Fig. 1: Distributions of the phenological observation sites in the study (a) and effects of late spring frost events (LSF) on flowering based on phenological observation records (b-c).
From: Flowering in the Northern Hemisphere is delayed by frost after leaf-out
Difference in ∆DOY between records in the presence and absence of LSF from all networks, the PEP725 network, USA-NPN, CPON, and RCNN (b) and for seven tree species in the PEP725 network (c). In a, Yellow dots represent the 3,969 sites selected from the PEP725 network. Green, blue, and red dots represent 918, 93, and 44 sites selected from USA-NPN, RCNN, and CNOP, respectively. In b, there are 57,001, 4,991, 734, and 1,239 qualified matched records in the PEP725 network, USA-NPN, CPON, and RCNN, respectively, in the presence of LSF. In the absence of LSF, these numbers are 203,950, 2,970, 10,017, and 5,465, respectively. In b and c, ∆DOY indicates ΔDOYflowering,LSF. Y and N indicate the presence and absence of LSF for each species in each year at each site, respectively. The black dashed line denotes when ∆DOY equals to zero. The boxes span from the first to the third quartile, with median values marked as the black lines in the middle of the boxes. The whiskers extend from the 5th to the 95th percentile. Values beyond the whiskers are hidden. The asterisk indicates a significant difference (two-side t-test, p < 0.05) in ∆DOY between LSF presence and absence using linear mixed models (Supplementary Data 5). Source data are provided in Source Data Fig. 1a–c.
