Cold hardiness dynamics contain adaptive traits that provide phenology information during the dormant season

In temperate and boreal environments, temperatures drop below the threshold for growth. Woody perennial plants then become dormant to ensure survival, while structures develop cold hardiness. Chilling requirements for dormancy must be met, and then cold hardiness must be lost, before growth can resume in spring. Rates of cold hardiness loss (deacclimation rates) have been shown to increase with chilling accumulation. To demonstrate effects of chilling and cold hardiness on plant phenology, here we combine data from three experiments: a natural temperature gradient on a mountain (“Mountain”, Mt. Washington, NH, USA), a natural temperature gradient across the continental USA (“Continental”, many locations between 32.8°N and 47.5°N), and an experimental temperature gradient (the Spruce and Peatland Responses Under Changing Environments, “SPRUCE”, MN, USA). For all three, buds were collected from woody perennial plants from late summer to early spring to measure field cold hardiness. Additionally, cuttings were collected and placed under forcing conditions (22 °C, 16h-day/8h-night) to measure deacclimation rates and time to budbreak. At “mountain”, buds were collected from several altitudes from base (490m) to treeline (1,615). At “SPRUCE”, warming ranges from ambient (+0°C) to constant +9°C above ambient. For “mountain” and “continental”, genotypic effects are expected due to local adaptation.

“Mountain”. Field cold hardiness showed a negative relationship with elevation: higher elevations plants showed had greater cold hardiness than lower elevations. However, the effect of elevation decreased in mid-winter. Until late fall, deacclimation rates were negligible, regardless of altitude, indicating chilling accumulation had not reached a threshold that would allow for growth resumption. In early December, plants from higher altitudes had accumulated enough chilling to start deacclimating at higher rates. From late December to February, all altitudes seemed to have reached a maximized deacclimation rate. Therefore, chilling is not a limiting factor in these environments. Higher elevations showed higher deacclimation rates, demonstrating the adaptive response to the shorter growing season in higher altitudes: once warm temperatures resume, cold hardiness is quickly lost for growth resumption.

“SPRUCE”. Cold hardiness is lesser in warming treatments during fall and spring. In mid-winter, no differences in cold hardiness are observed, regardless of degree of warming. However, the safety margin (distance from air temperature to cold hardiness) is only smaller in warmer treatments during spring, and for some species. At “SPRUCE” dormancy appears to progress faster in warmer treatments based on budbreak, but not based on deacclimation rates.

“Continental”. Colder locations advanced faster through dormancy than warmer locations. Additional artificial chilling was provided to samples from all locations. Additional chilling for warmer locations led to faster budbreak due to increased rates of deacclimation – a dormancy effect. For colder locations, however, additional chilling only resulted in faster budbreak due to cold hardiness loss during artificial chilling, thus not necessarily a dormancy effect.

In all experiments, loss of cold hardiness in field conditions precedes budbreak, and therefore provides progress information towards spring phenological events. The inclusion of cold hardiness dynamics in phenology models can provide better insight into causes of shifts in phenological timing.