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

extreme event plots showed an opposite pattern, as seen from the significant interaction between extreme event treatment and continentality index (Fig 5a). Thus, continental sites had originally higher biomass of subordinate species, but also had lower vegetation recovery of these species after vegetation removal. In contrast to subordinate species, we found a drastic reduction of mean E. nigrum biomass in extreme event plots when compared to control plots (Figs 3 and 5a). Also, we found a signif- icant interaction of extreme event treatment and continentality index, with a higher biomass of E. nigrum in coastal than continental extreme event plots, but not in control plots. Thus, E. nigrum recovery after vegetation removal was promoted in coastal as compared to continental sites. Bare soil and lichen cover were significantly higher in extreme event plots as compared to control plots, while moss cover varied overall across the climatic gradient, but not between extreme event and control plots, finding overall less moss cover in continental sites (Fig 5c, S2 Table). We found no difference in soil pH values between control plots and extreme event plots (F.value = 2.1, p = 0.92). The vegetation trait composition nine years after the extreme event was simulated was char- acterized by higher values of both SLA and LNC, but lower values of LDMC in extreme event plots when compared to control plots (Fig 5d, S2 Table). These results were consistent along the climatic gradient, i.e. there was no significant interaction of extreme event treatment and

continentality index (Fig 5d). In extreme event plots, mean SLA, LNC, and LDMC were 9.71 ±0.7mm, 14.18±0.5mg/g and 429.95 ± 3.21 mg/g respectively, while control plots had mean values of 6.88±0.5, 11.26±0.31, and 439.89 ±1.3 mg/g. To further study the vegetation community changes after the extreme event, we grouped the species as dwarf shrubs or grass/forbs/sedges (S1 Table). We found that dwarf shrubs accounted for over 80% of the subordinate species biomass in control plots (Fig 3a) and that it was dwarf shrub biomass that made up most of the sparse vegetation regrowth in extreme Fig 5. Standardized coefficients from the linear mixed effects models on the effects of extreme event treatment (EE), continentality index (CI) and their interaction (CI x EE) on data collected in 2019. a) mean species richness and biomass (total biomass, subordinate species biomass and E. nigrum biomass alone), b) mean seedling length of species used in bioassay experiment excluding peat treatment, c) mean cover of mosses and lichens, and bare soil, and d) community weighed means of selected traits, specific leaf area (SLA), leaf dry matter content (LDMC), leaf nitrogen content (LNC). Note the different scale of standardized coefficients in the different sections. https://doi.org/10.1371/journal.pone.0245929.g005 PLOS ONE Niche construction effects on tundra after extreme event PLOS ONE | https://doi.org/10.1371/journal.pone.0245929 February 4, 2021 9 / 14