Using easily reproducible twig experiments to quantify curvilinear temperature responses of flowering and leafing times

Abstract Premise Phenology—the timing of developmental events—is one of the most effective ways to study the impacts of climate change on plants and ecosystems. However, it remains unknown whether phenology shows a linear or curvilinear response to spring warming. Methods We measured days to leaf out and flowering for dormant twigs of 12 woody species in seven temperature conditions, representing a wider range of temperatures than in previous studies, and tested whether responses of leaf out and flowering follow linear or curvilinear trends by comparing model fit of linear and second‐order polynomial regressions. We also evaluated how the flower–leaf emergence sequence (FLS), the number of days between flowering and leaf out, is affected by temperature. Results Overall, warmer temperatures caused earlier flowering for all species and earlier leaf out for many, though there was some lack of response at the extremes of the temperature range. Species varied in whether responses were linear or curvilinear, though flowering responses of most species were best explained by curvilinear models. Due to intraspecific variability in flowering and leaf‐out sensitivities, two species exhibited different FLS patterns under different temperature treatments. Conclusions These findings suggest that studies investigating climate change and phenology should consider using curvilinear rather than linear models, particularly for flowering. Such curvilinear models may suggest alternative scenarios regarding how warming will impact phenology and ecosystem functions. Variable flowering and leaf‐out sensitivities can also affect FLS. Researchers should consider using a wider range of temperatures in dormant twig experiments.