En route to improved phenological models: can space-for-time substitution give guidance?

Changes in plant phenology—the seasonal timing of developmental events—are probably the best documented observed responses to ongoing climate change. Studies based on large global data sets have shown unequivocally that spring has advanced in the northern hemisphere (Parmesan and Yohe 2003, Root et al. 2003, Menzel et al. 2006), with an overall estimate across species of almost 3 days per decade (Parmesan 2007). Other studies have related the advancement linearly to the change in mean annual temperature (Wolkovitch et al. 2012). However, although these results provide convincing evidence that climate change is indeed affecting phenology of natural ecosystems, they shed little light on how to make forward predictions of phenological changes. The observed rates largely depend on the data set at hand (Parmesan 2007, Wolkovitch et al. 2012), and in any case, future predictions of warmer temperatures would involve extrapolation (Morin et al. 2009). On the other hand, the underlying mechanisms and causal relationships between environmental cues and pheno-logical events still remain poorly understood (Morin et al. 2009, Pau et al. 2011). In this issue, Jochner et al. (2013) set out to investigate the possibility of utilizing urbanization gradients of temperature as an analogue of climate change, to provide insights into climate change impacts on phenology. The case of birch (Betula pen-dula Roth) flowering is used as an example. Urbanization gradients are characterized by a decrease in temperature from the city centre towards the rural surroundings, but importantly, combined with simultaneous trends in CO 2 concentrations and other factors that are expected to accompany climate change (e.g., Zhang et al. 2004, Carreiro and Tripler 2005). While the inner-city parts of the urban gradient could be used to mimic the future environment in the area, the complete data sets with spatial trends could substitute time series with corresponding temporal trends. It has therefore been suggested that the urban gradient could be used to inform models about the quantitative relationships between climate and phenological events in current and future environments (Chung et al. 2009). To what extent the urban gradients actually represent predictions of future environmental factors still remains unclear, but more accurate information is becoming available that would allow for such comparisons (Rummukainen 2010). Urban gradients are an example of the method of space-for-time (SFT) substitution that has been widely used in ecology and forest science (Fukami and Wardle 2005). Important applications to forestry, growth and yield tables have been constructed …