The influence of altitude and urbanisation on trends and mean dates in phenology (1980-2009)

Long-term studies on urban phenology using network data are commonly limited by the small number of observation sites within city centres. Moreover, cities are often located on major rivers and consequently at lower altitudes than their rural surroundings. For these reasons, it is important (1) to go beyond a plain urban-rural comparison by taking the degree of urbanisation into account, and (2) to evaluate urbanisation and altitudinal effects simultaneously. Temporal phenological trends (1980-2009) for nine phenological spring events centred on the German cities of Frankfurt, Cologne and Munich were analysed. Trends of phenological onset dates were negative (i. e. earlier onset in phenology) for 96% of the 808 time series and significantly negative for 56% of the total number. Mean trends for the nine phenological events ranged between -0.23 days year ^-1 for beech and -0.50 days year ^-1 for hazel. The dependence of these trends and of mean dates on altitude and on the degree of urbanisation was explored. For mean dates, we demonstrated an earlier phenological onset at lower altitude and with a higher degree of urbanisation: altitude effects were highly significant and ranged between 1.34 days (100 m) ^-1 (beech) and 4.27 days (100 m) ^-1 (hazel).Coefficients for the log-transformed urban index were statistically significant for five events and varied greatly between events (coefficients from -1.74 for spruce to -5.08 for hazel). For trends in phenology, altitude was only significant for Norway maple, and no urban effects were significant. Hence, trends in phenology did not change significantly with higher altitudes or urbanised areas.