1. Introduction
In Québec province, Canada, windthrow events in balsam fir (
Abies balsamea (L.) Mill) stands can be considered as a significant source of disturbance and economic losses [
1], and are known to be linked to forest management [
2]. To improve forest management in these stands and mitigate the wind damage risk, the stability and vulnerability of balsam fir trees has been studied as a consequence of different types of thinning [
3,
4] and an existing British wind risk model: ForestGALES [
5,
6] has been tested, adapted, and improved to provide a decision support tool for forest managers [
3,
4,
7,
8].
During winter, freezing temperatures and snow cover are known to have an additional impact on wind damage. The ForestGALES model allows for the addition of snow in the calculation of the critical wind speed at which a tree will uproot or break. Snow is represented as an additional weight in the crown, changing the moment due to the overhanging crown mass when the tree is bent by the wind, and therefore increasing the turning moment applied to the trunk. However, what happens in a stand during winter is much more complex than this simple representation. In balsam fir forests, winter and snow cover can last for 6 to 7 months and snow can reach more than 200 cm depth. This thick layer of snow and ice produces a change in the wind profile as the snow offers less aerodynamic resistance to wind than ground vegetation and, therefore, the rate of change in wind speed is probably larger at ground level during the winter season [
9]. In the canopy, snow accumulates on branches and can form large clumps of snow and ice. The weight of the snow pressing on the branches causes a change in the shape of the crowns and a decrease in the canopy porosity likely to influence the stand wind profile as well. This snow loading of tree crowns combined with freezing temperatures can cause severe damages to trees, such as stem bending or stem breakage, but rarely uprooting, as frozen temperatures increases stem stiffness [
10,
11], and as the weight of the snow cover on the soil, and/or frozen soil, reinforce the root system [
12].
Most of the studies conducted on wind, snow, and conifers are from Northern Europe, and are based on long-term damage surveys. What is known from these studies is that the factors that most influence winter breakage or uprooting are related to tree shape and stand characteristics [
13]. More specifically, stem taper [
10,
14,
15], crown characteristics [
12], and the number of neighboring trees [
16] are the most important factors controlling the stability of trees in winter. In relation to this, the choice of thinning carried out in the stands seems to be crucial in their management with an increase in wind and snow damage risk following thinning in a stand [
17,
18].
There are to date no studies on the impact of snow intercepted by the canopy on the turning moment experienced by trees during wind loading, or any study of winter impact on wind damage in balsam fir stands, creating a lack of knowledge about winter damages in boreal stands in Northeastern America. This gap in knowledge means that we currently are not able to model what is happening in winter, or to estimate the potential impact of snow on the level of wind damage risk. Therefore, improving knowledge in this area is essential to improve the decision making process for forest management [
13]. Our measurements are the first to directly measure wind and snow loading on trees during winter.
The overall hypotheses for the study were:
- (1)
The additional weight of the snow on the crown will increase the lever arm on the trunk, and trees will experience an increased turning moment at a particular wind speed with an increase of snow thickness in their crowns compared to when there is no snow in the crown.
- (2)
The large negative temperatures will stiffen the trunk because of freezing, therefore trees will experience a globally lower turning moment at a particular wind speed in winter compared to during the summer because the crowns move less.
The first hypothesis deals specifically with the effect of snow and the second hypothesis with the effect of freezing temperatures.
4. Discussion
Our results show that the turning moment experienced by the trees was highly influenced by season, with a turning moment globally higher in winter, but apparently not strongly influenced by the snow thickness on the crowns. These results go against our hypotheses which were that for a similar wind speed, trees experience a higher turning moment when snow is present on their crown, and globally lower turning moment in winter. The
TreeVariables retained by our model selection also diverge between the two models. In the first model, including only the winter dataset, at least 4
TreeVariables could have been selected instead of the competition index
CBAL. However, all these variables are interrelated, based on
DBH, and relate to the social status of the tree in the forest. In the second model, which combines summer and winter datasets, the distance-dependent competition index
C12 appears to be the key variable that links the seasons. There is thus a direct dependency of the maximum hourly turning moment on the local competition, which confirms Hale et al. [
23] results, but also shows an influence of close competitors during the winter season.
Regarding the effect of snow thickness on the crown, care is needed with the interpretation of the results. The depth of snow on the crown does not seem to make a difference on the turning moment but the model selection for the winter dataset retained snow as a key variable. However, it is important to note that our way of monitoring the presence of snow on the crowns was not optimal. It was a first attempt to evaluate the effect of snow in the canopy on the turning moment in a practical way. Monitoring the presence of snow on the canopy is complicated by the structure of the canopy itself and the great variability in the interception of snow by the branches, as wind and temperature influence the moisture content of snow and therefore the degree of stickiness to the canopy [
12]. Therefore, the accumulation of snow on trees sufficient to cause snow damage depends upon the quantity and type of snow [
12] because the same snow thickness could result in two different weights on the branches. As the process of snow interception by trees is complex and involves components of throughfall, adhesion, cohesion, wind removal, sliding, melting, and vapor transport [
39], further work is required to evaluate the additional load on trees due to snow.
Continuous monitoring of wind at two different height showed a significant change in the wind profile between summer and winter, with an increase in wind speed at both anemometer heights and reduced calm periods during winter. This is probably due to the presence of snow in winter, which changes the geometrical structure of the stand and therefore the wind profile. This finding supports the fact that deep snow cover also offers less aerodynamic resistance to the wind than ground covered by vegetation and regeneration [
9].
The accumulation of a large amount of snow on the ground during the winter season, with depths up to 200 cm, also raises the effective level of the ground, while the trees themselves do not change in height and therefore have the same length of lever arm. As observed during fieldwork, the tree trunks remain free to move because their own radiation prevents snow from sticking to the trunk and creates a snow-free sleeve around the trunk down to the ground. The increase of the turning moment experienced by the trees in winter compared to summer for a similar wind speed could be due to several reasons. It might be due to increased stem stiffness due to freezing in the large negative temperatures or to increased root anchorage as the heavy snow depth during our data collection (130–150 cm depth) put a heavy load on the root system. It is quite possible that the root system in winter is completely locked in place by the snow cover and the frozen soil. The force of the wind on the tree would therefore be mostly applied to the trunk and with very little energy dissipated in the root system. In contrast, in summer, the soil is very wet, and the root system is more flexible, so some of the energy initiated in the trees by the wind is dissipated in the soil. In addition, the higher stiffness of stem and branches during winter decreases the streamlining under wind loads and, combined with a reduced crown porosity, the drag coefficient will be higher in winter. These are probably the two main reasons for the increase of the turning moment during winter.
An increase in the turning moment experienced by the trees, means an increased risk of stem breakage during winter season. With climate change, the relative importance of snow seasons against snow free seasons is likely to change, with an increase of unfrozen soil days during the windiest periods [
40]. If the soil remains unfrozen and without deep snow cover during these windiest periods, this could lead to an increase in overturning during early and late winter, when strong winds are more frequent. It will be important in the near future to have reliable predictions of the future winter climate in Quebec. It will also be useful to study the root anchorage of trees during the pre-winter and winter periods by carrying out destructive tree winching at different temperatures and with different snow cover on the ground, to evaluate if the changes we observed in this study are actually related to changes in root anchorage.