosa, S. virgaurea and B. vivipara were placed in each section. Seeds were germinated about one week before the experiment started and seedling growth was arrested in cold chambers at 2˚C once the root growth was 0.5 cm. The trays were placed in growth chambers under 24h of simulated daylight (Philips TLD 489 fluorescent tubes) with 12h at 18˚C and 12h at 12˚C. Seedlings were monitored every third day and water sprayed when necessary. The total shoot length of each seedling was measured after three weeks. PLOS ONE Niche construction effects on tundra after extreme event PLOS ONE | https://doi.org/10.1371/journal.pone.0245929 February 4, 2021 6 / 14