Figure 1.
Each station has been indicated by a red circle. Symbol sizes are proportional to the temporal length of the local longest pollen record.
Figure 2.
Trends of annual pollen index (API) by species.
Boxplots show the proportional annual change of yearly pollen sums for the 23 pollen taxa analyzed (reasons for selection given in the main text). Medians are significantly different from zero (Mann-Whitney test, * : , ** :
, *** :
, n.s.:
) for 11 taxa. On the right, the percentages of significant trends are indicated for each taxon (of which the percentages of positive trends are given in parentheses). The height of the boxplot is related to sample size, taxa are arranged in decreasing order of their medians.
Figure 3.
Boxplots show the proportional annual change of yearly pollen sums for 13 countries. Medians are significantly different from zero (Mann-Whitney test, * : , ** :
, *** :
, n.s.:
) for six countries. On the right, the percentages of significant trends are indicated for each country (of which the percentages of positive trends are given in parentheses). The height of the boxplot is related to sample size, countries are arranged in decreasing order of their medians.
Figure 4.
API trends against temperature trends by species.
Proportional annual change of yearly pollen sums was plotted against local temperature trends for 23 pollen taxa. Temperature trends were calculated for each location for the mean temperature of two seasons, January to April (associated with the flowering of Alnus, Betula, Carpinus, Corylus, Cupressaceae, Fagus, Fraxinus, Olea, Pinaceae, Platanus, Populus, Quercus, Salix, and Ulmus) or April to August (related to Ambrosia, Artemisia, Castanea, Chenopodiaceae, Plantago, Poaceae, Rumex, Tilia, and Urtica), over the years 1977–2009. A regression line has been superimposed for Betula and Carpinus, the only statistically significant relationships.
Figure 5.
Mean API against mean local temperature.
Log-scaled mean annual sum of airborne pollen was plotted against local mean temperature for 23 pollen taxa. Mean temperatures were calculated for two periods, January to April (associated with the flowering of Alnus, Betula, Carpinus, Corylus, Cupressaceae, Fagus, Fraxinus, Olea, Pinaceae, Platanus, Populus, Quercus, Salix, and Ulmus) or April to August (related to Ambrosia, Artemisia, Castanea, Chenopodiaceae, Plantago, Poaceae, Rumex, Tilia, and Urtica), over the period 1977–2009. Only significant regression lines are shown.
Figure 6.
API trends by environment type.
Boxplots show the proportional annual change of yearly pollen sums for different environments. Mann-Whitney tests show a significant increase (median different from zero, ) of airborne pollen in urban environments. The notches are calculated as
and the height of each boxplot is related to sample size. On the right, the percentages of significant trends are indicated for each type of environment (of which the percentages of positive trends are given in parentheses).
Figure 7.
Maximum duration of pollen series by location.
The local longest monitored period is shown as a red bar for each of the 97 locations considered. Missing years, occurring in few cases, have been omitted for clarity.