Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Postfire tree mortality and fire resistance patterns in pine forests of Ukraine

Postfire tree mortality and fire resistance patterns in pine forests of Ukraine The study was conducted in pure Scots pine (Pinus sylvestris L.) stands within the forest steppe physiographic region of Ukraine damaged by surface fires with different intensity. The aim of the research is to determine the effect of different fire intensity on pine stand and individual trees, considering tree morphometric parameters and type of damage. The intensity and duration of fire-related tree mortality was different in stands with different age. We found that tree fire resistance is driven by tree diameter, height of the rough bark, and natural degree of thickness. The pro- portion of dead trees one year after the spring fires in the middle-aged pine stands was 5 times lower and in mature pine stands even 10 times lower than after the summer fires. The critical damage to tree crowns in young pine trees causing their death is 80% of the needles burned. In the middle-aged pine trees, critical damage depended on the size of trees. The death of large, mature trees after smoldering summer fires was associated with the accumulation of a large stock of forest litter and duff near the tree-base, which contributed to the increased intensity of fire and its localization near the base part of the trees. Based on our findings, postfire tree mortality models have been developed for different age groups of pine stands. Key words: surface fires; smoldering fires; postfire mortality models; bark char; crown scorch; season of fire Editor: Tomáš Hlásny largest rivers, also in the Crimea Peninsula. More than 1. Introduction 90% of forest r fi es in Ukraine occur in pine forests (Voron Wildfires are one of the most dangerous factors for for- & Melnyk 2009; Voron & Sydorenko 2014). During the ests, leading to their destruction and degradation. Due last ten years (2009–2018), 19.9 thousand forest fires to global warming and increasing aridity, the risk of have occurred in Ukraine on an area of 37.2 hectares. The increased frequency and extent of wildfires is very high. economic losses amounted to 8.86 million Euros (Public Many regions of the world have experienced an increas- Report 2017). ing trend of excessive wildfires and an increasing occur- The rapid deterioration of the sanitary state of dam- rence of extremely severe fires (FAO 2006). aged trees leads to significant economic losses due to In 2017, wildfires burnt over 1.2 million hectares the decline of stand merchantability and the deteriora- of natural lands in the EU. The European Forest Fire tion of its technical quality. Timely diagnosis and accu- Information System estimated the amount of r fi e-related rate prediction of the postfire trees mortality are thus losses to be around 10 billion Euros (San-Miguel-Ayanz extremely important. Such studies can mitigate the et al. 2017). negative effects of wildfires and will be used to support In Europe, Scots pine forests now exceed 28 million decisions about forestry treatments in such forests. A hectares, covering over 20% of the productive forest area considerable amount of research papers is devoted to the (Mason & Alìa 2000). The most fire hazardous conifer- prediction of postfire loss of trees for different conifer ous forests occupy 43% of the total area of Ukraine, in species (Regelbrugge & Conard 1993; Usenya & Churylo particular, pine stands – 35%, which grow in the North 2001; Fites-Kaufman et al. 2008; Sah et al. 2019). It is of Ukraine (Polissya) and in the South (Steppe) along the well known that the value of postr fi e mortality rate affects *Corresponding author. Serhii Sydorenko, e-mail: serhii88sido@gmail.com, phone: +38 099 223 29 08 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 the predominant type of damage and r fi e intensity as well bark, m; number of r fi st-order roots above ground. Natu - as secondary disturbances such as drought, insects, and ral degrees of thickness (NDT) were additionally calcu- diseases (Ahafonov & Alekseev 1989). The research on lated for each plantation on the test plots. The date of r fi e, the impact of fire damage on postfire tree mortality in the date of measurement, the soil condition and the stand Ukraine is rather fragmented. The existing diagnostics age were recorded. We also examined a one-year postr fi e and predictions of postr fi e sanitary condition of damaged period, when the intensity of postfire tree mortality was trees is based on an assessment of the state of tree crowns the most intensive. and visible damages without considering the main factors The sanitary state of the pine stand was character- controlling the fire resistance of trees. ized by the index of sanitary condition, which was deter- The main goal of this study is to find additional crite- mined by the formula [2] (Sanitary Forests Regulations ria that reflect the intensity of damage and the morpho- in Ukraine, 2016): logical indicators of fire resistance of pine trees for the I = K ×n + K ×n + K ×n + … K ×n / N [2] development of postfire tree mortality models. c 1 1 2 2 3 3 6 6 where: I – index of sanitary state; K ...K – category of sanitary c 1 6 state for individual tree (from I to VI); n ...n – number of trees 1 6 with specic fi sanitary category, stems; N – total number of trees 2. Materials and methods on a study plot. The study was conducted in the pine forests of the for- The sanitary state of each tree was determined accord- est steppe of Ukraine. The climate is mild, moderately ing to 6 categories (Table 1). −1 continental. The total precipitation is 546 mm per year , The degree of differentiation of trees in the stand was of which 38–40% falls during the growing season. The evaluated according to the Kraft classification (Pogreb- duration of the growing season lasts 205 days. The period nyak 1968): with an average daily temperature of +5 °C to +15 °C is I – predominant, exceptionally large trees, which 80–90 days. Climatic stress factors (high temperatures in dominate the others, have the highest height and the summer months with the absence of rainfalls, irregu- diameter and well-developed crowns; lar rainfall, long dry periods, evaporation exceeding pre- II – dominant trees, which have relatively well-devel- cipitation) cause repeated occurrence of forest fires. oped crowns and about the same height as class I Study plots (SP) were established in pine stands trees; homogeneous by forest type conditions and stand com- III – low co-dominant trees, normally developed, smaller position. in height than the trees of the previous classes, with The evaluation of fire damage to the tree trunks was less developed, compressed crowns; determined by the following indicators: IV – dominated trees whose crowns are compressed and – the bark char height (minimum and maximum), m; the tops reach only the lower part of the crown of – fire damage of thin (light) bark; the dominant trees; – relative bark char, %; V – entirely overtopped trees completely under the Relative bark char was determined by the formula [1]: canopy of dominant trees, far behind in growth and development. H = (H / H) × 100% [1] rel m char The postfire growth of the damaged trees was inves- where H – relative bark char, %; H– height of tree, m; H rel m char tigated on 33 study plots (3 SP in young pine stands, 21 – the maximum height of the bark char, m. SP in middle-aged and premature stands, and 10 SP in The evaluation of the crown damage was based on mature and overmature stands). The stands on the SP the following: differed in age and forestry characteristics (Table 2). – discoloration, % (the proportion of needles that have Stand structure was characterized by a curve of trees lost their natural color, with an accuracy of 10%, no distribution by natural degrees of thickness (NDT). later than a week after the fire damage) (Eichhorn Natural degrees of thickness are an indicator of the tree et al. 2010); size by diameter (DBH), expressed in relative propor- – defoliation, % (accuracy up to 10%). tion of the average diameter of the stand. (Tyurin 1945; The following tree characteristics were determined: Mashkovsky 2015). tree height, m; DBH, cm; class of Kraft; height of rough Table 1. Criteria for determining the sanitary index of pine stands. Needle packing Range of the sanitary index Needles color The degree of stand damage [%] distribution of needles on shoots 1.00–1.50 90–100 without signs of violation green absent 1.51–2.50 66–90 without signs of violation green, light green weak 2.51–3.50 33–66 clustered pale green medium 3.51–4.50 33 clustered with a yellow tint or yellow-green strong 4.51–6.00 0 there are no living needles gray, yellow very strong 4.51–6.00 0 there are no living needles — very strong 22 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 Table 2. Characteristics of study plots damaged by surface was used. In ROC-analysis, model quality is considered fires. to be excellent at AUC (area under curve) value 0.9–1.0; very good at 0.8–0.9; good at 0.7–0.8; average at 0.6–0.7; No SP H , [m] A [years] D [cm] H [m] Relative density 3 −1 char [m ha ] poor at 0.5–0.6 (Fawcett 2004). Yong pine stands Summer fires or spring fires While studying the influence of Kraft class (KC) on 1 1.20 11 5.3 3.8 9 0.90 the resistance to r fi e damage of trees with the same height 2 0.90 11 7.3 4.1 9 0.90 of bark char, a sampling of trees was selected from the Control — 11 5.5 4.0 9 0.90 Middle aged stands whole set of trees within the range of bark char height Summer fires 0.51–1.00 m differing in KC. Despite a small number of 1 0.69 60 27.2 24.5 386 0.77 2 2.02 57 26.0 21.0 476 1.00 I–II and IV–V KC trees, they were joined into groups: 3 2.30 57 21.1 18.7 238 0.58 I–II, III and IV–V KC. One-way ANOVA was used. The 4 2.52 60 28.4 25.2 481 0.93 5 3.58 68 25.3 20.8 256 0.62 power of the influence of the factor was determined by 6 4.45 70 26.5 21.2 261 0.62 the method of Plokhinsky (Lakyn 1990). 7 0.78 61 28.6 25.7 268 0.50 8 1.26 61 21.8 19.6 220 0.60 9 1.85 65 28.4 22.2 375 0.91 10 1.98 65 26.6 22.5 401 0.96 11 0.95 65 25.3 23.1 419 0.94 3. results Spring fires 12 2.05 60 25.8 22.9 384 0.82 3.1. Postfire tree mortality in the young pine 13 2.22 60 29.5 23.5 466 0.97 14 3.06 47 24.0 21.8 327 0.74 stands 15 0.20 66 26.3 23.2 411 0.86 16 0.30 59 27.7 22.4 442 0.96 A significant correlation (r = 0.92; p = 0.05) was found 17 0.30 60 29.5 23.1 419 0.88 between the level of discoloration and the sanitary condi- 18 0.80 60 25.8 22.9 380 0.80 19 2.97 60 22.6 21.0 389 0.91 tion of the trees. The highest intensity of mortality was 20 2.90 55 17.1 18.5 185 0.88 determined in the year of the r fi e, when 23.8% of the total Mature and overmature Spring fires number of trees. In the following year, the mortality rate 21 1.10 81 31.2 23.0 370 0.70 decreased, reaching 5.6% of trees, and 1.7% of trees in 22 3.00 81 37.3 26.2 390 0.60 the third year after the fire. 23 2.33 88 30.1 27.3 642 0.99 24 2.46 88 31.2 24.8 480 0.93 It has been found that the lethal level of damage to 25 1.81 86 34.7 26.0 375 0.69 young pine trees is achieved when the relative bark char 26 2.40 86 33.2 26.6 476 0.96 27 1.90 86 31.1 25.2 398 0.77 (H ) reaches 61–70%. In this case, mortality reaches rel. Summer fires 82% of trees, corresponding with the 80% of the growing 28 0.48 95 42.6 27.2 361 0.60 29 2.02 116 50.4 31.1 477 0.72 stock (Table 3). 30 2.16 116 42.6 29.8 385 0.60 Small damages of tree crowns (up to 30%) did not Note: SP – sampling plot; H – the average height of the bark char, m; A – age; D – diameter; char lead to tree mortality. Significant deterioration of the H – height; M – stock volume per 1 ha. sanitary state with the subsequent death of a significant number of trees was observed in the case of a severe dam- Classification of pine stands by age group was per - age to the crown – more than 81%. Trees with over 91% formed according to the following criteria (Svyrydenko damage to the crown died within a year after the fire. et al. 2004): – young (pine stands up to 40 years old); Table 3. Distribution of dead trees depending on the type – middle-aged (41 to 80 years old); and value level of damage in the pine young stands of Vasy- – premature stands (from 81 to 90 years); shchevske Forestry, State Enterprise Zhovtneve, Kharkiv re- – mature and overmature (over 90 years old). gion. T r e e r e s i s t a n c e w a s a s s e s s e d u s i n g P e a r s o n ’ s Proportion of dead trees [%] The value of damage rel. [%] D [%] correlation. Correlation analysis was used to assess [%] N M N M possible linear associations between tree sanitary index 0–10 6 1 — — 11–20 7 1 — — and potential variables that reflect fire resistance: DBH 21–30 8 1 — — (diameter at breast height), NDT, tree height, tree age, 31–40 25 20 — — bark thickness, rough bark height, etc. 41–50 35 32 — — 51–60 71 75 — — Multiple regression analysis as well as logistic regres- 61–70 82 80 3 1 sion analysis (binary regression) were used to construct 71–80 — — 4 1 81–90 — — 38 42 predictive models for the tree mortality probability of 91–100 94 92 94 96 individual trees. The use of logistic regression is appro- Note: H – relative bark char, %; D – crown discoloration, %; N – number of trees, stems, %; rel. M – percent of the stock, %. priate to determine the likelihood of individual trees mortality. Logistic regression quality and accuracy were tested by ROC analysis using IBM’s SPSS 20. For the Larger pine trees were more r fi e resistant. Character - analysis of model quality and its cut-off threshold cor- istics such as diameter, height and class of Kraft were the rection ROC-analysis (receiver operating characteristic) best indicators of fire resistance in young pine stands. 23 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 It has been confirmed that a model that considers of the regression analysis show that in 83% of cases after tree diameter, relative bark char, and crown discolora- summer fires, the sanitary state of trees is determined tion (AUC = 0.95 ± 0.012) performs best in predicting by the average bark char on trees. It was also proved the the probability of tree mortality. proportion of dead trees after the fire in the studied tree groups increases in response to the increasing average A simplified version of model, which included height of the bark char (r = 0.87, t = 5.80, t = 3.17) f 0.01 only the magnitude of crown discoloration, had an AUC (Fig. 1). value of 0.93 ± 0.014. Both models have “excellent” qual- \ [ ity in classification (Fawcett 2004), so we suggest to use 5ð the simplified model (3) that includes discoloration and I HHV WU predicts postfire mortality with an accuracy of 89.9%. RSRUWLRQRIGHDG 3U This model correctly predicts the postfire status of sur - KHLJKW%DUNFKDU  P viving trees with an accuracy of up to 94.3% and of dead trees up to 80%. The risk of trees mortality is extremely Fig. 1. Relationship between the proportion of dead trees and high with a discoloration of over 80%. average bark char height on the trunks in middle-aged pine stands damaged by surface fires in summer. ݁ݔ݌ሺെͻǤͲͻͷ ൅ͳͳͳǤͲ ൈܦ ሻ ܲ ൌ  [3] ሺͳ ൅݁ݔ݌ ሺ ͷͻͲǤͻെ ൅ͳͳǤͲ ͳൈܦ ሻ Ʋ The postfire models were further improved by con - sidering the distribution of trees by NDT (ranking trees where P is the probability of postfire mortality; D – crown dis- in the planting relative to the average diameter) that con- coloration,%. tains both the class of tree development (Kraft class) in the stand and its size (diameter). A negative correlation (r = −0.54; p = 0.05) was found between the sanitary 3.2. Postfire tree mortality in the middle-aged state and the NDT; a less close relationship was found pine stands between the diameters and the sanitary condition of the Middle aged middle-aged pine trees were found to be trees (r = 0.40; p = 0, 05). characterized by an increase in fire resistance of trees For NDT 0.7–1.2, lethal damage occurs at the height with an increase in trunk diameter (a correlation between of up to 3 m on the trunk (the index of sanitary state reac- the diameter and the sanitary condition of damaged trees hing 4.1–4.6). For trees with NDT more than 1.3, lethal was revealed: r = −0.4; p = 0.05). damage is achieved only if the height of bark char exceeds Stands with the same level of damage caused during 4 m (Table 4). fires that occurred in different seasons of the year were exhibiting different intensity of mortality. We found that Table 4. Proportion of dead trees depending on the natural de- mortality rate after summer fires can be 10 times higher grees of thickness and average bark char height after summer than after spring fires. Trees with the same value of dam- surface fires. NDT age (mean height of bark char) and age but damaged in H [m] char >0.6 0.7–0.8 0.9–1.0 1.1–1.2 1.3–1.4 1.5> different seasons had a different reaction to fire damage. >1.0 22 11 8 1 0 0 The index of sanitary state after spring fires was 2.48; 1.1–2.0 47 12 10 1 0 0 2.1–3.0 71 30 34 3 9 — after summer fires – 4.4 (ANOVA results: F = 79.8 and 3.1–4.0 75 42 41 29 0 — F = 12.4, h = 0.77). 4.1> — 100 43 36 40 33 0.001 Note: H – m; NDT – natural degree of thickness. After spring fires, the index of sanitary state, even char with significant damage to the trunk (bark char height above 4 m), ranged from 2.6 to 2.8 (weakened stand). An increase in the proportion of dead trees with the After the spring surface fires, only Kraft class IV–V trees increasing average height of bark char was detected in the died out. The percentage of dead trees increased in stands groups of trees that were different in NDT and the aver- with a larger proportion of IV and V Class Kraft (CK) age height of bark char. The group of trees depressed in trees. Thus, in damaged pine trees, the fire accelerated growth (NDT 0.5–0.7) is characterized by an extremely the natural liquefaction of the stands. The significance high proportion of dead trees, which increases rapidly of the influence of the investigated factor was proved by (from 22 to 75%) with the growth of damage intensity. comparing trees that are homogeneous in their level of For trees with the NDT of 0.7–1.0, the critical bark char damage and taxation characteristics but different in Kraft height is more than 2 m (the share of dead trees comprises classes (F = 6.47; F = 4.88). 30–34%). f 0.05 A strong direct relationship between the bark char The best performing models based on the logistic on the tree trunks and the sanitary state of the trees was regression analysis included the natural degree of thick- established (r = 0.91, t = 6.80, t = 3.17). The results ness and the average bark char height (H ) (Fig. 2) (4). f 0.01 char 24 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 ୣ୶୮ሺଶǤ଺଻ȂହǤଶ଴ൈே஽்ା଴Ǥ଺ଵൈ ு ሻ ೎೐ ೛ ܲ ൌ  [4] ൫ ୮୶ୣȂଵ ሺ ǤହȂ଻଺Ǥଶ ଴ଶ ൈே ்஽ ା଴Ǥ଺ଵൈ ு ሻሻ Ʋ ೎೐ ೛ where P stands for the probability of postfire tree mortality (from 0 to 1); NDT is the natural degree of thickness; H is char the average bark char height on the tree trunks. ,QGH[RIVDQLWDU\FRQGLWLRQ URXJ KEDUN  ]RQHRI  WKLQ EDU N WUDQ VLWLRQ DGDP J H Fig. 4. Index of sanitary condition for groups of trees which have equal height of bark char. 3.3.Postfire tree mortality in the mature pine stands Fig. 2. Probability of postfire mortality in middle-aged pine It was found that in mature pine stands there is no sig- stands depending on the height of bark char and the natural nificant relationship between bark char height and the degree of thickness (NDT). The accuracy of the prediction model reaches 78.2%. The high quality of the proposed model trees´ sanitary state. Taking into account the role of the was confirmed by ROC analysis (AUC = 0.83 ± 0.020). The char height on the trunks as an indicator of fire damage, inclusion of more variables in the model did not increase its tables of predicted mortality were drawn up separately accuracy by more than 1%. According to the sanitary condi- for summer and spring fires (Table 5). tion classification, the dead trees belonged to the fifth and sixth categories of the sanitary condition. “The dying trees” Table 5. Sanitary state and postfire mortality in mature pine susceptible to stem tree beetles (category IV of sanitary state) stands damaged by surface fires. that will probably die with time have not been taken into ac- Season of fire count, so the cut-off limit for the above model has been re- Spring Summer char duced to 0.38. Proportion of dead trees [%] Proportion of dead trees [%] [m] I I s s N M N M 0–0.5 2.7 5.9 2.6 4.3 33.0 38.4 The one-way ANOVA analysis confirmed the sig- 0.5–1.0 3.0 7.4 3.2 4.5 54.0 55.9 nificance of fire damage of thin (light) bark. Trees with 1.1–1.5 2.9 5.4 1.7 4.1 56.0 65.6 1.5–2.0 2.9 5.4 3.5 4.3 31.0 27.0 average NDT ranging from 0.9 to 1.1, same age and in 2.1–2.5 2.9 1.5 1.2 4.7 64.0 52.3 bark char height on the trunks (2.8–3.1 m), were selected 2.6–3.0 2.9 4.8 3.4 4.7 64.0 72.1 3.1–4.0 2.7 3.0 4.2 5.7 83.0 72.3 for analysis. The sample was divided into three groups: Note: Is – index of sanitary state; H – bark char height, %; N – number of trees, %; M – the char “rough bark” – when the bark char did not exceed the share of the stock, %. height of the rough bark; “transition zone” – when the bark char reached one meter below the zone of transition It was found that the damage caused by the spring to thin bark or reached the zone of transition of the coarse r fi es did not pose a threat – the proportion of dead trees in bark; and the third group, “thin (light) bark damage” – the total stand stock did not exceed 3.4%. This indicates the height of bark char exceeded the zone of transition of that the least developed trees die after the spring fires. the rough bark to the thin one. The effect of thin bark area After the summer r fi es, the trees affected by any dam - damage on the trees´ sanitary state was statistically sig- age responded with rapid deterioration of their sanitary nificant (Ff = 5.98; Fst = 3.07; (p = 0.003) (Fig. 3, Fig. 4). state. No significant relationship was found between the T h u s , s e l e c t i n g d r y i n g - sanitary state of trees and damage value, although there is prone trees for cutting (as trees a tendency toward an increase in the proportion of dead that are likely to die), it is worth trees as the average height of bark char increases (from paying attention to the pres- 31 to 83% by the number of trees and from 27 to 72.3% ence of thin bark burns. In cases by the stock). The proportion of dead trees by the number where the height of the bark char was inferior to the share of dead trees by the stock. This height exceeds the height of the indicates that after the summer surface fires, the most coarse bark (thin bark r fi e dam - developed dominant and predominant trees have a higher age), the likelihood of postfire probability of mortality. trees mortality increases dra- The suppressed small trees in the pine stand died at matically. any level of damage (natural thinning by suppression) Fig. 3. Transition zone: rough bark (Table 6). At the same time, the most developed trees goes into thin light color bark. 25 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 of the highest natural degree of thickness (NDT) also base of the tree). The deterioration of tree sanitary state responded to the damage most radically – the mortality or even the death of trees after the surface fire is caused rate was 48–55% even with minimal damage (the height by damage to stem tissues; buds and needles in the crown of bark char up to one meter) and 100% for the bark char of the tree; damage to tree roots (Valendik et al. 2006; above one meter. Kosov 2008), but the authors do not provide critical levels Trees whose DBH was inferior to the average in the of damage for each age group of pine stands. The bark char indicator is widely used alone or in combination stand, namely with a natural degree of thickness 0.7–0.8, with crown scorch to predict postfire mortality of pine were found to be the most fire resistant. The mortality rate in this group is minimal (0–5%). Thus, the diameter forests. These two indicators are considered to be the of mature and overmature pine trees is not an indicator most signic fi ant ones (De Bano & Conrad 1978; McHugh of fire resistance: the slightly less developed and thinner & Kolb 2003). Kobechinskaya and Oturina (1997) trees have proved more resistant. indicated that the height of bark char is not necessarily the main criterion for the mortality in a stand. According to their research in the Crimea, the proportion of dead pine Table 6. Proportion of dead trees in mature pine stands, de- pending on the natural degrees of thickness (NDT) and bark trees varies greatly, with the same height and different char height (H ) after summer smoldering fires, %. combustion and destruction of cambium. In our opinion, char Bark char height (H ) Natural degrees of thickness char such variation is caused by differences in the stand and >0.6 0.7–0.8 0.9–1.0 1.1–1.2 1.3–1.4 [m] individual trees characteristics. >1.0 43 5 48 54 55 1.1–2.0 — 0 69 100 100 According to Dieterich (1979), the amount of damage 2.1–3.0 — 0 100 100 100 to the crowns during r fi es has been identie fi d as the main 3.1–4.0 — 0 100 100 100 cause of coniferous tree mortality and has been success- fully used to predict postfire mortality. In our opinion, A tendency was established toward an increase in the crown scorch should be included in the prediction models share of dead trees with an increase in their diameter. of postfire tree mortality, especially for predicting mor - Correlation analysis established a strong direct reliable tality in young pine stands, where this type of damage is relationship (t = 5.84; t = 3.71) between these indica- f 0.01 typical during surface fires and closely correlates with tors. One of the reasons for the higher mortality of large- the deterioration of the sanitary state of damaged trees. sized trees was the bigger amount of litter accumulated The surface fires in the young pine stands are extremely near tree basis. Thus, in the undamaged parts of the pine dangerous. Because of the low height crown attachment, stands measurements were taken of the thickness of the the r fi e causes severe damage to both tree trunks and tree litter and duff at different distances from the trunk. crowns. Due to this, for young pine stands it is advisable The thickness of litter and duff layers near the tree to use also “relative bark char” (the ratio of maximum basis varied within 7 to 14 cm (average value –11.0 ± bark char height to the height of the tree) as an additional 0.52 cm) and decreased as the distance from the trunk indicator. This criterion is not new and was used Regel- increased to 2.6 ± 0.30 cm, which is statistically con- brugge and Conard (1993), but it is appropriate to use it firmed (F = 56,8; F = 2.7). Rough bark reaches only f 0,001 for young pine stands. This technique allows to predict the level of the litter, therefore surface r fi es severely dam - postfire mortality of trees more accurately, compared age the cambium, leading to the weakening of the tree to using the “char height” indicator. According to our and ultimately to dying. It was found that the thickness results, the lethal damage value of the crown scorch for of the bark on the pine root paws can be 2.5 times smaller young pine trees is 80% and relative bark char – 60%. than on the trunk above – 8.8 ± 0.99 mm versus 22.2 ± Critical for pine stands damaged during summer 2.29 mm. The reliability of the difference in bark thick- fires can be considered the average height of bark char ness on model trees was statistically conr fi med (t = 5.54; above two meters, the sanitary index being 3.4 and the t = 2.26). The presence of open root paws, on the one 0.05 proportion of dead trees reaching 21%. As the average hand, indicates damage to the first order roots, while on bark char height increased, so did the proportion of dead the other, is an indicator of a strong degree of forest lit- trees from 21% for bark char height of 1.5–2 m to 70% ter combustion, which causes also damage to fine roots. for bark char of more than 5 m the sanitary index ranging from 2.6 to 2.8. The rate of postfire tree mortality was insignificant (1–6%). The probability of postfire mortal- 4. Discussion ity of individual middle-aged trees after summer surface Our results show that the response of pine trees of fires is determined depending on the diameter (DBH) of different age groups to the effect of surface fire differs the trees and the value of damage to the trunk. Trees of greatly. In young pine stands, damage to the crown the lowest natural degree of thickness (NDT 0.6–0.7) are with simultaneous damage to the tree trunk prevails. killed by minor damage to the trunk (up to 1 m). The risk In middle-aged pine trees prevails damage to the trunk; of postfire mortality for trees with the NDT of 1.0–1.1 in mature pine stands – damage to the trunk by heat is significant only when the height of bark char is more radiation and root systems (roots of the first order at the than 3.5 m. 26 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 them, the rest of the higher order roots, which are related As is evidenced in major research papers (De Bano & Conrad 1978; Pinard & Huffman 1997; Stephens & to the first order roots, will also die (Guo et al. 2008). J. Finney 2002), tree mortality rate is inversely related to Varner (2009) demonstrates that the lethal temperature the diameter of the trees, since larger trees typically have during a fire, spreads deeper than 20 cm into the soil. a thicker bark, which is a good insulator. According to O’Brien found (O’Brien et al. 2010), that the fine roots of Pinus palustris Miller were almost evenly distributed in our research, the height of the coarse bark may be an additional indicator of fire resistance of trees, especially the lower layers of forest duff and developed in the upper for middle-age trees. The trees whose height of the bark 30 cm of soil. Although the thermal conductivity of the char reaches the zone of the thin bark (its thickness being soil is negligible, the fire can, in certain conditions, sig- 0.1–0.9 mm), respond to damage by deterioration of the nic fi antly damage the root systems. Therefore, if the trees sanitary state and have a higher probability of mortality. in the pine stand form a surface root system, a smolder- The thermal insulation properties of the bark layer ing fire can significantly affect the root systems. At the depend on its thickness, structure, bulk density and same time, a steady smoldering fire that damaged the humidity. These bark characteristics vary widely across cambium near the root (tree basis zone) can cause a long- tree species. The bark of coniferous tree species from fam- term weakening of the trees, reduce resistance to beetles´ ily Pinaceae (Pinus sylvestris L., Pinus pallasiana D. and attacks and lead to the death of the pine stand. Similar others) is the most successfully adapted for the thermal processes were recorded by Varner (2009) in mature pine isolation of living tissues from fire damage (Valendik et trees that were damaged during smoldering fire in the al. 2006). Some authors (Hare 1965) provide data about summer; such trees died even with slight damage (up to r fi e resistance, which is related to morphological features 0.5 m of bark char height). Varner, in his work, came to and increases with the size of the tree (diameter) and the same conclusion, n fi ding that in low intensity surface fires, mature pine trees can withstand fire damage, but a its age. Other authors have reported that the diameter and thickness of the bark are poorly correlated with the smoldering fire in planting with a thick layer of litter and postfire mortality rate (Menges & Deyrup 2001). Basi - duff can cause more than 80% of trees to die (Thies et al. cally, the conclusions are often contradictory. On the 2006; Varner et al. 2009). other hand, authors of similar works (Hood et al. 2007) The intensity of postfire mortality depends also on the season of fire (spring, summer, autumn). This phe - claim that trees with a larger diameter, on the contrary, are less resistant to fires. In our research, such conclu - nomenon was not clear due to a number of reasons. For sions are only partially confirmed. In mature pine trees, example, Menges and Deyrup (2001) argue that the especially after the summer fires and smoldering fires, intensity of post fire mortality is the highest after the the fire resistance of the most developed trees was negli- autumn and winter fires. Harrington (1987) claimed gible, they died after exposure to even minimal fire dam- that more intense postfire mortality occurred during age. Tree mortality may take place in case of the small spring fires (active season) than after autumn ones (dry thickness of the bark and its high thermal conductiv- season), while Thies (2005), on the other hand, deter- ity in trees with smaller diameter (DBH) (Kosov et al. mined greater tree mortality after fall than after spring. It 2005), so young and middle-aged pine trees larger in is logical to assume that the season itself, as a factor, has size proved more fire-resistant and had a better sanitary little effect on tree mortality, the main role being played by state one year after fire. In our opinion, the authors came the intensity and characteristic of the damage. That is, in to different conclusions because they did not take into the summer, the intensity of the fire is greater, especially account the fire intensity, damage to the roots and tree during prolonged droughts when forest fuel dries up to a morphological changes relative to its age. Our results critical level. This hypothesis has been confirmed in the have shown that the high rate of mortality and probability research of Thies (2006). Our results have shown that summer r fi es have a more negative impact on pine stands. of mortality of more developed large trees in mature pine stands is associated with the accumulation of a relatively Summer fires for middle-aged, mature and overmature larger stock of litter and duff at the tree base near the pine stands have more significant consequences than trunks. This contributes to the increased intensity of fire spring fires. The intensity of postfire tree mortality after and its localization near the tree trunk. The bark on the summer surface fires in mature pine trees was 10 times higher than after spring r fi es. Since spring r fi es are rapid, root paws (8.8 ± 0.99 mm) in mature pines is 2.5 times thinner than on the trunk 10 cm above ground (22.2 ± only the top thin layer of litter is burned during the fire, 2.29 mm). Therefore, summer surface fires damage the and the wetter ones do not burn out. Such type of damage cambium in this location, which leads to the weakening of is less significant than in summer, all the litter and duff the trees and their subsequent dying. According to Fowler burning, and damage to the trunk is also accompanied and Cieg (2004), damage to fine roots can be crucial for by damage to the root and the destruction of the most their postr fi e survival. Prolonged heat during smoldering physiologically active n fi e roots in the topsoil. In addition, fires can cause root damage when first order roots near trees damaged in early spring do not experience lack of the base of the trunk are damaged and die off. Along with moisture, and they recover faster after being damaged. 27 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 Guo, D. L., Mitchell, R. J., Withington, J. M., Fan, P. P., 5. Conclusions Hendricks, J. J., 2008: Endogenous and exogenous We conclude that fire resistance patterns in pine trees controls of root lifespan, mortality and nitrogen flux change during their growth and vary greatly between in a longleaf pine forest: root branch order predomi- different age groups. We have found evidence that in nates. Journal of Ecology, 96:737–745. young, middle-aged and premature pine stands the main Hare, R. C., 1965: The contribution of bark to fire resist- indicators of fire resistance are connected to the tree size ance of southern trees. Journal of Forestry, 4:248– – large trees appear more resistant. The main indicators of fire resistance were thus tree diameter, tree height, Harrington, M. G., 1987: Ponderosa pine mortality from NDT, Kraft class, and other related indicators. In mature spring, summer, and fall crown scorching. Journal of stands, on the contrary, larger trees were less resistant Applied Forestry, 2:14–16. to fire damage. The reason is that trees with larger diam- Hood, S. M., Smith, S. L., Cluck, D. R., 2007: Delayed eter in mature pines stands accumulate a relatively larger conifer tree mortality following fire in Califor- stock of fuel (litter and duff) near the tree trunks. This nia. USDA Forest Service Gen. Tech. Rep. PSW- contributes to the local increase of fire intensity and its GTR-203. USA, Albany, Calif., 261–283 p. localization near the tree trunk. Such local changes in Kobechinskaya, V. G., Oturina, I. P., 1997: Ecological fire behavior lead to severe damage to the lower part of consequences of the impact of fires on the vegetation cover of mountain Crimea. Issues of bioindication tree trunks and the death of damaged trees. The height and ecology of Zaporozhye, 2:28–31. of the coarse bark may be an additional indicator of pine Kosov, I. V., 2008: The mechanism of the impact of tree fire resistance. Trees whose height of the bark char grassroots fires on trees of coniferous species. Fires does not exceed the zone of the coarse bark have a better in Siberia’s Forest Ecosystems – Vseros Materials. sanitary condition and a higher probability of survival. Conf. with international. participation, 17–19 Sep- tember, 2008, Krasnoyarsk, p. 146–149. Kosov, I. V., Kisilyahov, E. K., Rybnikov, V. Yu., 2005: references The mechanism of damage to the stand during litter- Ahafonov, A. F., Alekseev, Y. A., 1989: Drying of pure humus fires. Nerd. researched in Siberia, Krasno - pine on fires. Forestry, 12:37–39. yarsk, 13:97–101. De Bano, L. F., Conrad, C. E., 1978: The effect of fire Lakyn, H. F., 1990: Biometrics. 4th, rework. Moscow, on nutrients in a chaparral ecosystem. Ecology, Vyshcha shkola, 352 p. 59:89–97. Mashkovsky, V. P., 2015: Commoditization of the rated Dieterich, J. H., 1979: Recovery potential of fire-dam - cutting by assortment tables with tree row distribu- aged southwestern ponderosa pine. USDA Forest tion according to natural diameter classes. Problems Service – Rocky Mountain Forest and Range Exper- of Forest and Forestry: collection of scientic fi works of iment Station Research Note RM-379. Ft. Collins, Institute of Wood of NAS of Belarus, 75:340–348 p. Colorado, 8 p. Mason, W. L., Alìa, R., 2000: Investigación agraria. Sis- Eichhorn, J., Roskams, P., Ferretti, M., Mues, V., Szepesi, temas y recursos forestales, 9:3–17. A., Durrant, D., 2010: Visual assessment of crown McHugh, C. W., Kolb, T. E., 2003: Ponderosa pine mor- condition and damaging agents. Manual Part IV. In: tality following r fi e in northern Arizona. International Manual on Methods and Criteria for Harmonized Journal of Wildland Fire, 12:7–22. Sampling, Assessment, Monitoring and Analysis of Menges, E. S., Deyrup, M. A., 2001: Postfire survival the Effects of Air Pollution on Forests – UNECE-ICP in south Florida slash pine: interacting effects of Forests Programme Co-ordinating Centre, Hamburg, fire intensity, fire season, vegetation, burn size, and 49 p. bark beetles. International Journal of Wildland Fire, Fawcett, T., 2004: ROC Graphs: Notes and Practical 10:53–63. Considerations for Researchers. Kluwer Academic O’Brien, J. J., Hiers, J. K., Mitchell, R. J., Morgan Var- Publishers, 38 p. ner III, J., Mordecai, K., 2010: Acute physiological Fites-Kaufman, J. A., Weixelman, D., Merrill, A. G., stress and mortality following r fi e in a long-unburned 2008: One-Year Postfire Mortality of Large Trees in longleaf pine ecosystem. Fire Ecology, 6:1–12. Low- and Moderate-Severity Portions of the Star Fire Pinard, M. A., Huffman, J., 1997: Fire resistance and bark in the Sierra Nevada 1. Session K – Postr fi e Mortality properties of trees in a seasonally dry forest in eastern of Large Trees in Sierra Nevada. USDA Forest Service Bolivia. Journal of Tropical Ecology, 13:727–740. General Technical Reports PSW-GTR-189:279–284. Pogrebnyak, P. S., 1968: General forestry. Moscow, Fowler, J. F., Sieg Fowler, C. H., 2004: Postfire mortality Kolos, 440 p. of ponderosa pine and Douglas-fir: a review of meth- Regelbrugge, J. C., Conard, S. G., 1993: Modeling tree ods to predict tree death. Gen. Tech. Rep. RMRS- mortality following wildfire in Pinus ponderosa for- GTR-132. USDA Forest Service, Fort Collins, Rocky ests in the central Sierra Nevada of California. Inter- Mountain Research Station, 25 p. national Journal of Wildland Fire, 3:139–148. 28 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 Sah, J., Ross, M., Ogarcak, D., Snyder, J., 2006: Fuel Tyurin, A. V., 1945: Forest mensuration. Moscow, Gos- Loads, Fire Severity, and Tree Mortality in Florida lestekhizdat, 376 p. Keys Pine Forests. International Journal of Wildland Usenya, V. V., Churylo, V. S., 2001: The dynamics of Fire, 15:463–478. post-fire mortality in pine stands. Problems of Forest San-Miguel-Ayanz, J., Durrant, T., Boca, R., Libertà, G., and Forestry: collection of scientic fi works of Institute Branco, A., de Rigo, D., Ferrari, D.et al., 2017: Forest of Wood of NAS of Belarus, 52:209–214. Fires in Europe, Middle East and North Africa, 142 p. Valendik, E. N., Suhinin, A. I., Kosov, I. V. 2006: Effect State Forest Resources Agency of Ukraine, 2017: Forests of forest r fi es on the stability of conifers Krasnoyarsk. and forestry in Ukraine. Sci. Bull. UNFU, 27: 10–15. Krasnoyarsk, Scientic fi Center of the Siberian Branch Stephens, S. L., Finney, M. A., 2002: Prescribed fire of the RAS, 96 p. mortality of Sierra Nevada mixed conifer tree species: Varner, J. M., Putz, F. E., O’Brien, J. J., Hiers, J. K., effects of crown damage and forest o fl or combustion. Mitchell, R. J., Gordon, D. R., 2009: Post-fire tree Ecology and Management, 162:261–271. stress and growth following smoldering duff fires. Svyrydenko, V. Ye., Babich, O. H., Kyrychok, L. S. 2004: Forest Ecology and Management, 258:2467–2474. Forestry. Textbook. Kyiv, Aristey, 544 p. Voron, V. P., Melnyk, Ye. Ye., 2009: Tendencies of fires Thies, W. G., Westlind, D. J., Loewen, M., 2005: Season development in the forests of green belt of Kharkov. of prescribed burn in ponderosa pine forests in east- Forestry and Forest Melioration, 115:207–214. ern Oregon: impact on pine mortality. International Internet sources: Journal of Wildland Fire, 14:223–231. FAO, 2006: Fire management: Review of international Thies, W. G., Westlind, D. J., Loewen, M., Brenner, G., cooperation. Fire Management Working Paper 2006: Prediction of delayed mortality of r fi e-damaged FM18E. Available at: https://www.fao.org/3/ ponderosa pine following prescribed fires in eastern j9406e/j9406e00. Oregon, USA. International Journal of Wildland Fire, Sanitary Forests Regulations in Ukraine, 2016: Approved 15:19–29. by Cabinet of Ministers of Ukraine. Kyiv, 20 p. Avail- able at: https://zakon.rada.gov.ua/laws/show/555- 95-%D0%BF#Text. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Forestry Journal de Gruyter

Postfire tree mortality and fire resistance patterns in pine forests of Ukraine

Download PDF
9 pages

Loading next page...
 
/lp/de-gruyter/postfire-tree-mortality-and-fire-resistance-patterns-in-pine-forests-wMe0BtGrqE
Publisher
de Gruyter
Copyright
© 2021 Serhii Sydorenko et al., published by Sciendo
ISSN
0323-1046
eISSN
0323-1046
DOI
10.2478/forj-2020-0029
Publisher site
See Article on Publisher Site

Abstract

The study was conducted in pure Scots pine (Pinus sylvestris L.) stands within the forest steppe physiographic region of Ukraine damaged by surface fires with different intensity. The aim of the research is to determine the effect of different fire intensity on pine stand and individual trees, considering tree morphometric parameters and type of damage. The intensity and duration of fire-related tree mortality was different in stands with different age. We found that tree fire resistance is driven by tree diameter, height of the rough bark, and natural degree of thickness. The pro- portion of dead trees one year after the spring fires in the middle-aged pine stands was 5 times lower and in mature pine stands even 10 times lower than after the summer fires. The critical damage to tree crowns in young pine trees causing their death is 80% of the needles burned. In the middle-aged pine trees, critical damage depended on the size of trees. The death of large, mature trees after smoldering summer fires was associated with the accumulation of a large stock of forest litter and duff near the tree-base, which contributed to the increased intensity of fire and its localization near the base part of the trees. Based on our findings, postfire tree mortality models have been developed for different age groups of pine stands. Key words: surface fires; smoldering fires; postfire mortality models; bark char; crown scorch; season of fire Editor: Tomáš Hlásny largest rivers, also in the Crimea Peninsula. More than 1. Introduction 90% of forest r fi es in Ukraine occur in pine forests (Voron Wildfires are one of the most dangerous factors for for- & Melnyk 2009; Voron & Sydorenko 2014). During the ests, leading to their destruction and degradation. Due last ten years (2009–2018), 19.9 thousand forest fires to global warming and increasing aridity, the risk of have occurred in Ukraine on an area of 37.2 hectares. The increased frequency and extent of wildfires is very high. economic losses amounted to 8.86 million Euros (Public Many regions of the world have experienced an increas- Report 2017). ing trend of excessive wildfires and an increasing occur- The rapid deterioration of the sanitary state of dam- rence of extremely severe fires (FAO 2006). aged trees leads to significant economic losses due to In 2017, wildfires burnt over 1.2 million hectares the decline of stand merchantability and the deteriora- of natural lands in the EU. The European Forest Fire tion of its technical quality. Timely diagnosis and accu- Information System estimated the amount of r fi e-related rate prediction of the postfire trees mortality are thus losses to be around 10 billion Euros (San-Miguel-Ayanz extremely important. Such studies can mitigate the et al. 2017). negative effects of wildfires and will be used to support In Europe, Scots pine forests now exceed 28 million decisions about forestry treatments in such forests. A hectares, covering over 20% of the productive forest area considerable amount of research papers is devoted to the (Mason & Alìa 2000). The most fire hazardous conifer- prediction of postfire loss of trees for different conifer ous forests occupy 43% of the total area of Ukraine, in species (Regelbrugge & Conard 1993; Usenya & Churylo particular, pine stands – 35%, which grow in the North 2001; Fites-Kaufman et al. 2008; Sah et al. 2019). It is of Ukraine (Polissya) and in the South (Steppe) along the well known that the value of postr fi e mortality rate affects *Corresponding author. Serhii Sydorenko, e-mail: serhii88sido@gmail.com, phone: +38 099 223 29 08 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 the predominant type of damage and r fi e intensity as well bark, m; number of r fi st-order roots above ground. Natu - as secondary disturbances such as drought, insects, and ral degrees of thickness (NDT) were additionally calcu- diseases (Ahafonov & Alekseev 1989). The research on lated for each plantation on the test plots. The date of r fi e, the impact of fire damage on postfire tree mortality in the date of measurement, the soil condition and the stand Ukraine is rather fragmented. The existing diagnostics age were recorded. We also examined a one-year postr fi e and predictions of postr fi e sanitary condition of damaged period, when the intensity of postfire tree mortality was trees is based on an assessment of the state of tree crowns the most intensive. and visible damages without considering the main factors The sanitary state of the pine stand was character- controlling the fire resistance of trees. ized by the index of sanitary condition, which was deter- The main goal of this study is to find additional crite- mined by the formula [2] (Sanitary Forests Regulations ria that reflect the intensity of damage and the morpho- in Ukraine, 2016): logical indicators of fire resistance of pine trees for the I = K ×n + K ×n + K ×n + … K ×n / N [2] development of postfire tree mortality models. c 1 1 2 2 3 3 6 6 where: I – index of sanitary state; K ...K – category of sanitary c 1 6 state for individual tree (from I to VI); n ...n – number of trees 1 6 with specic fi sanitary category, stems; N – total number of trees 2. Materials and methods on a study plot. The study was conducted in the pine forests of the for- The sanitary state of each tree was determined accord- est steppe of Ukraine. The climate is mild, moderately ing to 6 categories (Table 1). −1 continental. The total precipitation is 546 mm per year , The degree of differentiation of trees in the stand was of which 38–40% falls during the growing season. The evaluated according to the Kraft classification (Pogreb- duration of the growing season lasts 205 days. The period nyak 1968): with an average daily temperature of +5 °C to +15 °C is I – predominant, exceptionally large trees, which 80–90 days. Climatic stress factors (high temperatures in dominate the others, have the highest height and the summer months with the absence of rainfalls, irregu- diameter and well-developed crowns; lar rainfall, long dry periods, evaporation exceeding pre- II – dominant trees, which have relatively well-devel- cipitation) cause repeated occurrence of forest fires. oped crowns and about the same height as class I Study plots (SP) were established in pine stands trees; homogeneous by forest type conditions and stand com- III – low co-dominant trees, normally developed, smaller position. in height than the trees of the previous classes, with The evaluation of fire damage to the tree trunks was less developed, compressed crowns; determined by the following indicators: IV – dominated trees whose crowns are compressed and – the bark char height (minimum and maximum), m; the tops reach only the lower part of the crown of – fire damage of thin (light) bark; the dominant trees; – relative bark char, %; V – entirely overtopped trees completely under the Relative bark char was determined by the formula [1]: canopy of dominant trees, far behind in growth and development. H = (H / H) × 100% [1] rel m char The postfire growth of the damaged trees was inves- where H – relative bark char, %; H– height of tree, m; H rel m char tigated on 33 study plots (3 SP in young pine stands, 21 – the maximum height of the bark char, m. SP in middle-aged and premature stands, and 10 SP in The evaluation of the crown damage was based on mature and overmature stands). The stands on the SP the following: differed in age and forestry characteristics (Table 2). – discoloration, % (the proportion of needles that have Stand structure was characterized by a curve of trees lost their natural color, with an accuracy of 10%, no distribution by natural degrees of thickness (NDT). later than a week after the fire damage) (Eichhorn Natural degrees of thickness are an indicator of the tree et al. 2010); size by diameter (DBH), expressed in relative propor- – defoliation, % (accuracy up to 10%). tion of the average diameter of the stand. (Tyurin 1945; The following tree characteristics were determined: Mashkovsky 2015). tree height, m; DBH, cm; class of Kraft; height of rough Table 1. Criteria for determining the sanitary index of pine stands. Needle packing Range of the sanitary index Needles color The degree of stand damage [%] distribution of needles on shoots 1.00–1.50 90–100 without signs of violation green absent 1.51–2.50 66–90 without signs of violation green, light green weak 2.51–3.50 33–66 clustered pale green medium 3.51–4.50 33 clustered with a yellow tint or yellow-green strong 4.51–6.00 0 there are no living needles gray, yellow very strong 4.51–6.00 0 there are no living needles — very strong 22 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 Table 2. Characteristics of study plots damaged by surface was used. In ROC-analysis, model quality is considered fires. to be excellent at AUC (area under curve) value 0.9–1.0; very good at 0.8–0.9; good at 0.7–0.8; average at 0.6–0.7; No SP H , [m] A [years] D [cm] H [m] Relative density 3 −1 char [m ha ] poor at 0.5–0.6 (Fawcett 2004). Yong pine stands Summer fires or spring fires While studying the influence of Kraft class (KC) on 1 1.20 11 5.3 3.8 9 0.90 the resistance to r fi e damage of trees with the same height 2 0.90 11 7.3 4.1 9 0.90 of bark char, a sampling of trees was selected from the Control — 11 5.5 4.0 9 0.90 Middle aged stands whole set of trees within the range of bark char height Summer fires 0.51–1.00 m differing in KC. Despite a small number of 1 0.69 60 27.2 24.5 386 0.77 2 2.02 57 26.0 21.0 476 1.00 I–II and IV–V KC trees, they were joined into groups: 3 2.30 57 21.1 18.7 238 0.58 I–II, III and IV–V KC. One-way ANOVA was used. The 4 2.52 60 28.4 25.2 481 0.93 5 3.58 68 25.3 20.8 256 0.62 power of the influence of the factor was determined by 6 4.45 70 26.5 21.2 261 0.62 the method of Plokhinsky (Lakyn 1990). 7 0.78 61 28.6 25.7 268 0.50 8 1.26 61 21.8 19.6 220 0.60 9 1.85 65 28.4 22.2 375 0.91 10 1.98 65 26.6 22.5 401 0.96 11 0.95 65 25.3 23.1 419 0.94 3. results Spring fires 12 2.05 60 25.8 22.9 384 0.82 3.1. Postfire tree mortality in the young pine 13 2.22 60 29.5 23.5 466 0.97 14 3.06 47 24.0 21.8 327 0.74 stands 15 0.20 66 26.3 23.2 411 0.86 16 0.30 59 27.7 22.4 442 0.96 A significant correlation (r = 0.92; p = 0.05) was found 17 0.30 60 29.5 23.1 419 0.88 between the level of discoloration and the sanitary condi- 18 0.80 60 25.8 22.9 380 0.80 19 2.97 60 22.6 21.0 389 0.91 tion of the trees. The highest intensity of mortality was 20 2.90 55 17.1 18.5 185 0.88 determined in the year of the r fi e, when 23.8% of the total Mature and overmature Spring fires number of trees. In the following year, the mortality rate 21 1.10 81 31.2 23.0 370 0.70 decreased, reaching 5.6% of trees, and 1.7% of trees in 22 3.00 81 37.3 26.2 390 0.60 the third year after the fire. 23 2.33 88 30.1 27.3 642 0.99 24 2.46 88 31.2 24.8 480 0.93 It has been found that the lethal level of damage to 25 1.81 86 34.7 26.0 375 0.69 young pine trees is achieved when the relative bark char 26 2.40 86 33.2 26.6 476 0.96 27 1.90 86 31.1 25.2 398 0.77 (H ) reaches 61–70%. In this case, mortality reaches rel. Summer fires 82% of trees, corresponding with the 80% of the growing 28 0.48 95 42.6 27.2 361 0.60 29 2.02 116 50.4 31.1 477 0.72 stock (Table 3). 30 2.16 116 42.6 29.8 385 0.60 Small damages of tree crowns (up to 30%) did not Note: SP – sampling plot; H – the average height of the bark char, m; A – age; D – diameter; char lead to tree mortality. Significant deterioration of the H – height; M – stock volume per 1 ha. sanitary state with the subsequent death of a significant number of trees was observed in the case of a severe dam- Classification of pine stands by age group was per - age to the crown – more than 81%. Trees with over 91% formed according to the following criteria (Svyrydenko damage to the crown died within a year after the fire. et al. 2004): – young (pine stands up to 40 years old); Table 3. Distribution of dead trees depending on the type – middle-aged (41 to 80 years old); and value level of damage in the pine young stands of Vasy- – premature stands (from 81 to 90 years); shchevske Forestry, State Enterprise Zhovtneve, Kharkiv re- – mature and overmature (over 90 years old). gion. T r e e r e s i s t a n c e w a s a s s e s s e d u s i n g P e a r s o n ’ s Proportion of dead trees [%] The value of damage rel. [%] D [%] correlation. Correlation analysis was used to assess [%] N M N M possible linear associations between tree sanitary index 0–10 6 1 — — 11–20 7 1 — — and potential variables that reflect fire resistance: DBH 21–30 8 1 — — (diameter at breast height), NDT, tree height, tree age, 31–40 25 20 — — bark thickness, rough bark height, etc. 41–50 35 32 — — 51–60 71 75 — — Multiple regression analysis as well as logistic regres- 61–70 82 80 3 1 sion analysis (binary regression) were used to construct 71–80 — — 4 1 81–90 — — 38 42 predictive models for the tree mortality probability of 91–100 94 92 94 96 individual trees. The use of logistic regression is appro- Note: H – relative bark char, %; D – crown discoloration, %; N – number of trees, stems, %; rel. M – percent of the stock, %. priate to determine the likelihood of individual trees mortality. Logistic regression quality and accuracy were tested by ROC analysis using IBM’s SPSS 20. For the Larger pine trees were more r fi e resistant. Character - analysis of model quality and its cut-off threshold cor- istics such as diameter, height and class of Kraft were the rection ROC-analysis (receiver operating characteristic) best indicators of fire resistance in young pine stands. 23 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 It has been confirmed that a model that considers of the regression analysis show that in 83% of cases after tree diameter, relative bark char, and crown discolora- summer fires, the sanitary state of trees is determined tion (AUC = 0.95 ± 0.012) performs best in predicting by the average bark char on trees. It was also proved the the probability of tree mortality. proportion of dead trees after the fire in the studied tree groups increases in response to the increasing average A simplified version of model, which included height of the bark char (r = 0.87, t = 5.80, t = 3.17) f 0.01 only the magnitude of crown discoloration, had an AUC (Fig. 1). value of 0.93 ± 0.014. Both models have “excellent” qual- \ [ ity in classification (Fawcett 2004), so we suggest to use 5ð the simplified model (3) that includes discoloration and I HHV WU predicts postfire mortality with an accuracy of 89.9%. RSRUWLRQRIGHDG 3U This model correctly predicts the postfire status of sur - KHLJKW%DUNFKDU  P viving trees with an accuracy of up to 94.3% and of dead trees up to 80%. The risk of trees mortality is extremely Fig. 1. Relationship between the proportion of dead trees and high with a discoloration of over 80%. average bark char height on the trunks in middle-aged pine stands damaged by surface fires in summer. ݁ݔ݌ሺെͻǤͲͻͷ ൅ͳͳͳǤͲ ൈܦ ሻ ܲ ൌ  [3] ሺͳ ൅݁ݔ݌ ሺ ͷͻͲǤͻെ ൅ͳͳǤͲ ͳൈܦ ሻ Ʋ The postfire models were further improved by con - sidering the distribution of trees by NDT (ranking trees where P is the probability of postfire mortality; D – crown dis- in the planting relative to the average diameter) that con- coloration,%. tains both the class of tree development (Kraft class) in the stand and its size (diameter). A negative correlation (r = −0.54; p = 0.05) was found between the sanitary 3.2. Postfire tree mortality in the middle-aged state and the NDT; a less close relationship was found pine stands between the diameters and the sanitary condition of the Middle aged middle-aged pine trees were found to be trees (r = 0.40; p = 0, 05). characterized by an increase in fire resistance of trees For NDT 0.7–1.2, lethal damage occurs at the height with an increase in trunk diameter (a correlation between of up to 3 m on the trunk (the index of sanitary state reac- the diameter and the sanitary condition of damaged trees hing 4.1–4.6). For trees with NDT more than 1.3, lethal was revealed: r = −0.4; p = 0.05). damage is achieved only if the height of bark char exceeds Stands with the same level of damage caused during 4 m (Table 4). fires that occurred in different seasons of the year were exhibiting different intensity of mortality. We found that Table 4. Proportion of dead trees depending on the natural de- mortality rate after summer fires can be 10 times higher grees of thickness and average bark char height after summer than after spring fires. Trees with the same value of dam- surface fires. NDT age (mean height of bark char) and age but damaged in H [m] char >0.6 0.7–0.8 0.9–1.0 1.1–1.2 1.3–1.4 1.5> different seasons had a different reaction to fire damage. >1.0 22 11 8 1 0 0 The index of sanitary state after spring fires was 2.48; 1.1–2.0 47 12 10 1 0 0 2.1–3.0 71 30 34 3 9 — after summer fires – 4.4 (ANOVA results: F = 79.8 and 3.1–4.0 75 42 41 29 0 — F = 12.4, h = 0.77). 4.1> — 100 43 36 40 33 0.001 Note: H – m; NDT – natural degree of thickness. After spring fires, the index of sanitary state, even char with significant damage to the trunk (bark char height above 4 m), ranged from 2.6 to 2.8 (weakened stand). An increase in the proportion of dead trees with the After the spring surface fires, only Kraft class IV–V trees increasing average height of bark char was detected in the died out. The percentage of dead trees increased in stands groups of trees that were different in NDT and the aver- with a larger proportion of IV and V Class Kraft (CK) age height of bark char. The group of trees depressed in trees. Thus, in damaged pine trees, the fire accelerated growth (NDT 0.5–0.7) is characterized by an extremely the natural liquefaction of the stands. The significance high proportion of dead trees, which increases rapidly of the influence of the investigated factor was proved by (from 22 to 75%) with the growth of damage intensity. comparing trees that are homogeneous in their level of For trees with the NDT of 0.7–1.0, the critical bark char damage and taxation characteristics but different in Kraft height is more than 2 m (the share of dead trees comprises classes (F = 6.47; F = 4.88). 30–34%). f 0.05 A strong direct relationship between the bark char The best performing models based on the logistic on the tree trunks and the sanitary state of the trees was regression analysis included the natural degree of thick- established (r = 0.91, t = 6.80, t = 3.17). The results ness and the average bark char height (H ) (Fig. 2) (4). f 0.01 char 24 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 ୣ୶୮ሺଶǤ଺଻ȂହǤଶ଴ൈே஽்ା଴Ǥ଺ଵൈ ு ሻ ೎೐ ೛ ܲ ൌ  [4] ൫ ୮୶ୣȂଵ ሺ ǤହȂ଻଺Ǥଶ ଴ଶ ൈே ்஽ ା଴Ǥ଺ଵൈ ு ሻሻ Ʋ ೎೐ ೛ where P stands for the probability of postfire tree mortality (from 0 to 1); NDT is the natural degree of thickness; H is char the average bark char height on the tree trunks. ,QGH[RIVDQLWDU\FRQGLWLRQ URXJ KEDUN  ]RQHRI  WKLQ EDU N WUDQ VLWLRQ DGDP J H Fig. 4. Index of sanitary condition for groups of trees which have equal height of bark char. 3.3.Postfire tree mortality in the mature pine stands Fig. 2. Probability of postfire mortality in middle-aged pine It was found that in mature pine stands there is no sig- stands depending on the height of bark char and the natural nificant relationship between bark char height and the degree of thickness (NDT). The accuracy of the prediction model reaches 78.2%. The high quality of the proposed model trees´ sanitary state. Taking into account the role of the was confirmed by ROC analysis (AUC = 0.83 ± 0.020). The char height on the trunks as an indicator of fire damage, inclusion of more variables in the model did not increase its tables of predicted mortality were drawn up separately accuracy by more than 1%. According to the sanitary condi- for summer and spring fires (Table 5). tion classification, the dead trees belonged to the fifth and sixth categories of the sanitary condition. “The dying trees” Table 5. Sanitary state and postfire mortality in mature pine susceptible to stem tree beetles (category IV of sanitary state) stands damaged by surface fires. that will probably die with time have not been taken into ac- Season of fire count, so the cut-off limit for the above model has been re- Spring Summer char duced to 0.38. Proportion of dead trees [%] Proportion of dead trees [%] [m] I I s s N M N M 0–0.5 2.7 5.9 2.6 4.3 33.0 38.4 The one-way ANOVA analysis confirmed the sig- 0.5–1.0 3.0 7.4 3.2 4.5 54.0 55.9 nificance of fire damage of thin (light) bark. Trees with 1.1–1.5 2.9 5.4 1.7 4.1 56.0 65.6 1.5–2.0 2.9 5.4 3.5 4.3 31.0 27.0 average NDT ranging from 0.9 to 1.1, same age and in 2.1–2.5 2.9 1.5 1.2 4.7 64.0 52.3 bark char height on the trunks (2.8–3.1 m), were selected 2.6–3.0 2.9 4.8 3.4 4.7 64.0 72.1 3.1–4.0 2.7 3.0 4.2 5.7 83.0 72.3 for analysis. The sample was divided into three groups: Note: Is – index of sanitary state; H – bark char height, %; N – number of trees, %; M – the char “rough bark” – when the bark char did not exceed the share of the stock, %. height of the rough bark; “transition zone” – when the bark char reached one meter below the zone of transition It was found that the damage caused by the spring to thin bark or reached the zone of transition of the coarse r fi es did not pose a threat – the proportion of dead trees in bark; and the third group, “thin (light) bark damage” – the total stand stock did not exceed 3.4%. This indicates the height of bark char exceeded the zone of transition of that the least developed trees die after the spring fires. the rough bark to the thin one. The effect of thin bark area After the summer r fi es, the trees affected by any dam - damage on the trees´ sanitary state was statistically sig- age responded with rapid deterioration of their sanitary nificant (Ff = 5.98; Fst = 3.07; (p = 0.003) (Fig. 3, Fig. 4). state. No significant relationship was found between the T h u s , s e l e c t i n g d r y i n g - sanitary state of trees and damage value, although there is prone trees for cutting (as trees a tendency toward an increase in the proportion of dead that are likely to die), it is worth trees as the average height of bark char increases (from paying attention to the pres- 31 to 83% by the number of trees and from 27 to 72.3% ence of thin bark burns. In cases by the stock). The proportion of dead trees by the number where the height of the bark char was inferior to the share of dead trees by the stock. This height exceeds the height of the indicates that after the summer surface fires, the most coarse bark (thin bark r fi e dam - developed dominant and predominant trees have a higher age), the likelihood of postfire probability of mortality. trees mortality increases dra- The suppressed small trees in the pine stand died at matically. any level of damage (natural thinning by suppression) Fig. 3. Transition zone: rough bark (Table 6). At the same time, the most developed trees goes into thin light color bark. 25 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 of the highest natural degree of thickness (NDT) also base of the tree). The deterioration of tree sanitary state responded to the damage most radically – the mortality or even the death of trees after the surface fire is caused rate was 48–55% even with minimal damage (the height by damage to stem tissues; buds and needles in the crown of bark char up to one meter) and 100% for the bark char of the tree; damage to tree roots (Valendik et al. 2006; above one meter. Kosov 2008), but the authors do not provide critical levels Trees whose DBH was inferior to the average in the of damage for each age group of pine stands. The bark char indicator is widely used alone or in combination stand, namely with a natural degree of thickness 0.7–0.8, with crown scorch to predict postfire mortality of pine were found to be the most fire resistant. The mortality rate in this group is minimal (0–5%). Thus, the diameter forests. These two indicators are considered to be the of mature and overmature pine trees is not an indicator most signic fi ant ones (De Bano & Conrad 1978; McHugh of fire resistance: the slightly less developed and thinner & Kolb 2003). Kobechinskaya and Oturina (1997) trees have proved more resistant. indicated that the height of bark char is not necessarily the main criterion for the mortality in a stand. According to their research in the Crimea, the proportion of dead pine Table 6. Proportion of dead trees in mature pine stands, de- pending on the natural degrees of thickness (NDT) and bark trees varies greatly, with the same height and different char height (H ) after summer smoldering fires, %. combustion and destruction of cambium. In our opinion, char Bark char height (H ) Natural degrees of thickness char such variation is caused by differences in the stand and >0.6 0.7–0.8 0.9–1.0 1.1–1.2 1.3–1.4 [m] individual trees characteristics. >1.0 43 5 48 54 55 1.1–2.0 — 0 69 100 100 According to Dieterich (1979), the amount of damage 2.1–3.0 — 0 100 100 100 to the crowns during r fi es has been identie fi d as the main 3.1–4.0 — 0 100 100 100 cause of coniferous tree mortality and has been success- fully used to predict postfire mortality. In our opinion, A tendency was established toward an increase in the crown scorch should be included in the prediction models share of dead trees with an increase in their diameter. of postfire tree mortality, especially for predicting mor - Correlation analysis established a strong direct reliable tality in young pine stands, where this type of damage is relationship (t = 5.84; t = 3.71) between these indica- f 0.01 typical during surface fires and closely correlates with tors. One of the reasons for the higher mortality of large- the deterioration of the sanitary state of damaged trees. sized trees was the bigger amount of litter accumulated The surface fires in the young pine stands are extremely near tree basis. Thus, in the undamaged parts of the pine dangerous. Because of the low height crown attachment, stands measurements were taken of the thickness of the the r fi e causes severe damage to both tree trunks and tree litter and duff at different distances from the trunk. crowns. Due to this, for young pine stands it is advisable The thickness of litter and duff layers near the tree to use also “relative bark char” (the ratio of maximum basis varied within 7 to 14 cm (average value –11.0 ± bark char height to the height of the tree) as an additional 0.52 cm) and decreased as the distance from the trunk indicator. This criterion is not new and was used Regel- increased to 2.6 ± 0.30 cm, which is statistically con- brugge and Conard (1993), but it is appropriate to use it firmed (F = 56,8; F = 2.7). Rough bark reaches only f 0,001 for young pine stands. This technique allows to predict the level of the litter, therefore surface r fi es severely dam - postfire mortality of trees more accurately, compared age the cambium, leading to the weakening of the tree to using the “char height” indicator. According to our and ultimately to dying. It was found that the thickness results, the lethal damage value of the crown scorch for of the bark on the pine root paws can be 2.5 times smaller young pine trees is 80% and relative bark char – 60%. than on the trunk above – 8.8 ± 0.99 mm versus 22.2 ± Critical for pine stands damaged during summer 2.29 mm. The reliability of the difference in bark thick- fires can be considered the average height of bark char ness on model trees was statistically conr fi med (t = 5.54; above two meters, the sanitary index being 3.4 and the t = 2.26). The presence of open root paws, on the one 0.05 proportion of dead trees reaching 21%. As the average hand, indicates damage to the first order roots, while on bark char height increased, so did the proportion of dead the other, is an indicator of a strong degree of forest lit- trees from 21% for bark char height of 1.5–2 m to 70% ter combustion, which causes also damage to fine roots. for bark char of more than 5 m the sanitary index ranging from 2.6 to 2.8. The rate of postfire tree mortality was insignificant (1–6%). The probability of postfire mortal- 4. Discussion ity of individual middle-aged trees after summer surface Our results show that the response of pine trees of fires is determined depending on the diameter (DBH) of different age groups to the effect of surface fire differs the trees and the value of damage to the trunk. Trees of greatly. In young pine stands, damage to the crown the lowest natural degree of thickness (NDT 0.6–0.7) are with simultaneous damage to the tree trunk prevails. killed by minor damage to the trunk (up to 1 m). The risk In middle-aged pine trees prevails damage to the trunk; of postfire mortality for trees with the NDT of 1.0–1.1 in mature pine stands – damage to the trunk by heat is significant only when the height of bark char is more radiation and root systems (roots of the first order at the than 3.5 m. 26 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 them, the rest of the higher order roots, which are related As is evidenced in major research papers (De Bano & Conrad 1978; Pinard & Huffman 1997; Stephens & to the first order roots, will also die (Guo et al. 2008). J. Finney 2002), tree mortality rate is inversely related to Varner (2009) demonstrates that the lethal temperature the diameter of the trees, since larger trees typically have during a fire, spreads deeper than 20 cm into the soil. a thicker bark, which is a good insulator. According to O’Brien found (O’Brien et al. 2010), that the fine roots of Pinus palustris Miller were almost evenly distributed in our research, the height of the coarse bark may be an additional indicator of fire resistance of trees, especially the lower layers of forest duff and developed in the upper for middle-age trees. The trees whose height of the bark 30 cm of soil. Although the thermal conductivity of the char reaches the zone of the thin bark (its thickness being soil is negligible, the fire can, in certain conditions, sig- 0.1–0.9 mm), respond to damage by deterioration of the nic fi antly damage the root systems. Therefore, if the trees sanitary state and have a higher probability of mortality. in the pine stand form a surface root system, a smolder- The thermal insulation properties of the bark layer ing fire can significantly affect the root systems. At the depend on its thickness, structure, bulk density and same time, a steady smoldering fire that damaged the humidity. These bark characteristics vary widely across cambium near the root (tree basis zone) can cause a long- tree species. The bark of coniferous tree species from fam- term weakening of the trees, reduce resistance to beetles´ ily Pinaceae (Pinus sylvestris L., Pinus pallasiana D. and attacks and lead to the death of the pine stand. Similar others) is the most successfully adapted for the thermal processes were recorded by Varner (2009) in mature pine isolation of living tissues from fire damage (Valendik et trees that were damaged during smoldering fire in the al. 2006). Some authors (Hare 1965) provide data about summer; such trees died even with slight damage (up to r fi e resistance, which is related to morphological features 0.5 m of bark char height). Varner, in his work, came to and increases with the size of the tree (diameter) and the same conclusion, n fi ding that in low intensity surface fires, mature pine trees can withstand fire damage, but a its age. Other authors have reported that the diameter and thickness of the bark are poorly correlated with the smoldering fire in planting with a thick layer of litter and postfire mortality rate (Menges & Deyrup 2001). Basi - duff can cause more than 80% of trees to die (Thies et al. cally, the conclusions are often contradictory. On the 2006; Varner et al. 2009). other hand, authors of similar works (Hood et al. 2007) The intensity of postfire mortality depends also on the season of fire (spring, summer, autumn). This phe - claim that trees with a larger diameter, on the contrary, are less resistant to fires. In our research, such conclu - nomenon was not clear due to a number of reasons. For sions are only partially confirmed. In mature pine trees, example, Menges and Deyrup (2001) argue that the especially after the summer fires and smoldering fires, intensity of post fire mortality is the highest after the the fire resistance of the most developed trees was negli- autumn and winter fires. Harrington (1987) claimed gible, they died after exposure to even minimal fire dam- that more intense postfire mortality occurred during age. Tree mortality may take place in case of the small spring fires (active season) than after autumn ones (dry thickness of the bark and its high thermal conductiv- season), while Thies (2005), on the other hand, deter- ity in trees with smaller diameter (DBH) (Kosov et al. mined greater tree mortality after fall than after spring. It 2005), so young and middle-aged pine trees larger in is logical to assume that the season itself, as a factor, has size proved more fire-resistant and had a better sanitary little effect on tree mortality, the main role being played by state one year after fire. In our opinion, the authors came the intensity and characteristic of the damage. That is, in to different conclusions because they did not take into the summer, the intensity of the fire is greater, especially account the fire intensity, damage to the roots and tree during prolonged droughts when forest fuel dries up to a morphological changes relative to its age. Our results critical level. This hypothesis has been confirmed in the have shown that the high rate of mortality and probability research of Thies (2006). Our results have shown that summer r fi es have a more negative impact on pine stands. of mortality of more developed large trees in mature pine stands is associated with the accumulation of a relatively Summer fires for middle-aged, mature and overmature larger stock of litter and duff at the tree base near the pine stands have more significant consequences than trunks. This contributes to the increased intensity of fire spring fires. The intensity of postfire tree mortality after and its localization near the tree trunk. The bark on the summer surface fires in mature pine trees was 10 times higher than after spring r fi es. Since spring r fi es are rapid, root paws (8.8 ± 0.99 mm) in mature pines is 2.5 times thinner than on the trunk 10 cm above ground (22.2 ± only the top thin layer of litter is burned during the fire, 2.29 mm). Therefore, summer surface fires damage the and the wetter ones do not burn out. Such type of damage cambium in this location, which leads to the weakening of is less significant than in summer, all the litter and duff the trees and their subsequent dying. According to Fowler burning, and damage to the trunk is also accompanied and Cieg (2004), damage to fine roots can be crucial for by damage to the root and the destruction of the most their postr fi e survival. Prolonged heat during smoldering physiologically active n fi e roots in the topsoil. In addition, fires can cause root damage when first order roots near trees damaged in early spring do not experience lack of the base of the trunk are damaged and die off. Along with moisture, and they recover faster after being damaged. 27 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 Guo, D. L., Mitchell, R. J., Withington, J. M., Fan, P. P., 5. Conclusions Hendricks, J. J., 2008: Endogenous and exogenous We conclude that fire resistance patterns in pine trees controls of root lifespan, mortality and nitrogen flux change during their growth and vary greatly between in a longleaf pine forest: root branch order predomi- different age groups. We have found evidence that in nates. Journal of Ecology, 96:737–745. young, middle-aged and premature pine stands the main Hare, R. C., 1965: The contribution of bark to fire resist- indicators of fire resistance are connected to the tree size ance of southern trees. Journal of Forestry, 4:248– – large trees appear more resistant. The main indicators of fire resistance were thus tree diameter, tree height, Harrington, M. G., 1987: Ponderosa pine mortality from NDT, Kraft class, and other related indicators. In mature spring, summer, and fall crown scorching. Journal of stands, on the contrary, larger trees were less resistant Applied Forestry, 2:14–16. to fire damage. The reason is that trees with larger diam- Hood, S. M., Smith, S. L., Cluck, D. R., 2007: Delayed eter in mature pines stands accumulate a relatively larger conifer tree mortality following fire in Califor- stock of fuel (litter and duff) near the tree trunks. This nia. USDA Forest Service Gen. Tech. Rep. PSW- contributes to the local increase of fire intensity and its GTR-203. USA, Albany, Calif., 261–283 p. localization near the tree trunk. Such local changes in Kobechinskaya, V. G., Oturina, I. P., 1997: Ecological fire behavior lead to severe damage to the lower part of consequences of the impact of fires on the vegetation cover of mountain Crimea. Issues of bioindication tree trunks and the death of damaged trees. The height and ecology of Zaporozhye, 2:28–31. of the coarse bark may be an additional indicator of pine Kosov, I. V., 2008: The mechanism of the impact of tree fire resistance. Trees whose height of the bark char grassroots fires on trees of coniferous species. Fires does not exceed the zone of the coarse bark have a better in Siberia’s Forest Ecosystems – Vseros Materials. sanitary condition and a higher probability of survival. Conf. with international. participation, 17–19 Sep- tember, 2008, Krasnoyarsk, p. 146–149. Kosov, I. V., Kisilyahov, E. K., Rybnikov, V. Yu., 2005: references The mechanism of damage to the stand during litter- Ahafonov, A. F., Alekseev, Y. A., 1989: Drying of pure humus fires. Nerd. researched in Siberia, Krasno - pine on fires. Forestry, 12:37–39. yarsk, 13:97–101. De Bano, L. F., Conrad, C. E., 1978: The effect of fire Lakyn, H. F., 1990: Biometrics. 4th, rework. Moscow, on nutrients in a chaparral ecosystem. Ecology, Vyshcha shkola, 352 p. 59:89–97. Mashkovsky, V. P., 2015: Commoditization of the rated Dieterich, J. H., 1979: Recovery potential of fire-dam - cutting by assortment tables with tree row distribu- aged southwestern ponderosa pine. USDA Forest tion according to natural diameter classes. Problems Service – Rocky Mountain Forest and Range Exper- of Forest and Forestry: collection of scientic fi works of iment Station Research Note RM-379. Ft. Collins, Institute of Wood of NAS of Belarus, 75:340–348 p. Colorado, 8 p. Mason, W. L., Alìa, R., 2000: Investigación agraria. Sis- Eichhorn, J., Roskams, P., Ferretti, M., Mues, V., Szepesi, temas y recursos forestales, 9:3–17. A., Durrant, D., 2010: Visual assessment of crown McHugh, C. W., Kolb, T. E., 2003: Ponderosa pine mor- condition and damaging agents. Manual Part IV. In: tality following r fi e in northern Arizona. International Manual on Methods and Criteria for Harmonized Journal of Wildland Fire, 12:7–22. Sampling, Assessment, Monitoring and Analysis of Menges, E. S., Deyrup, M. A., 2001: Postfire survival the Effects of Air Pollution on Forests – UNECE-ICP in south Florida slash pine: interacting effects of Forests Programme Co-ordinating Centre, Hamburg, fire intensity, fire season, vegetation, burn size, and 49 p. bark beetles. International Journal of Wildland Fire, Fawcett, T., 2004: ROC Graphs: Notes and Practical 10:53–63. Considerations for Researchers. Kluwer Academic O’Brien, J. J., Hiers, J. K., Mitchell, R. J., Morgan Var- Publishers, 38 p. ner III, J., Mordecai, K., 2010: Acute physiological Fites-Kaufman, J. A., Weixelman, D., Merrill, A. G., stress and mortality following r fi e in a long-unburned 2008: One-Year Postfire Mortality of Large Trees in longleaf pine ecosystem. Fire Ecology, 6:1–12. Low- and Moderate-Severity Portions of the Star Fire Pinard, M. A., Huffman, J., 1997: Fire resistance and bark in the Sierra Nevada 1. Session K – Postr fi e Mortality properties of trees in a seasonally dry forest in eastern of Large Trees in Sierra Nevada. USDA Forest Service Bolivia. Journal of Tropical Ecology, 13:727–740. General Technical Reports PSW-GTR-189:279–284. Pogrebnyak, P. S., 1968: General forestry. Moscow, Fowler, J. F., Sieg Fowler, C. H., 2004: Postfire mortality Kolos, 440 p. of ponderosa pine and Douglas-fir: a review of meth- Regelbrugge, J. C., Conard, S. G., 1993: Modeling tree ods to predict tree death. Gen. Tech. Rep. RMRS- mortality following wildfire in Pinus ponderosa for- GTR-132. USDA Forest Service, Fort Collins, Rocky ests in the central Sierra Nevada of California. Inter- Mountain Research Station, 25 p. national Journal of Wildland Fire, 3:139–148. 28 S. Sydorenko et al. / Cent. Eur. For. J. 67 (2021) 21–29 Sah, J., Ross, M., Ogarcak, D., Snyder, J., 2006: Fuel Tyurin, A. V., 1945: Forest mensuration. Moscow, Gos- Loads, Fire Severity, and Tree Mortality in Florida lestekhizdat, 376 p. Keys Pine Forests. International Journal of Wildland Usenya, V. V., Churylo, V. S., 2001: The dynamics of Fire, 15:463–478. post-fire mortality in pine stands. Problems of Forest San-Miguel-Ayanz, J., Durrant, T., Boca, R., Libertà, G., and Forestry: collection of scientic fi works of Institute Branco, A., de Rigo, D., Ferrari, D.et al., 2017: Forest of Wood of NAS of Belarus, 52:209–214. Fires in Europe, Middle East and North Africa, 142 p. Valendik, E. N., Suhinin, A. I., Kosov, I. V. 2006: Effect State Forest Resources Agency of Ukraine, 2017: Forests of forest r fi es on the stability of conifers Krasnoyarsk. and forestry in Ukraine. Sci. Bull. UNFU, 27: 10–15. Krasnoyarsk, Scientic fi Center of the Siberian Branch Stephens, S. L., Finney, M. A., 2002: Prescribed fire of the RAS, 96 p. mortality of Sierra Nevada mixed conifer tree species: Varner, J. M., Putz, F. E., O’Brien, J. J., Hiers, J. K., effects of crown damage and forest o fl or combustion. Mitchell, R. J., Gordon, D. R., 2009: Post-fire tree Ecology and Management, 162:261–271. stress and growth following smoldering duff fires. Svyrydenko, V. Ye., Babich, O. H., Kyrychok, L. S. 2004: Forest Ecology and Management, 258:2467–2474. Forestry. Textbook. Kyiv, Aristey, 544 p. Voron, V. P., Melnyk, Ye. Ye., 2009: Tendencies of fires Thies, W. G., Westlind, D. J., Loewen, M., 2005: Season development in the forests of green belt of Kharkov. of prescribed burn in ponderosa pine forests in east- Forestry and Forest Melioration, 115:207–214. ern Oregon: impact on pine mortality. International Internet sources: Journal of Wildland Fire, 14:223–231. FAO, 2006: Fire management: Review of international Thies, W. G., Westlind, D. J., Loewen, M., Brenner, G., cooperation. Fire Management Working Paper 2006: Prediction of delayed mortality of r fi e-damaged FM18E. Available at: https://www.fao.org/3/ ponderosa pine following prescribed fires in eastern j9406e/j9406e00. Oregon, USA. International Journal of Wildland Fire, Sanitary Forests Regulations in Ukraine, 2016: Approved 15:19–29. by Cabinet of Ministers of Ukraine. Kyiv, 20 p. Avail- able at: https://zakon.rada.gov.ua/laws/show/555- 95-%D0%BF#Text.

Journal

Forestry Journal – de Gruyter

Published: Mar 1, 2021

Keywords: surface fires; smoldering fires; postfire mortality models; bark char; crown scorch; season of fire

There are no references for this article.