Thresholds for spring freeze: measuring risk to improve predictions in a warming world.

Plant distribution and productivity are shaped by environmental stressors, particularly freezing events in extra-tropical regions. In early spring, a progressive loss of cold hardiness with phenological development leaves emerging tissues vulnerable to freezing events. In many regions, climate warming is advancing phenology to a greater degree than the date of the last spring freeze, increasing the period of vulnerability to spring freezes. Studies describing critical temperatures at which spring freeze damage occurs are numerous and diverse in both the system studied and type of thresholds used (empirical vs model). Here, we review trends in previously reported thresholds for spring freeze damage across a range of plant groups (row crops, fruit trees, and forest species) over the course of phenological development in spring, and analyze potential sources of variation causing discrepancies between empirically determined and model-derived thresholds. Our analysis shows consistent reporting of higher (less-hardy) thresholds in model-based studies when compared with empirical measurements. These differences highlight a need to improve model estimations of damage by accounting for both microclimatic complexities (humidity, topography, wind speed, etc.) and physiological considerations (phenology and species) that influence the translatability between precise, empirically measured thresholds and critical temperatures used to describe damage in the field.