Inter-annual variability and decadal trends in alpine spring phenology: A multivariate analysis approach

Plant phenological observations are of increasing value as indicators of climate change and variability. We developed a robust multispecies estimate for Swiss Alpine spring phenology for the period 1965-2002 by applying empirical orthogonal function (EOF) analysis on a combination of 15 spring phases. The impact of climate parameters such as temperature and precipitation on the phenological development was investigated using a multivariate statistical model. This multispecies estimate proved to be a good approach to assess the pattern of spring appearance during the last 40 years. It revealed an earlier onset of spring in recent years, mainly since 1988 when a clear shift in spring appearance occurred. The mean overall trend of 1.5 days per decade was clearly driven by winter and spring temperatures whereas precipitation showed no significant influence. The dominant EOF patterns suggested a general climate forcing for the observed inter-annual variability independent of single plant phases. A more regional phenology signal was found in the second EOF mode, indicating slightly weaker phenological trends in southern Switzerland as well as in higher altitudes. Both, temperature and precipitation contributed to this pattern significantly. Analysis of single phases confirmed the pattern of the multispecies estimate. All species showed trends towards earlier appearance ranging from -1 to -2.8 days per decade and the appearance dates had a very high covariance with temperature.