Supplemental Data: Rapid advancement of spring in the High Arctic
snowmelt, the snow cover in each plant and arthropod sampling plot decreased from 100 % to 0% within 2–3 weeks. Plots were visited at least once a week and we estimated the date of snowmelt in each plot as the date of 50 % snow cover by linear interpolation between the latest observation where snow...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Text |
| Language: | English |
| Subjects: | |
| Online Access: | http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.622.1991 http://download.cell.com/current-biology/mmcs/journals/0960-9822/piis0960982207013255.mmc1.pdf |
| Summary: | snowmelt, the snow cover in each plant and arthropod sampling plot decreased from 100 % to 0% within 2–3 weeks. Plots were visited at least once a week and we estimated the date of snowmelt in each plot as the date of 50 % snow cover by linear interpolation between the latest observation where snow cover was above 50 % and the earliest observation where snow cover was below 50%. In some cases snowmelt occurred prior to arrival at Zackenberg. Since the spatial pattern of snowmelt is relatively consistent across years [S3] we predicted missing values from plots with similar pattern of snowmelt. Plants. The ratio of buds to open flowers was calculated from weekly counts of 200 randomly chosen flowers per plot in a total of 25 plots of six different plant species during June, July and August in each year during 1996–2005. The date when 50 % of the flower buds had opened on each plot was chosen to represent onset of flowering and was estimated by using linear interpolation between weekly estimates of the ratio of buds relative to open flowers. Estimates of onset of flowering based on less than 10 flowers were omitted from the analyses. Arthropods. We used both window (one plot) and pitfall traps (five plots) which were emptied weekly during June, July and August. All plots were operated during 1996–2005 except for one |
|---|