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

event plots as well (Fig 3b). E. nigrum was the most important contributor to the dwarf shrub group in the three most coastal sites (Fig 3b). In the two most continental sites, B. nana and E. nigrum were equally abundant. Regenerative capacity of the E. nigrum heathlands Seed germination from both the seed bank and seed rain was low for all sites (Table 2). We reg- istered no germinating seeds from the seedbank samples in two of the study sites. From the 2011 seed sowing experiment in the field, only three out of the seven species were registered as seedlings during the study (A.flexuosa, S.virgaurea and B.vivipara). Seedling counts were very low and decreased exponentially throughout the years (S1 Fig, S3 Table). On average, we registered a maximum of 5 seedlings per plot during the first year, and about one seedling the last year. The seedlings we registered over the years did not manage to establish into adult individuals. We found no difference between control plots and extreme event plots but found that continental sites had less registered germinating seedlings throughout the years (S1 Fig). The bioassay in the phytotron showed that all four species had arrested growth when sown on soil collected from both extreme event and control plots when compared to soil from com- mercially available peat (S2 Fig). Furthermore, when comparing mean shoot length of all spe- cies in extreme event and control soil alone (excluding seedlings grown in peat), we found no effect of extreme event or continentality index except for a small effect on S. virgaurea (Fig 5b, S2 Table).

Discussion The results from our simulated extreme event on E. nigrum tundra heathlands along a sharp climatic gradient, shows extremely low vegetation recovery in all disturbed plots along the entire climatic gradient. Ten years after simulating the extreme event, the vegetation biomass and the species richness had not reached pre-disturbance levels and no new species had estab- lished in the disturbed plots. Bare soil and erosion were apparent in extreme event plots across the entire climatic gradient. Furthermore, our results showed a limited regenerative capacity of the E. nigrum heathlands in the form of both germinable seed bank and seed rain, and the addition to the plots of seeds of a range of faster growing vascular plant species did not contrib- ute to a faster revegetation of the disturbed plots at any site. Overall, the consistency of the results along the climatic gradient confirms the strong legacy E. nigrum has on the soil envi- ronment and suggests that, the outcome of future extreme events on E. nigrum dominated heathlands could be mediated by the niche constructor ability of E. nigrum independent of the climatic conditions. Our main hypothesis was that the niche constructor ability of E. nigrum would mediate any positive effects of climate on the recovery of E. nigrum heathlands after an extreme event. Table 2. Germinable seeds m-2 from the 144 seedbank samples and average seeds m-2 per year from seed rain mats from each study site. Study site Rebbenes Skogsfjord Snarby Skibotn Guhkesja´vri

Seed bank Monocotyledon 99.6 ± 13.7 0 0 24 ± 12.1 82 ± 65 Dicotyledon 60 ± 29.3 0 0 80 ± 2.5 56 ± 35.6 Seed rain Monocotyledon 35.6 ± 10 9.2 ± 7 10.8 ± 2.5 7.2 ± 3 22.8 ± 12 Dicotyledon 20 ± 6 1.6 ± 0.5 27 ± 6 5 ± 1.4 13.1 ± 5 Seedlings were not identified to species but classified as monocotyledon or dicotyledon. The mean is presented with 95% confidence intervals. https://doi.org/10.1371/journal.pone.0245929.t002 PLOS ONE Niche construction effects on tundra after extreme event PLOS ONE | https://doi.org/10.1371/journal.pone.0245929 February 4, 2021 10 / 14