The damage caused by wind in middle-aged Scots pine stands on permanent thinning experimental plots

Wind belongs to the most important disturbing factors having an influence on the forest growth (Mitchell 2013). Its importance may increase in the future according to predicted trends in climate (Peltola et al. 2010). Full protection against this factor is impossible, especially in case of extreme winds (Wood 1995). So, there are three different strategies of forest management in the context of wind threat (Gardiner and Quine 2000): 1)

acceptation of loss caused by wind;

The first and second strategies are the most popular in countries frequently touched by storms (Savill 1983) and the last one is preferred in regions of moderate risk of wind catastrophe (Zajączkowski 1991). Between silvicultural operations, thinning seems to be of key importance to shape trees and tree communities resistant to atmospheric factors. Although each thinning brings a short period of instability (Quine 1995; Jalkanen and Mattila 2000; Valinger and Fridman 2011), but in the long run, it may increase tree resistance if properly performed (Huss 1983, 1993; Zajączkowski 1991; Suvanto et al. 2016). The problem of thinning is especially important in the case of coniferous species, which are generally more susceptible to wind damage than broadleaves (Konôpka et al. 1987; Zajączkowski 1991; Peltola et al. 2000; Valinger and Fridman 2011). Contradictory recommendations on thinning regime are made due to following different silvicultural strategies connected to stand stability, focused on individual or group resistance (Slodičak 1995). The first strategy leads to strong growth and tapered trees by wide initial spacing (Donis et al. 2020) and rarely repeated heavy thinnings (Abetz 1976; Huss 1983, 1993). The second one is oriented on preventing the penetration of wind into interior of the forest. Therefore, moderate and frequent thinnings are preferred, without breakage of the forest canopy, respecting the natural groups of strong trees (Fleder 1990; Zajączkowski 1990). It is also possible to mix both these strategies by performing heavy thinning interventions in the young stand and light ones in the maturing stand (Slodičak 1995). The forming of stand edge in both cases has a great importance (Dupont and Brunet 2008). However, independent of silvicultural measures, their influence on forest stability decreases with increasing wind speed. Trees are prepared to resist only the particular strength and direction of wind like they ‘remember’ from their history (Quine 1995; Wood 1995; Talkkari et al. 2000), forming their crowns and stems according to wind exposure (Tomczak et al. 2014).

The aim of this work was to check the influence of thinning methods applied in a permanent experimental plot on the level of damage caused by strong wind accidents that appeared in the years 2002 and 2011.

The research plot was located in Ostrów Mazowiecka Forest District (21°50′E, 52°45′N), where in 1968, a 19-year-old Scots pine (Pinus sylvestris L.) stand was established with direct sowing. A similar experiment was conducted in 1970 in Myszyniec Forest District (21°38′E, 53°20′N) in a 21-year-old Scots pine stand established by planting. Forest habitat in both cases is defined as fresh coniferous forest on haplic arenosol soil, site class I (Myszyniec) and I.5 (Ostrów Mazowiecka). On both plots, thinning treatments were applied with 5–6 year intervals. The treatments are as follows:

K – control plot (without thinning);

In the case of selective thinning treatments, future crop trees had been chosen in the number of 500 pcs per hectare and cuts were concentrated on their strongest competitors. Intensity of thinning, measured with basal area units, varied between 20% and 30%. In the case of schematic cuts, each fifth row was removed at the beginning of the experiment. In the case of thinning from below, only the trees belonging to third and lower Kraft’s class were removed.

Experimental design was the random block system with five replications. The surface of each research unit was equal to 0.1 ha. On the plot Myszyniec, the random block design was not completed since 1990 after four units were excluded from the experiment because of big gaps caused by the fungal pathogens Heterobasidion annosum (Fr.) Bref. and Armillaria mellea (Vahl) P. Kumm. However, all the plots were measured. Each 5 years, the diameters of all trees had been measured and the height of 25 trees per unit was used to calculate the mean height. The last measurement was done in July 2001 and thinning was prescribed. Trees chosen for removal in the plot had been cut already in the winter of 2001/2002. Cuttings on the rest of the plot were left to the next season.

On 4 July 2002, strong windbreak occurred in some forest district in the northeast of Poland. The reason was the cold front that occurred between two air masses causing storms accompanied by a sudden violent squall wind from the SW direction (Mikułowski 2002; Gil and Mikułowski 2003; Lorenc 2012). The wind speed exceeded 40 m s⁻¹. The forest district Myszyniec belonged to this area and the storm touched the thinning experimental plot too. In fact, this plot was located on the border between the zone of totally destroyed stand and the zone where it was only partially damaged. It caused visible differences in the amount of damage between blocks of experiment. That is why, this factor was also taken into account during the analyses of results. Trees that survived the storm had been noticed in autumn 2002, including the units that were left out of the experiment. Basal area, mean dbh and height were calculated for the stand remaining and removed, based on the data collected the year before. It was the last measurement on this plot which disappeared after the storm.

On the plot Ostrów Mazowiecka, full measurement was done in 2008. Thinning was not prescribed at that time, and so, only sanitary cuts were performed. On 20 July 2011, a local strong convection storm occurred in the part of forest district. The strength of this wind is unknown because of the distance to the closest meteorological station. The experimental plot belonged to the area subjected to damage. The inventory of this plot was performed in spring 2014. Based on previous measurements and inventory of damage, calculations of basal area, mean height and dbh of broken and uprooted trees were done. This plot exists to this time.

To analyse the differences between the amounts of damage, in particular thinning treatments, it was necessary to use non-parametric test (Kruskal–Wallis) because the distribution of this feature differed from the normal one. Additionally, to compare the differences between basal area of damaged trees on two parts of Myszyniec plot (thinned last year before the storm and 5 years earlier), the Mann–Whitney U test was applied. Analyses have been done using STATISTICA work package (Statistica 2007).

Wind damage appeared in both cases of experimental plots where history of silvicultural interventions was well documented. It gave the opportunity to check different theories connected to damage prevention by different thinning methods. Results obtained by this are, however, not sufficient to formulate trusted silvicultural conclusions. Generally, they confirmed the opinion that in case of extreme wind, the possibility to prevent damage is limited (Quine 1995; Talkkari et al. 2000). Such a situation appeared on the plot Myszyniec, where some parts were subjected to total destruction, independent of the thinning treatment. Similar summer storm, with ‘bow echo’ effect, touched 15 years later, in August 2017 in the forests of northwestern Poland (Dmyterko and Bruchwald 2020) and caused large-scale destruction of forest stands. Another summer storm that struck in June 2016 destroyed parts of forests in eastern Poland (Bruchwald and Dmyterko 2019).

The difference in damage between thinned and non-thinned parts of the plot might be caused by the temporal instability caused in the short period after intervention (Quine 1995; Jalkanen and Mattila 2000; Valinger and Fridman 2011). However, the material is too poor for drawing hard conclusions on this basis. An equal probable reason might be a local variability of wind gusts blowing from untypical direction, especially in the area close to stand edge (Quine 1995; Wood 1995). Results obtained on this plot did not confirm the rule that trees suffering from the wind are thinner than average (Abetz 1976; Huss 1983, 1993; Skatter and Kucera 2000). It was probably caused by the extremal strength of the wind, when such regularity may disappear (Valinger et al. 1993).

Quite a different situation appeared in Ostrów Mazowiecka, where the obtained results showed relatively low level of damage caused by wind of unknown speed (see Materials and Methods). According to the wind classification by Lorenc (2012) and based on the damage level, it may achieve a speed between 25 and 28 m s⁻¹. The differences between treatments were not significant; however, it could be observed that the lowest amount of damage was in treatment SD, where the stand was thinned selectively in the first stage and from below in the second stage. It is compatible to the opinion of the authors who suggested heavy interventions in young stands and weak or moderate interventions in medium-aged stands, without breakage of forest canopy (Zajączkowski 1990; Slodičak 1995). In comparison to Myszyniec plot, broken trees had usually smaller diameter than average. It can be stated that wind eliminated some kind of ‘risk trees’ (Abetz 1976), allowing the rest to resist. Therefore, we cannot deny the hypothesis about positive influence of thinning in Scots pine stands from the point of view of mechanical stability. It should be only emphasised that it is true until the wind does not exceed a definite threshold. However, even in the case of extremal wind, silvicultural measures can minimise losses in places lying on the border of calamity zones (Kohnle and Gauckler 2003).