Methods
For each of our research objectives we have conducted separate experiments. On this page we describe the experimental methods for our first objective:
1. What is the range wide variation in heat sums for white spruce? What variables control this variation?
1. What is the range wide variation in heat sums for white spruce? What variables control this variation?
Experimental Design
For the heating requirement experiment 33 provenances were sampled to capture the range of climatic variation across Canada that could influence this trait (Figure 2). Two branches were collected from one tree per block. Collection occurred in the spring after winter chilling had completed: between April 29 and May 15, 2021.
Following collection branches were transferred indoors into glass jars filled with water that was refreshed once a week. Indoor heating was monitored using an Elitech RC-4HC Temperature and Humidity Data Logger. On observation days buds on individual branches were rated using a standardized scale for white spruce (Dhont et al., 2010). Observation of the buds took place immediately after they were transferred indoors and continued every day until the branch had died. |
Statistical Analysis
Heating requirements were determined by calculating growing degree days using a model where heat accumulates uniformly above 0 C (Man & Lu, 2010). In the heating requirements experiment, the amount of outdoor heating experienced by the branches varied according to when they were collected. Outdoor heating was calculated using data from a nearby weather station, Rock Island Lake Auto (55° 19' 35.04", -113° 27' 37.44"). Weather station data was obtained from the Alberta Climate Informative Service’s (ACIS) Current and Historical Weather Station Data Viewer https://acis.alberta.ca (July 2021). Outdoor heating accumulation was calculated beginning January 1 and ending on the date of branch collection. Indoor heat sum accumulation was calculated with the same model using data from the temperature data logger.
A sigmoidal function was fitted on the cleaned observational data for each provenance to produce continuous budbreak development curves. These curves were used to determine the heatsum requirements for the budbreak stages for provenance.
Using these values, variance partitioning was performed to determine the extent that different environmental variables contributed to differences in heatsum requirements for budbreak. A variety of environmental variables from the origin of the provenances were first assessed using the correlation coefficients and only three were eventually used in variance partitioning. They were: growing degree days (GDD), region, and period. Data for growing degree days were obtained from ClimateNA (Wang et al., 2016). Region refers to the region of Canada the provenance originated from. Period was calculated as the average number of growing degree days between the first day in spring above 0 C and the last day in spring below 0 C.
A sigmoidal function was fitted on the cleaned observational data for each provenance to produce continuous budbreak development curves. These curves were used to determine the heatsum requirements for the budbreak stages for provenance.
Using these values, variance partitioning was performed to determine the extent that different environmental variables contributed to differences in heatsum requirements for budbreak. A variety of environmental variables from the origin of the provenances were first assessed using the correlation coefficients and only three were eventually used in variance partitioning. They were: growing degree days (GDD), region, and period. Data for growing degree days were obtained from ClimateNA (Wang et al., 2016). Region refers to the region of Canada the provenance originated from. Period was calculated as the average number of growing degree days between the first day in spring above 0 C and the last day in spring below 0 C.