Fire sensitivity of broadleaf tree species in plantations of Kurdistan, Iran

Sabri Baqer Rasooli · Shiva Rashidi ·Naghi Shabanian

Abstract To better manage wildf ires and plantations in Kurdistan, native tree species commonly used in planting and reviving forests such as Quercus brantii Lindl., Q. infectoria Olivier, Q. libani Olivier, Pistachio atlantica Desf.,Fraxinus rotundifolia Vahl, and Robinia pseudoacacia L.were evaluated and compared in terms of f ire sensitivity and f ire resistance. To determine f ire sensitivity, indices such as the f lammability index, ignition time, f lame durability,moisture content, carbonized surface, mass reduction, bulk density, as well as dry weight of wood, bark, and leaves were used. The data were subjected to ANOVA. The means obtained, after examining homogeneity or heterogeneity of the variances, were compared using parametric and nonparametric comparison tests. There was a signif icant diff erence at the level of 5%. Based on the analysis of these indices, the species could be divided into three groups: species sensitive to f ire such as Robinia pseudoacacia, species moderately sensitive to f ire such as F. rotundifolia, and species relatively resistant to f ire, including Q. brantii, Q. libani, Q. infectoria,and P. atlantica.

Keywords Fire · Flammability index · Time to ignition ·Flame durability · Moisture content

Introduction

Wildf ire always threatens a wide part of the world’s forests.This occurrence destroys thousands of hectares of vegetation (Hirschberger 2016) so that the annual average f ires globally has been estimated to occur over 6 to 14 million ha(Encinas et al. 2007). Wildf ire events are one of the major destructive factors to forest ecosystems, inf licting irreversible damage (Marozas et al. 2007). In addition to causing economic losses and environmental pollution, and aff ecting regional climates, f ire plays a major role in destroying fauna and f lora as well as the forested landscape with its Effects felt at every level of the ecosystem (Sowmya and Somashekar 2010). In addition, with high intensity, extent, and frequency of f ire, the qualitative value of species can change and lead to the introduction of species of low economic value (Podur et al. 2002). Therefore, because of the occurrence of f ire and on-oing destruction of the world’s forests, special attention has been given to the establishment of plantations and creating green spaces around cities in many countries. This issue is highly important in a country like Iran which is poor in forest areas and green spaces and should be given priority.Therefore, in addition to monitoring and identifying various aspects of f ire, the occurrence of f ire might be largely constrained by taking preventive measures and choosing Species that are more resistant to f ire for plantations (Jaiswal et al. 2002). Unfortunately, in many locations in Iran, where the establishment ion of plantations in areas such as urban green spaces, precise information about the species and their f ire sensitivity is lacking and many failures have occurred at considerable cost. Therefore, given the importance of this issue, a solution for reducing damage caused by wildf ires is needed. An Effective solution could be the selection of f ire resistant species for plantations and for the restoring of natural forests (Barrett et al. 2011).

For plantations, species should be selected which, in addition to achieving the planting objectives, will have the highest compatibility with climatic, environmental, economic,and social conditions of the region, whether the species are native or introduced (Koneshloo 2001).

Numerous studies have been carried out on various species in terms of f ire adaptation from amongst which are the following examples: Dubey et al. ( 1999) measured thermal conductivity levels of 11 tree species in India and established the relationship between thermal conductivity and bulk density; they also showed that the thermal conductivity of broad-leafed species would increase with an increase in bulk density. Scarff and Westboy ( 2006) in a survey in southeastern Australia onAcacia aneuraF.Muell. ex Benth.,Acacia doratoxylon A.Cunn.,andAcacia omalophyllaA.Cunn. ex Benth.,reported that the ignitibility of leaf litter depended on biomass. Leaf size was a major factor aff ecting f ire resistance, with the ignitibility of larger leaves being higher Proenca et al. ( 2010). examined f ire intensity in the canopies of broadleaf forests (Quercus roburL., Ilex aquifoliumL.) and coniferous forests (Pinus pinasterAit., Pinus sylvestrisL.) in Portugal and observed that the intensity of f ire in broadleaf stands was less than in coniferous forests.Alvarez et al. ( 2011) identif ied potential crown f ires and fuel types in Aleppo pine forests (Pinus halepensisMill.) of the Mediterranean basin. Data from a wildf ire was used to categorize forest structure into fuel classes as a function of crown f ire potential. Fourteen forest structures were identif ied featuring various forest types in the Western Mediterranean region based on canopy coverage, the number of tree layers, total tree density, and the percentage of each tree layer. By utilizing a f ire type patterns which burn most forest structures, four f ire hazard classes were identif ied based on fuel types. The 1st and 2nd classes were the lowest active crown f ire risk, being diff erent in the proportion of passive crown f ires and surface f ires. The 3rd f ire hazard class included the threshold among the structures with high and low f ire patterns, while the 4th was based on a high crown f ire risk Nasiri ( 2012). calculated the time required for burning 10 tree species,Carpinus betulusL., Alnus glutinosa(L) Gaertn., Parrotia persica(DC.), Fagus orientalisLipsky, Pterocarya fraxinifolia(Lam.) Spach, Acer cappadocicumGled, Acer velutinumBoiss, Cupresus sempervirencevarhorizontalis(Mill.) G. Don, Quercus castaneifoliaC.A.Mey.,andZelkova carpinifolia(Pall.) K.Koch, in two dry wood and wet wood (with natural moisture) states, as well as with bark or without bark. The f indings showed thatP. persicaandC. betulushad the highest and lowest levels of f ire resistance, respectively. In addition, wet wood resistance to burning was 2.3 times forQ. castaenifolia, compared to 2.9 timesP. persicaof dry wood. In addition, resistance to burning increased withAcer velutinum, Acer cappadocicum, F. orientalis,andZelkova carpinifolia(Pall.) K.Koch.

According Nasiri ( 2012), the presence of tree bark increased the burning resistance by some species so that the diff erence in burning times between wet and dry wood ofF. orientalis,C. sempervirencevarhorizontalis, A. velutinum, A. cappadocicum,andZ. carpinifolia, amounted to 46, 82, 110, 116,and 169 s, respectively. Kauf et al. ( 2014 ) reported leaf litter test results, accompanied by past research on similar materials, and suggested that materials with low ignition frequencies could be very f lammable under specif ic conditions. In addition, they noted that information measured with a unique f lammability grade, although in certain instances useful,could not be often meaningful and should be regarded with caution. Kauf et al. ( 2015) studied the relationship between f lammability parameters and morphological features of leaf litter by considering local and seasonal variabilities. Factors such as fuel moisture content, specif ic leaf area and average mass showed high f luctuations in terms of f lammability.The Effects of location and sampling time were also highly signif icant. Because of the standard test circumstances, such Effects were related to alterations to the internal features of the leaf litter. Since the Effects were inconsistent and specif ic to some species, the f indings might possibly restrict the generalization of the f lammability rankings. Wyse et al. ( 2016)quantif ied the ignitibility of 60 exotic and indigenous species in New Zealand and contrasted their experientially obtained rankings with expert evaluations. The most ignitable species included the introduced common gorse (Ulex europaeusL.), followed by manna gum (Eucalyptus viminalisLabill.),kumarahou (Pomaderris kumerahoA.Cunn. ex Fenzl), rimu(Dacrydium cupressinumSoland. ex Forst.f.) and silver beech (Lophozonia menziesii(Hook.f.) Heenan & Smissen).The rankings were highly correlated with the expert evaluations, thereby giving support to both methods. Giraldo et al.( 2016) evaluated seven Mexican tropical species for their f ire behavior. The results showed diff erent behavior patterns by all species, with highly signif icant diff erences in a number of cases. Although some correlations existed between the hardness of wood and density of the stands and their f ire behavior, such elements are often uncertain. In other species,some features, including the composition and anatomy of the wood are more Effective in achieving more predictable f ire behavior patterns.

In this study, in order to better manage f ire and for the successful implementation of forest plantation programs in Kurdistan, a number of important tree species often used for forest plantations and restoration plans were selected.These broad-leaved species such asQ. brantii, Q. infectoria, Q. libani, Pistasia atlanticaDesf., Fraxinus rotundifolia Vahl,andRobinia pseudoacaciaL. are to be evaluated and compared in terms of f ire sensitivity and f ire resistance.To determine f ire sensitivity, indices such as f lammability index, ignition time, f lame durability, moisture content, carbonized surface, mass reduction, bulk density, as well as dry weight of wood, bark, and leaves were used.

Materials and methods

Native and introduced species

Quercus brantiiLindl., is native and from the Fagaceae family. This family has 8 genus and 1000 species, of whichQuercushas about 400 species of trees and shrubs around the world. The height of this species in Zagros forests is about 8 m.

Quercus infectoriaOlivier also is native and from the Fagaceae family and less widely spread in Zagros forests thanQ. brantiiLindl. as it is specif ic to soil and climate.

Quercus libaniOlivier is another species of the Fagaceae family and is sensitive to climatic and soil factors and is more limited in Zagros.

Pistasia atlanticaDesf. is from theAnacardiaceaefamily and afterQuercus,is the second most important species of the Zagros forests, being of high economic and social status due to its valuable resin and fruit.

Robinia pseudocaciaL. is from thePapilnaceaefamily,a non-native, introduced species, one of the best species in dry and cold regions, and could be used as a pioneer species.It is relatively fast growing in the ecological conditions of Kurdistan, having been planted successfully in the cities of Sanandaj and Marivan.Fraxinus rotundifoliaVahl, is from theOleaceafamily,a non-native, introduced species, one with low water demand and drought tolerant and could be used as a pioneer species(Feraydooni 2012).

Study area

The indigenous species were from the Shuy village region.[In addition, the planted species were from Hassan-Abad Forest Park] (Fig. 1). The Shuy habitat is seven km northwest of the city of Baneh and between 45°50ʹ E and 45°54ʹE longitude and 36°2ʹ N and 36°6ʹ N latitude, has a northwestern direction and a slope of 20% at an altitude of 1525 m a.s.l. Average annual precipitation and average annual temperatures are 657 mm and 13.8 °C, respectively. Based on the De Martonne’s climatic classif ication, the region has a semi-humid climate (Salehzadeh et al. 2016).

Hassan-Abad Forest Park is located between 46°52ʹE and 46°55ʹE longitude and 35°15ʹ N and 35°19ʹ N latitude, and was established in 1968 to create a green space over 144.7 hectares on the eastern slope of the Abidar Mountain, 5 km south of Sanandaj. Average slope is 30% with altitudes of 1550 m a.s.l. Tree species includeF. rotundifolia, R. pseudoacacia, Biota orientalis(L.) Endl., Pinus nigraJ.F.Arnold,P. eldaricaMedw., Cupressus arizonicaGreene,andCeltis orientalisL.. Minimum temperature is − 5 °C in February with a maximum of 37 °C in August. Average annual precipitation is 484 mm (Aminpour et al. 2007).

Methods

Wood and bark samples were collected in late November and leaf samples in late July from 15-cm saplings and diameters> 10 cm. The samples to were cut into 15 cm × 5 cm pieces with 1-cm diameter and transferred to a wood f lammability measuring device. The number of replications for each of the investigated parameters was f ive. Leaf samples were transferred to the Forest Biology Lab and tested immediately after being detached from trees. In tests on the resistance and resistivity of the leaf samples, the only measurement was the ignition time due to the low resistance of leaf textures to f lames. A 0.5 g mass of each species was placed on a metal mesh over an alcohol burner for the direct contact of f lames with the samples, and the time for ignition was recorded using a chronometer.

The collected data were examined for normality using the Kolmogorov-Smirnov test and subjected to an ANOVA in a fully randomized method. In the case of accepting hypothesis 1, (the existence of a signif icant diff erence among the samples), to compare the means and select the best sample,the means were compared after examining the homogeneity of variances using Duncan’s multiple range test, in the case of the homogeneity of variances, or using Dunnett’s test, in the case of the heterogeneity of variances. Statistical analysis was performed using SPSS software, and Excel software was used to draw the f igures.

Results

Woods

The woods of the species were tested using the methods mentioned, with the results shown in the following tables and f igures.

Table 1 shows, in terms of all parameters in all species studied a signif icant diff erence at 5%. In other words, the null hypothesis is rejected for all features studied.

In order to compare the means for the features examined for all species and to determine the best sample, the means were compared with each other using the mean comparison test, with the results of which are shown in Fig. 2.

Fig. 1 Study area in Kurdistan province, Iran

Bark

Bark samples were examined and compared as mentioned,with the results demonstrated in the following table and f igure.

As Table 2 shows, for all features examined, there is a signif icant diff erence at 5%. In order to compare the means for the features examined and to determine the best sample, the means were compared together using the Duncan’s multiple range test, with the results shown in Fig. 3.

Leaves

As noted previously, because of their thin texture, leaves have low resistance to f lames. Therefore, among the parameters examined, ‘time to ignition’ was the only measurable parameter (Table 3).

As Table 3 shows, in terms of the feature ‘time to ignition’, there is a signif icant diff erence at 5% among the samples and the null hypothesis is rejected.

To compare the means obtained for ‘time to ignition’ of the various species and to determine the best treatment, the means were compared with each other using the mean comparison test (Fig. 4).

Discussion

Fuel moisture content and f lammability are the main features inf luencing f lammability and f ire behavior (Chuvieco et al.2004; Anderson and Anderson 2009; White and Zipperer 2010; Bilgili et al. 2019). The variance analysis results show signif icant diff erence in terms of moisture content among the woods of the species studied (Table 1). The higher the moisture content of the wood, the greater amount of heat and time needed to evaporate its water, with its resistance to f ire being higher (Atreya and Abu-Zaid 1991; Nasiri 2012). Considering this, f ire resistance byP. atlanticahas been the highest,with the other species in the following order:Q. libani, Q.brantii, Q. infectoria, F. rotundifolia,andR. pseudoacacia.However, it should be noted that, in addition to moisturecontent, other factors including the type and amount of chemicals in the wood such as resins, are involved in wood ignition; thus, the level of f ire resistance by species should not be based only on moisture content. However, moisture content could be an appropriate criterion for the assessment of f ire resistance or sensitivity (Mitsopoulos and Dimitrakopoulos 2014). The results of the present study are compatible with those of Nasiri ( 2012 ). He also concluded that the f ire resistance of wet wood was between 2.3 times (Quercus castaneifoliaC.A.Mey.) to 2.9 times (Carpinus betulusL.)greater than dry wood. After investigating the moisture content in layers of leaf litter and its possible Effects on the occurrence of f ire in Alaska, Jandt et al ( 2005) concluded that, with an increase in the moisture content, the possibility of f ire would decrease.

Table 1 Data variance analysis of the wood features of the species studied

In terms of f lammability index, there are signif icant differences among the wood samples (Table 1). As (Fig. 2 b)demonstrates,P. atlantica(1032.6 s) andR. pseudoacacia(75.4 s) required the highest and lowest times to reach attain f lammability, respectively. In general, species in a region,even under similar environmental conditions, may be categorized in diff erent groups in terms of the f lammability index.These results support the f indings of Petriccione et al. ( 2006)in which, after determining the f lammability indices of species in the Mediterranean region, concluded that diff erent species under similar environments diff ered in terms of time needed to reach f lammability.

In terms of the ‘time to ignition’, there is signif icant differences among the species in the present study (Table 1).According to the results from the comparison of the means and (Fig. 2 b),Q. libaniandR. pseudoacaciahad the longest and shortest ignition time.

The higher the f lammability index and the time to ignition, the longer it will require to burn, i.e., its f ire resistance will be higher (Vallette 1992). However, this is not a def initive conclusion because other factors such as species, wood texture, and resin content can aff ect experiment results.

SinceP. atlanticahas the highest moisture content and requires a longer time to reach its f lammability index, it can be concluded that a species with a high moisture content also has a high f lammability index, and will have high f ire resistance.

In terms of ‘f lame durability’, the wood ofR. pseudoacacia(112.6 s) had the highest f lame durability among the species (Fig. 2 b). In addition, the results obtained by other researchers indicate that diff erent species have diff erent f lammability indices, time to ignition, and f lame durability,with these being due to diff erences in their moisture content,wood texture, density, and hydrocarbon compositions. Scarffand Westboy ( 2006) carried out a study on speciesAcaciaaneuraF.Muell. ex Benth., Acacia doratoxylonA.Cunn.,andAcacia omalophylla,in southeastern Australia, and reported that the f lammability of leaf litter depended on its composition, biomass, as well as its physical and chemical features.

Fig. 2 Results of the means of the features examined; a: moisture content, b: f lammability index, time to ignition, and f lame durability, c: carbonized surface, d: mass reduction of wood, e: bulk density, f: dry weight of wood

With regards to the carbonized surface area,P. atlantica(32.3 cm 2 ) had the largest area. The ‘carbonized surface’and the ‘f lammability index’ are positively correlated. The higher the f lammability index, the longer the wood will be exposed to direct f lames; hence, a wider surface will be subject to direct f lames and will be carbonized.

Of the wood samples studied,F. rotundifolia(99.51%)andP. atlantica(98.24%) had the highest and lowest amounts of mass reduction (Fig. 2 d). The amount of mass reduction in various woods depends on factors such as texture, hydrocarbons, materials extracted from them and moisture content (Ebrahimi 2016). However, considering theresults in the present study,F. rotundifolia, showed higher mass reduction under test conditions and the lowest level of f ire resistance; in contrast,P. atlanticawas more stable under these conditions. The results obtained for this parameter are opposite to the ones for ‘f lammability index’ and‘carbonized surface’. In other words, this study has shown that species with a higher f lammability index and a higher carbonized surface has a lower percentage of mass reduction.In addition, in terms of ‘bulk density’ and the ‘dry weight of wood’,P. atlanticahad the highest amounts, implying that denser species have the lowest percentage of mass reduction.

Table 2 Data variance analysis of bark features

As Table 1 illustrates, there is a statistically signif icant difference in ‘bulk density’ among the wood samples.According to (Fig. 2 e),P. atlantica(1.398 g/cm 3 ) andF.rotundifolia, (0.802 g/cm 3 ) had the highest and lowest bulk density. In general, wood is a thermal insulator, so the lower the bulk density, the lower their thermal conductivity, and the lower their f ire resistance. Hence porous and light woods are less thermally conductive, i.e., they have low f ire resistance (Ebrahimi 2016). These results are comparable with Dubey et al. ( 1999), in which the thermal conductivity of 11 tree species in India was investigated and concluded that the thermal conductivity of broadleaf species would increase upon an increase in bulk density, i.e., f ire resistance would increase.

In terms of the ‘dry weight of wood’, as (Fig. 2 f) shows,P. atlantica(64.4 g) andF. rotundifolia, (52.71 g) had the highest and lowest amounts of ‘dry weight of wood’ among the samples. According to (Fig. 2 e and f), there is a positive correlation between bulk density and the dry weight of wood. Species with a high dry weight have high bulk density, being more f ire resistant, i.e., they require more time to burn and have a high f lammability index.

Variance analysis results (Table 2) show a signif icant diff erence in moisture content of the bark of the species studied. Mean comparison results (Fig. 3 a) show that the highest and the lowest percentage of bark moisture were withQuercus infectoria, (26.97%) andR. pseudoacacia(2.51%). To express the f ire resistance levels by considering only moisture without considering other factors,Q.infectoriaandR. pseudoacaciahad the highest and lowest levels of f ire resistance, which is consistent with the results of Vallette et al. ( 1992). In their research on the f lammability index of various Mediterranean species during various seasons of the year, they found a negative correlation between the dryness level and the ignition time. In other words, the higher the degree of dryness, the less time for ignition. In summers when moisture levels are lower, the f lammability of most species is near the maximum. However, as noted previously, there are other factors such as the type of species and their texture, as well as the amount of hydrocarbons and extracted materials which are highly Effective in determining f ire resistance and species stability. For example, Vallette et al. (1992) noted that the shrub,Cytisus trif lorusLam., had a low f lammability index even in summer.

Fig. 3 Results of the means of the features examined in the bark samples; a: moisture content, b: f lammability index, time to ignition, and f lame durability, c: carbonized surface, d: mass reduction, e: bulk density, f: dry weight of bark

Table 3 Data variance analysis of the time to ignition of leaf samples

The highest and lowest f lammability indices in the bark samples were found inQ. infectoria(362 s) andR. pseudoacacia.(16 s) (Fig. 3 b).

Q. libani,(120.4 s) andR. pseudoacacia(20.4 s) had the highest and the lowest levels of ‘time to ignition’, i.e., the bark ofQ. libani,reached the ignition point later than other species; hence, it has a higher f ire resistance.

Fig. 4 ‘Time to ignition’ in leaf samples

According to (Fig. 3 b), in terms of f lame durability,P.atlantica(87.2 s) andR. pseudoacacia(20.8 s) had the highest and lowest f lame durability.

With regards to the area carbonized,Q. infectoria(16.83 cm 2 ) andR. pseudoacacia(2.27 cm 2 ) had the highest and lowest carbonized surfaces (Fig. 3 c. According to the results of (Fig. 3 b) and (Fig. 3 c),Q. infectoriaOlivier andR. pseudoacacialso had the highest and lowest ‘flammability index’. Therefore, species with a high f lammability index take longer to burn and they have high carbonized surfaces as well.

With regards to ‘mass reduction’ (Fig. 3 d),F. rotundifolia(99.56%), andP. atlantica(98.24%) have the highest and lowest percentage of mass reduction, respectively. Based on this f inding, lighter woods have lower percentage of mass reduction as well.

According to (Fig. 3 e),P. atlantica(0.282 g/cm 3 ) had the highest bulk density.

(Fig. 3 b and e) indicate that species with high bulk densities have high f lammability and ignition indices; i.e., they require more time to burn due to high amounts of wood f iber and have high f ire resistance. In addition, according to (Fig. 3 d), denser species have lower percentage of mass reduction in the burning process with a higher amount of mass left after burning. According to the results of this study, species with high bulk densities have more wood f iber and therefore high levels of f ire resistance.

In terms of the dry weight the bark samples,P. atlantica(26.53 g) andR. pseudoacacia(1.68 g) had the highest and lowest dry weights, respectively (Fig. 3 f).

Based on the results of this study, species with high bulk densities have high dry weights, i.e., species with high dry weights have high levels of f ire resistance.

According to (Fig. 4) Q brantii(32.2 s) andF. rotundifolia(12.4 s) had the longest and shortest ignition times.It may be concluded that leaf shape and size highly inf luence its f lammability index. In a study by Papio and Trabaud( 1990) on the f lammability index of Mediterranean shrub species, includingPistacia lenticus L., Phillyrea angustifoliaL., andRosmarinus offi cinalisL.,the f lammability index increased with an increase in their surface-to-size ratio.

Conclusions

Based on the results of this study, is a signif icant diff erence was shown among the species in terms of the indices examined, and based on the analysis of the indices altogether, the species can be divided into three general categories: (1) highly f ire-sensitiveRobinia pseudoacaciaL.;(2) moderately f ire-sensitiveFraxinus rotundifoliaVahl;and (3) low f ire-sensitive or f ire resistantQuercus brantiiOlivier, Quercus libaniOlivier, Quercus infectoriaOlivier,andPistacia atlantica.Given the region’s history of f ire frequency, if ecological conditions are favorable, species with high f ire resistance should be planted.

AcknowledgmentsSpecial thanks goes to Mr. Rhmani, the head of the Forest Biology Laboratory of the Faculty of Natural Resources of the University of Kurdistan for his valuable support.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affi liations.