This post is part of the series Simulating Climates in Growth Chambers.
It remains difficult to simulate realistic winters. The more expensive growth chambers will allow programmed temperatures of -2 °C at night (with lights off). Fluorescent and incandescent lights themselves generate heat and the cooling requirement is energy-demanding. LED lights are costly. While a temperature of -2°C is technically possible, I have found out the hard way that this may not actually be what you want. When submitting plants to cold temperatures during the day (+2°C) and freezing temperatures at night (-2°C) it has happened that my soil volume froze, and did not thaw during the day. This resulted in wilting. In nature, plants would form roots deeper down, which have access to soil water. In the chamber, soil volume is very shallow, and assumes the temperature of the surrounding air, while buffering only the highest and lowest extremes.
Our interest has mostly been in growth, so a six-week “chilling period” is applied after one growth season to encourage more uniform flushing in the second growing season. During the chilling period, temperatures are held at a constant 4°C and day length is reduced to 8 hours or less, under reduced light intensity. This has been sufficient in our case, since the plants had received the correct signals to set bud and prepare for winter by the time our simulated end of the growing season (“October 15”) arrived.
In certain situations it is necessary to have real frost. We found this was true for testing cold hardiness of pine and spruce. We use the electrolytic leakage method described in Hannerz et al. (1999)1. Day length is the most important trigger in signalling the end of the growing season and initiating dormancy. However, in nature these shorter days are correlated with cooling temperatures. When the factors are uncoupled in the growth chamber, we find that the cooler temperatures are a required component of the acclimation procedure. We have successfully induced hardiness by subjecting plants which had already set bud to brief periods of night frost. To this aim, we gradually, over a period of 4 days, dropped night temperatures to a minimum of -2°C, until we kept it there for two hours.
A week after this acclimation took place, pretests revealed that the plants had responded sufficiently to reveal the population signal we were interested in, and yield suitable testing temperatures. Full tests were scheduled for the second week after the frost. Higher daytime temperatures (maxima between 23 and 30 °C) did not undo or mask the nighttime frost signal, and the soil never risked freezing with these brief exposures.
1 Hannerz, M., S.N. Aitken, J.N. King, and S. Budge. 1999. Effects of genetic selection for growth on frost hardiness in western hemlock. Can. J. For. Res. 29: 509–516