Quantifying the trade-off between spring phenology and lethal frost risk: a meta-analysis

We preformed a global meta-analysis in synthesising the plant freezing resistance during spring phenology, and provide the corresponding lethal frost risk based on safety margins which defined as the differential between freezing resistance and ambient minimum temperatures during spring phenology. We searched for journal articles using the ISI Web of Science and the China National Knowledge Infrastructure following the key word combinations: 'freezing resistance* frost resistance* cold resistance' OR 'freezing tolerance* frost tolerance* cold tolerance' OR 'freezing hardiness* frost hardiness* cold hardiness' OR 'hardening* de-hardening' OR 'cold acclimation* cold de-acclimation* cold re-acclimation' AND 'leaf* foliage* bud* stem* branch'. Studies included in our dataset had to satisfy the following criteria: (i) only observations at species-level and under nature conditions were considered, those conducted in gardens or glasshouses were excluded due to human management or intervention; (ii) for experimental studies, only 'control data' were included; (iii) freezing test using standardized methods (Supplementary Figure 1) were conducted exclusively during spring phenology, those based on visual freezing damage scores were excluded due to inconsistent quality and reliability; (iv) studies with no clear information on the location of the study sites were excluded. In total, our dataset included 304 observations from 89 studies, encompassing 193 plant species, belonging to 336 orders, 54 families, 122 genera across 126 sites.