Impact of climate warming on phenological asynchrony at the European scale
Our model simulations predict that uniform warming by +4°C will advance OAB and TDM by similar median values of 19 and 24 days, respectively. Yet, the variability in the response of OAB and TDM to warming differs. While the advancement of OAB with warming varies substantially between geographic locations and lake types (20th-80th percentiles 0-35 days; Fig. 3a-d; Fig. S5), the advancement of TDM is spatially more uniform and considerably less variable between lake types (20th-80th percentiles 19-30 days; Fig. 3e-h; Fig. S6). Consequently, the net effect of warming on phenological asynchrony (PLD) is on average close to zero (median -3 days), but can range from -60 to +60 days (Fig. 3i-l; Fig. 4b; Fig. S7).
How phenological asynchrony (expressed as PLD) changes with warming depends on the dominant process controlling OAB (Fig. 4a, b) and thus on lake type and geographic location, because these properties determine which process controls OAB before and after warming (Fig. 3i-p; Fig. S7-S9). In northern, eastern, or high-elevation lakes with optical depth ≤ 6, where OAB is controlled by ice-off (Fig. 3m; Fig. S8), asynchrony will increase because OAB advances more with warming than TDM (Fig. 4a-f; Fig. S9). In contrast, in low-elevation, southern and western lakes with optical depth 6-18, where OAB is controlled by incident radiation (Fig. 3n, o; Fig. S8), asynchrony will decrease because only TDM advances with warming (Fig. 4a-f; Fig. S9). In lakes with optical depth ≥ 36, where OAB is controlled by the onset of stratification (Fig. 3p, Fig. S8), asynchrony will not change systematically because OAB and TDM advance similarly with warming (Fig. 4a-f; Fig. S9). Finally, lakes in which the dominant process controlling OAB will shift with warming – either from ice-cover to radiation (optical depth 6-12) or from radiation to stratification (optical depth 24-30) – OAB will advance at a slightly faster or slower pace than TDM, respectively, leading to intermediate shifts in asynchrony (Fig. 4a-b, Fig. S9). Intriguingly, warming will thus decrease asynchrony in lakes in which it is currently largest (radiation controlled lakes), but will not change asynchrony in lakes in which it is currently shortest (stratification controlled lakes; Fig. 4b; Fig. S9).
Overall, the OAB controlling factors explain more than 60% of the variance in the warming-induced changes in both OAB and PLD in the 16 lake types across Europe (Fig. 4g, i). The controlling factors thus describe the impact of warming on phenological asynchrony equally well as does the combination of geographic coordinates and optical depth (Fig. 4g, i).