Foliar phenology as a sensitive indicator to climate change: a case study to investigate its applicability to the Mediterranean Area

Leaf phenology is shifting in response to global warming YONGSHUO and FU et al. (2014). While changes in phenological spring events due to global warming have been widely documented, changes in autumn phenology, and therefore in growing season length, are less studied and poorly understood VITASSE et al. (2009). European phenological records indicate that leaf unfolding has advanced an average of 2-3 days per decade for the last five decades PEÑUELAS and FILELLA (2001); PEÑUELAS et al. (2002); MENZEL et al. (2006); GORDO and SANZ (2009) and that leaf senescence or fall has been delayed by 1-2.5 days per decade MENZEL and FABIAN (1999); PEÑUELAS et al. (2002); MATSUMOTO et al. (2003); GORDO and SANZ (2009). Changes in the dates of leaf senescence and fall in the phenological records, however, are slower, more heterogeneous and less consistent than those for leaf unfolding PEÑUELAS et al. (2002); MENZEL et al. (2006). This alteration affects surface albedo, ecosystem carbon balance and evapotranspiration, forest productivity, the timing of the transitions between active and dormant stages. In a previous study JOLLY et al. (2005) the use of a generalized bioclimatic index (growing season index, GSI), based on photoperiod, vapour pressure deficit and minimum temperature, was used to asses autumn canopy foliar dynamics in different ecosystems. This study investigates the use of this index at four Mediterranean (central of Italy) meteorological stations with long time series datasets (1955-2013) set on different climatic area. A preliminary analysis shows that photoperiod is the most important factor affecting GSI. The index shows little interannual variations for all the stations, except the one located in the mountains. This station is characterized by a colder climate than the others and for this reason minimum temperature becomes more important than photoperiod affecting GSI interannual variability. Our results suggest that further investigations on the definition of the growing season index are needed in order to apply it to the assessment of autumn canopy foliar dynamics in a Mediterranean.