Forest fire occurrence shows clear relationships with the numbers of nearby inhabitants and tourists in the Czech Republic

doi:10.21203/rs.3.rs-2321318/v1

Download PDF

Research Article

Forest fire occurrence shows clear relationships with the numbers of nearby inhabitants and tourists in the Czech Republic

https://doi.org/10.21203/rs.3.rs-2321318/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

The vulnerability of forests to fire results from complex interactions among climate, fuel availability (fuel load and moisture content of the vegetation), and ignition sources. The number of forest fires (FFs) has increased in many regions, therefore, it is necessary to reduce and monitor the fire risk. Based on data from 2006 to 2015, we used Generalized Additive Models to determine the degree to which local climate, the forest–urban interface, the percentage of conifers, the number of overnight tourists, and the number of local human inhabitants (residents) are related to the FF frequency in the Czech Republic (Central Europe).

Results

On a monthly scale, the FFs incidence showed distinct spring (April) and summer (July-August) peaks. No distinct pattern was identified on an annual scale, yet the highest number of FFs occurred in 2015, the hottest year in our records. The used predictors explained 45 and 46% of the variability in FFs on monthly and summer scales, respectively, and 69% on an annual scale. The number of FFs was related to the number of residents and the number of overnight tourists ha^− 1 y^− 1 of the forest. The effect of climate was manifested on monthly and summer scales only, with warmer and drier conditions associated with higher FF frequency. A higher proportion of conifers and the length of the forest-urban interface were positively associated with FF too. Finally, FF frequency was associated with the population density and number of overnights, suggesting the importance of human behavior in fire risk.

Conclusions

The significant relationships between the numbers of FFs and the number of residents and overnight tourists ha^− 1 y^− 1 of forest suggest that the risk of FFs could be controlled by increasing public awareness and implementing stricter regulations on tourist and local inhabitants’ behaviour.

Central Europe

Climate

Fire risk

Human influence

Wildfire

Climate change and its effects on forest structure and composition will potentially significantly affect forest fire (FF) risk in the future (Field et al. 2016; Kasischke and Turetsky, 2006; Turetsky et al., 2015; Westerling 2016). The need for improved FFs monitoring, modelling, and prediction is increasingly recognized worldwide (Abatzoglou and Williams 2016). Forest managers need FF risk assessment to establish effective fire prevention systems and allocate firefighting resources (e.g., Calheiros et al. 2021; Müller et al. 2020; Elia et al. 2020).

Forest vulnerability to fire (Miller and Ager 2013) results from complex interactions between climate, fuel availability (fuel load and moisture content of the vegetation), and ignition sources (Zachar 2009; Emmons 1973; Osvald 1997; Roy 2003; Thomas and McAlpine 2010; Moritz et al. 2011). Ignition sources may be natural (predominantly lightning) or anthropogenic (i.e., resulting from human activities). The probability of human-caused fire ignition results from the direct or indirect presence of human activity in the landscape (Martinez et al. 2009). Research has indicated that roughly 90% of FF are human-caused and that only a small percentage are caused by lightning (Cardille et al. 2001; Grissino-Mayer et al. 2004; Mollicone et al. 2006; Vacik et al. 2011). Lightning-caused fires are ignited by cloud-to-ground lightning strikes. In contrast, human-caused fires result from a variety of human activities, including recreation (e.g., camping, hiking, and hunting) and industry (e.g., timber production, railway and highway transportation, and oil and gas exploration) (Wotton et al. 2003).

FFs in the temperate forest zone are mainly associated with human activities, while the climatic effects are not so obvious (Clark and Merkt 1989; Donis et al. 2017; Ellenberg 1996; Tinner et al. 2005; Niklasson et al. 2010; Rodrigues et al. 2020). In Europe, human activities account for most fire ignitions (Leone et al. 2002; Catry et al. 2009; Martinez et al. 2009). Although human factors are important when analyzing forest fire ignition, little attention has been paid to their quantitative assessment, with human activity data in forests being surprisingly scarce (Martell et al. 1987; Vega-Garcia et al. 1995; Parks et al. 2018). The one of the reason is the relatively high complexity of human behavior, which is often difficult to measure. As a result, the human contribution to wildfire ignition is difficult to assess (see Vasilakos et al. 2007; Martinez et al. 2009). In the past, estimations of the contribution of human activities to wildfire ignition have been based on indirect assessments (Donogue and Main 1985; Martinez et al. 2009). Direct assessments of the human contribution to wildfire ignition must be included in FF risk models.

The forests of Central Europe are undergoing a major change due to an ongoing bark beetle outbreak, which is destroying vegetation on an unprecedented scale, and which is therefore generating large quantities of dead wood (e.g., Vanická et al. 2020). When a forest stand is attacked by bark beetles, the stand loses its canopy; therefore, sunlight reaches the soil surface and supports the growth of grasses, and over time, small branches fall from the dry trees. These changes significantly increase the flammable fuel in the stand (Hicke et al. 2012; Hart et al. 2015). In connection with a change in human lifestyles caused by the current pandemic, the recreational use of the forest is increasing (Weinbrenner et al. 2021). It seems likely that a higher human presence in forests can increase the occurrence of FFs.

This research aims to determine the drivers of FF risk in the Czech Republic, mainly to disentangle the natural effects, such as forest composition and climate, and effects related to human activity. We hypothesize that FF frequency has increased in the recent decade, and the number of FFs is related to factors such as human population density and recreation rather than natural factors. Because of the distribution of forest fires (Fig. 1), we performed three types of analysis: month scale, summer scale, and annual scale.

Study area

The Czech Republic has a mild four-season climate that is between the oceanic and continental. The country` climate is characterized by the prevailing western winds, intense cyclonic activity, and relatively high precipitation. The average daily temperature ranges from about 3°C in January to 17°C in July, and the average annual rainfall ranges from 600 to 800 mm in most of the country (Tolasz et al. 2007). The altitude ranges from 115 to 1,603 m a.s.l. with a median of 430 m a.s.l. The predominant type of relief is hills and highlands. The average population density is 133 people.km^− 2 (Pokorná et al. 2019).

In the absence of humans, mixed beech forests would dominate in most of the Czech Republic, with oak-dominated deciduous forests in the lowlands and coniferous forests with spruce at higher altitudes (Chytrý 2012). Because of the intensive forest management performed since the 19th century, the current forest composition differs significantly from the natural state. Forests currently cover 33.9% of the territory and consist predominantly of Picea abies (52%), Pinus sylvestris (17%), Fagus sylvatica (7%), Quercus spp. (7%), Larix decidua (4%), Betula pendula (3%), and Abies alba (1%). Other deciduous species (e.g., Carpinus betulus, Acer spp., Fraxinus spp., Populus spp., Salix spp., and Tilia spp.) occupy about 8% of the forested area (Ústav pro hospodářskou úpravu lesů 2019).

Forest Fire Database

Detailed information on FFs were obtained from the Fire Rescue System of the Czech Republic (Hasičský záchranný sbor České republiky 2019), which keeps statistics on all fires in the country that required intervention. The database covers the whole area of the Czech Republic in the period 2006–2015 (except for military areas which represent < 5% of the area). From 2006 to 2010, the database indicated the districts where FFs occurred. The exact geographical locations were not indicated until after 2010.

Because of problems with the clear identification of some FFs, all records (approx. 10,000) were manually checked for misclassification; based on the descriptive information provided for each fire, the fire was excluded or included in the final database (FFD) (Berčák et al. 2018; Holuša et al. 2018). Also, FFs that occurred between January-March and November-December were not included in the final FFD (around 100 FFs in total), because these FFs were mostly associated with events such as accidental fuel leaks or fireworks. In addition, due to the minimal forest area, FFs from the administrative district of the capital city (Prague with more than 1,500,000 inhabitants) were also removed from the analysis.

In total, the FFD contains data on 7,105 FFs, with an average of 710 FFs per year over the reporting period.

The Dependent Variable

As the dependent variable, we used the numbers of FFs in individual municipality districts (n = 205) (State Administration of Land Surveying and Cadastre 2019) rather that than accurate spatial coordinates of the FFs for the following reasons: (i) data for the period 2006–2010 indicate FF occurrence in districts but without coordinates within districts; and (ii) socio-economic data (population, tourism, etc.) are only available for districts; and (iii) burned area is not very useful variable for understanding FF characteristics (Mohammadi et al. 2021). FFs in the Czech Republic are often extinguished quickly, and the small resulting burned areas reflect the effectiveness of the intervention rather than the nature of the FF ignition (Berčák et al. 2018). Moreover, burned area depends on fuel quality, but it is not related to the ignition.

Predictor Variables

We used several predictors, which represent key factors driving the fire risk in the temperate forests; climate data, forest-related data, socio-economic data, and landscape context data (Appendix 1) (Clark and Merkt 1989; Emmons 1973; Ellenberg 1996; Osvald 1997; Roy 2003; Flannigan et al., 2005; Tinner et al. 2005; Martinez et al. 2009; Zachar 2009; Thomas and McAlpine 2010; Niklasson et al. 2010; Aldersley et al. 2011; Zumbrunnen et al. 2012; Donis et al. 2017; Vacchiano et al. 2018; Dupire et al. 2019; Müller et al. 2020; Elia et al. 2020; Calheiros et al. 2021).

On a monthly and summer scale, we used the average monthly air temperature and monthly rainfall as predictor variables. On an annual scale, we used the average annual air temperature, the annual total rainfall, and the number of days with snow cover per year as predictor variables. The climate data and their differentiation were used from the climate atlas and specific data for the studied period measured by Czech Hydrometeorological Institution (2019).

Forest-related data at the district level were obtained from the national forest statistics (Ústav pro hospodářskou úpravu lesů 2019). The predictor variables used were the percentage of forest area, the percentage of conifers (%), and the percentage of pine (eAGRI 2019). In the Eurasian boreal zone, FFs predominantly occur in pine forests (Engelmark 1987; Tanskanen 2007), because pine trees are the most flammable (Ubysz and Valette 2010). In addition, pine forests usually grow under relative dry conditions, produce resinous and easily combustible debris, and in the case of older stands have a relatively thin canopy that allows the debris to dry (Lecomte et al. 2005). The values for these variables were kept constant throughout the study period.

The interface between forest and urban areas and between forest and agricultural areas is related to the incidence of human-caused fires (Rodrigues et al. 2020; Ganteaume et al. 2021). Values for these variables were derived for districts using spatial analysis of Corine LandCover data European Environment Agency (2019). The used variables were the lengths of the borders relative to the area of the forest (m.ha^− 1) between the forest and urban and agricultural areas, i.e., the forest-urban interface and the forest-agricultural land interface. The values for these variables were kept constant throughout the study period.

The two variables used were the number of residents per forest area (inhabitants. ha^− 1) and the number of overnight tourists per forest area represented by the number of overnight guests in recreational facilities in the district in 2012–2016 (overnight guests. ha^− 1y^− 1). Information on the total population in individual districts on 1 January 2011 and data on overnight guests was obtained from the public database of the Czech Statistical Office (2019).

Analyses

Relationships between the predictors and the number of FFs were analyzed using generalized additive models (GAM) (Wood 2017). This model predicts values of the dependent variable based on a linear combination of predictor variables that are approximated by the so-called smoother functions. The degree of smoothness of the function (the number of nodes, k) is determined separately for each predictor variable by cross-validation to avoid overfitting (over-adaptation of the function to data and loss of ability to generalize). The main results of the analysis are the deviance explained by the model, the statistical significance of the individual predictor, and the shape of the smooth function along with the Effective Degrees of Freedom, which reflect the degree of non-linearity of a curve. The mgcv library of statistical program R was used for the analysis.

Given the high number of candidate predictor variables (Appendix 1), their mutual substitutability (concurvity) was evaluated, and redundant variables were discarded. Subsequently, several models involving different predictor variables were iteratively tested, and variables that had a statistically significant association with the number of FF were used for the interpretation. The influence of first-order interactions among the variables was tested too. The quality of the created model was assessed based on the percent of deviance explained, distribution of residuals, their autocorrelation, and the correlation of predicted and measured data.

Temporal forest fire pattern

On a monthly basis, the occurrence of FFs showed two clear peaks. One peak occurred in spring; another peak occurred in summer (Fig. 1). Although the general monthly pattern in Fig. 1 is evident, there are differences between individual years (Appendix 2). For example, the summer peak was most pronounced in 2015, which was particularly warm and dry (Appendix 2). On an annual scale, the most wildfires occurred in 2015, which was the hottest and driest year in the study period. In the years 2012 and 2007, which ranked second and third in number of FFs, a high number of FFs occurred in spring. The spring of 2007 was one of the driest and hottest in the study period (Appendix 2).

Forest Fires Drivers

The regression models indicated that the number of FFs ha^− 1 of forest was significantly influenced by several predictor variables, and these effects differed depending on the time scale (annual, summer season, monthly). The used predictors explained 45 and 46% of the variability in FFs on a monthly scale and summer scale respectively, and 69% of the variability on an annual scale (Table 1).

Table 1

Results of regression models describing the relationships between numbers of forest fires per forest area (the dependent variable) and various predictors. The values indicated for predictor variables are estimated degrees of freedom. *, **, *** indicate p < 0.05, < 0.01, and < 0.001 respectively. n.s indicates not significant (p > 0.05).
Scale	R² (adj.)	Dev. explained (%)	Forest-urban Interface (m.ha^− 1)	Overnights (numbers of person ha^− 1)	Percentage of conifers (%)	Precipitation (mm)	Residents (numbers of person ha-1)	Temperature (°C)	Temperature x Precipitation
Monthly	0.449	48.7	3.385***	n.s.	1.847***	1.272***	n.s.	3.7***	11.28***
Summer	0.459	53.9	4.223***	3.033***	3.604***	n.s.	n.s.	3.202***	n.s.
Annual	0.687	71.4	3.554***	1.0***	2.946***	n.s.	2.558**	n.s.	n.s.

On annual and monthly scales, the percentage of conifers and the forest-urban interface were positively related to the occurrence of FFs; numbers of FFs showed a particularly strong relationship with the forest-urban interface (Figs. 2–3). On an annual scale, numbers of FFs were not significantly related to any of the climate variables. On the other hand, factors such as the human population (residents) and the number of nights spent in tourist facilities (overnights) were significantly associated with the number of FFs. Although FFs were positively related to the number of residents (except for the most densely populated areas), FFs were negatively related to the number of overnights (Table 1, Fig. 2).

In contrast to the occurrence of FFs on an annual scale, the occurrence of FFs on a monthly scale was significantly related to climate variables. The occurrence of FFs increased as the monthly air temperature increased above 20°C (Fig. 3c) and decreased as total month rainfall increased (Fig. 3d).

On a summer scale (i.e., considering data from summer seasons only), the predictor variables explained approximately 50% of the variability in the numbers of FFs (Table 1). Although the percentage of conifers was a significant predictor variable, the relationship was difficult to explain in that the number of FFs tended to decrease as the percentage increased from 15 to 40% but then increased as the percentage increased from 50 to 88% (Fig. 4b). Numbers of FFs increased as the length of the forest-urban interface increased (Fig. 4a). The numbers of FFs increased in response to the overnights number 0 to 25 overnight people per hectare of the forest per year (majority of the data) but decreased for higher overnight values (Fig. 4c). The number of FFs was positively associated with temperature, showing a sharper increase under temperature above 18°C (Fig. 4d).

Table 2

Summary of main trends concerning the relationships between the numbers of forest fires and predictor variables in the Czech Republic at three scales and according to Figs. 2–4. Values at which the trend changes are given in brackets.
Scale	Forest-urban Interface (m.ha^− 1)	Overnights (Numbers of of person ha^− 1 y^− 1 )	Percentage of conifers (%)	Precipitation (mm)	Residents (Numbers of inhabitants ha^− 1)	Temperature (°C)
Monthly	ꭈ (500)		↗	↘		┘ (18)
Summer	↗	∩ (25,000)	┘ (60)			┘ (18)
Annual	↗	↘	┘ (60)		∩ (30,000)

↘ downward trend

↗ upward trend

∩ unimodal response

┘The response is constant first, then rising

ꭈ the response is rising, then constant

Climatic signal in the forest fire incidence

On an annual scale, the climatic characteristics were not significantly related to the numbers of FFs; precipitation was not related to FFs numbers in summer, but the number of FFs decreased as rainfall increased on a monthly scale. Although the number of FFs in the summer months increased with temperature, the number in individual months was unrelated with temperature until the mean monthly temperature increased to 18°C; at that point, the number of FFs clearly increased as the temperature increased (Table 2).

Like the current study, many previous studies have found that the numbers of FFs are negatively related with precipitation (e.g., Engelmark 1993; Pew and Larsen 2001; Koutsias et al. 2013; Futao et al. 2016). Precipitation increases the moisture content of fuel and therefore reduces the probability of fire ignition (Flannigan and Harrington 1988; Flannigan et al. 2016). Many studies have reported that prolonged drought combined with high air temperature, high wind speed, and low relative humidity often lead to fires (e.g., Moritz 2003; McKenzie et al. 2004; Keeley 2004; Ubysz et al. 2012; 2013).

A linear relationship between number of FFs and temperature was found when the monthly temperature exceeded 18°C. Such average monthly temperatures were common for June, July, and August between 2011–2020; between 1961–1970, however, average monthly temperatures in those 3 months seldom increased to 18°C. Unal et al. (2013) found that the probability of FF occurrence increases with daily temperatures but not with monthly average temperatures. In Poland and Serbia, 60% of the FFs occurred when the daily temperature was above 24°C and 25°C, respectively, air humidity was < 40% and precipitation was lacking (Vasić 1992; Ubysz et al. 2012). This helps explain why we did not find a significant relationship between number of FFs and annual total precipitation or average annual temperature.

Effect Of Forest Conditions

The numbers of FFs were positively related to the percentage of conifers on a monthly scale or remained constant and then increased only when the share of conifers exceeded 60% on a summer and an annual scale (Table 2).

FFs are more likely to occur in coniferous forests than in deciduous forests in Mediterranean, temperate, and boreal regions (Díaz-Delgado et al. 2004; Sturtevant et al. 2004; Parisien et al. 2011). We assumed that the cover of the pines would have positive effects, but it was not confirmed in our study probably because, pine represents only 17% of the forests in the Czech Republic (Ústav pro hospodářskou úpravu lesů 2019).

According to Adámek et al. (2018), birch (Betula spp.) was one of the factors determining forest fires in the Czech Republic; based on our findings, we strongly disagree with such a conclusion. In the northern Czech Republic, we identified the tree composition of forest stands where FFs occurred in the years 2011–2015; we found that no fire occurred in birch forests (Vaněk 2017). We suggest that the finding on fires associated with birch presence is relates the inaccurate tree species data. Moreover, birch is often used to create firebreaks in young pine stands, likely causing the confusion (Ubysz et al. 2012).

Landscape-scale Drivers

The extent of the forest-urban interface was positively related with the number of FFs. As the forest-urban interface increased, the number of FFs increased on annual and summer scales and initially increased and then plateaued on a monthly scale (Table 2). This factor is partly related to heat sources. It is surprising that few studies have analysed the forest-urban interface although it was found to be behind several noticeable effects, such as winter fire progression or moderate increase in fire incidence linked to human-caused fires (e.g., Rodrigues et al. 2020). The increase in the forest-urban interface is a result of higher fragmentation of the country resulting in higher numbers of smaller woods. The greater accessibility for people (collectors of mushrooms and other wild foods or recreation) to smaller woods adjacent to houses and gardens can plays a role here. The longer boundaries between forest and the surroundings means an increased human access to the forest. Data concerning the collecting of mushrooms, which we expect to be a significant factor, are available only at the regional level, and mean values over several years are available (Šišák et al. 2016). But generally, the main mushrooms collected are Boletus spp. (Kovalčík 2014; Vasile et al. 2017), and one of the most collected species, Boletus reticulatus Schaeff. 1774, occurs in habitats affected by human activity such as parks or pond dikes (Šutara et al. 2009), i.e., in very small groups of trees.

The reason why the number of FFs does not increase as the forest-urban interface exceeds 1000 m.ha^− 1 is that the landscape is already very structured and a further increase in the forest-urban interface no longer leads to higher human traffic.

Socio-economic Drivers

Socio-economic factors have obvious relationships with ignition. In summer, numbers of FFs were not significantly associated with the numbers of local humans ha^− 1 of forest but were unimodally associated with the numbers of overnights ha^− 1 y^− 1 of the forest (Table 2). The relationship between the numbers of FFs and the size of the local human populations was also unimodal at an annual scale; FF numbers initially increased as residents increased but then decreased as human inhabitants exceeded 30,000 ha^− 1 (Table 2).

The importance of humans as igniters of FFs is obvious. Although fires are a natural phenomenon, human activity has been the leading cause of ignition in populated landscapes (Niklasson and Granström 2000; Vannière et al. 2008; Molinari et al. 2013). In Europe, 97% of FFs were directly or indirectly caused by humans (Thomas and McAlpine 2010; Ganteaume et al. 2013).

Both numbers of residents and numbers of overnight tourists have been related to the numbers of FFs in other studies (Cardille and Ventura 2001; Pew and Larsen 2001; Zumbrunnen et al. 2012; Ganteaume et al. 2013; Martınez-Fernandez et al. 2013; Futao et al. 2016; Adámek et al. 2018). In the annual scale analysis, the decline in the relationship between the number of FFs and the number of overnights following an initial increase seems illogical (Fig. 4c). The likely explanation is that the higher numbers of overnights do not indicate the real attendance of the forest; the higher numbers of overnights probably involve large cities (annual scale analysis) or visits to historical monuments, attractions, and water areas (summer scale analysis), but not the real presence of tourists in the forests.

Higher attendance of forests by humans should lead to a higher risk of ignition. In Latvia, the geographical distribution of FFs has two distinct clusters near the two largest cities, which also suggests the dominance of human causes of ignition (Donis et al. 2017). That the relationship between the number of FFs and the human population is unimodal with a maximum at around 30,000 people per ha of forests is interesting. While a smaller population means less forest attendance, larger cities should logically have more attendance. However, this may not be the case in modern society. The residents of larger cities have many ways to spend their leisure time (Shena and Wall 2021), and therefore may spend less time in nature. It may also be that human city dwellers may be more fearful of fires and thus more careful with fires.

Some forest manages have recently suggested that roads are needed to prevent fires and to keep forests healthy, but several international studies have shown that FF occurrence is negatively related to the distance to roads (Vega-Garcia et al. 1995; Goldammer 2003; Kalabokidis et al. 2002; Catry et al. 2009; Martinez et al. 2009; Arndt et al. 2013). Historical fire maps indicate, however, that forests with and without roads have burned at similar rates. The apparent neutrality of roads with respect to fire occurrence may be due to the higher rates of human-caused ignition near roads being offset by the benefits that roads provide in fighting fires (Healey 2020). The average density of forest transport roads in forests of the Czech Republic is 18.00 m.ha^− 1, which is balanced throughout the country (Žáček and Klč 2008). We do not think that the forest roads deserve substantial analysis regarding their effects on FFs in the Czech Republic. Forest roads are indirect indicators of the attendance of forests by humans. We assume that the attendance by humans has a more direct relationship than roads on numbers of FFs.

The current study revealed the possible effects of climatic, forestry, socioeconomic, and landscape factors on the number of FFs in the Czech Republic. Among the factors, the forest-urban interface, overnights, percentage of conifers, precipitation, residents, and temperature helped explain the variance in the numbers of FFs on a monthly, summer, or annual scale. Because the relationships are logical, they should improve FF forecasting and forest management.

Although humans are working to mitigate the effects of climate change on FFs, both directly by affecting the climate and indirectly by changing the structure and composition of forests or the fragmentation of the landscape, reducing the probability of wildfire ignition will require time. It is unlikely that we can easily reduce the number of FFs, in part because the conditions for the ignition of FFs will become increasingly favourable. Forest attendance, for example, will probably continue to increase due to the covid pandemic. That is why it is important that people in Central Europe realize that FFs are a problem for all Central Europe, not only the “South”. To reduce the numbers of FFs, we must change the public’s awareness of FFs. We therefore recommend that an educational program be established to make both the public and foresters more aware about how to reduce the risk of fire ignition.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests.

Funding

This contribution was supported by projects of the Czech University of Agriculture No. IGA A_20_11 and IGA A_20_08.

Authors' contributions

JH: Conceptualization; Formal analysis; Investigation; Methodology; Validation; Writing - original draft; Writing - review & editing; Supervision; RB: Formal analysis; Investigation; Methodology; TH: Conceptualization; Formal analysis; Validation; KR: Funding acquisition; Project administration; JT: Validation; Methodology; Writing - original draft; Writing - review & editing. All authors significantly contributed to this manuscript, have read and approved the manuscript in its current form.

Acknowledgements

The authors thank Bruce Jaffee (USA) for editorial and linguistic improvement of manuscript.

Abatzoglou, JT, and AP Williams. 2016. Impact of anthropogenic climate change on wildfire across western US forests. Proceedings of the National Academy of Sciences 113: 11770-11775. doi. 10.1073/pnas.1607171113
Adámek, M, Z Jankovská, V Hadincová, E Kula, and J Wild. 2018. Drivers of forest ﬁre occurrence in the cultural landscape of Central Europe. Landscape Ecology 33:2031–2045. doi. 10.1007/s10980-018-0712-2
Agee, JK. 1998. Fire and Pine Ecosystems. In Ecology and Biogeography of Pinus, ed Richardson, DM, Cambridge University Press
Albers, J. 2012. Comparative analysis of the forest fire situation in Central-Eastern Europe, Master Thesis. BOKU, Vienna, Austria
Aldersley, A, SJ Murray, SE Cornell. 2011. Global and regional analysis of climate and human drivers of wildfire. Science of Total Environment 409: 3472–3481. doi. 10.1016/j.scitotenv.2011.05.032
Arndt N, H Vacik, V Koch, A Arpaci, and H Gossow. 2013. Modeling human-caused forest fire ignition for assessing forest fire danger in Austria. iForest – Biogeosciences and Forestry 6: 315-325. doi. 10.3832/ifor0936-006
Ballabio C, P Panagos, and L Monatanarella. 2016. Mapping topsoil physical properties at European scale using the LUCAS database. Geoderma 261:110–123. doi. 10.1016/j.geoderma.2015.07.006
Balog, K, and M Kvarčák. 1999. Dynamika požáru. Ostrava: SPBI
Berčák, R, J Holuša, K Lukášová, Z Hanuška, P Agh, J Vaněk, E Kula, and I Chromek. 2018. Forest fires in the Czech Republic – characteristics, prevention and firefighting: review. Zprávy Lesnického Výzkumu 63: 184-194. Eid. 2-s2.0-85052139017
Calheiros, T, MG Pereira, and JP Nunes. 2021. Assessing impacts of future climate change on extreme fire weather and pyro-regions in Iberian Peninsula. Science of The Total Environment 754: 142233. doi. 10.1016/j.scitotenv.2020.142233
Cardille, J, and SJ Ventura. 2001. Environmental and Social Factors Influencing Wildfires in the Upper Midwest, United States. Ecological Applicationss 11: 111–127. doi. 10.2307/3061060
Catry, FX, FC Rego, FL Bacao, and F Moreira. 2009. Modelling and mapping wildfire ignition risk in Portugal. International Journal of Wildland Fire 18: 921–931. doi:10.1071/WF07123
Catry, FX, FC Rego, JS Silva, F Moreira, A Camia, C Ricotta, and M Conedera. 2010. Fire Starts and Human Activities, 9-22. In Towards Integrated Fire Management – Outcomes of the European Project Fire Paradox, ed. JS Silva, FS Rego, P Fernandes, E. European Forest Institute Report 23: 1-228.
Chytrý, M. 2012. Vegetation of the Czech Republic: diversity, ecology, history and dynamics. Preslia 84:427–504
Clark J, and J Merkt. 1989. Post-glacial fire, vegetation, and human history on the northern alpine forelands, south-western Germany. Journal of Ecology 77:897–925
Clark, JS, and PD Royall. 1996. The Role of Fire During Climate Change in an Eastern Deciduous Forest at Devil’s Bathtub, New York. Ecology 77, 2148–2166.
Corona, P, B Ferrari, R Cartisano, A Barbati. 2014. Calibration assessment of forest flammability potential in Italy. Iforest - Biogeosciences and Forestry 7: 300–305. doi. 10.3832/IFOR1123-007
Czech Hydrometeorological Institution. 2019. http://portal.chmi.cz/ Accessed 1 Apr 2019
Czech Statistical Office. 2019. https://www.czso.cz/ Accessed 1 Apr 2019
Díaz-Delgado, R, F Lloret, and X Pons. 2004. Spatial patterns of fire occurrence in Catalonia, NE, Spain. Landscape Ecology 19: 731–745. doi. 10.1007/s10980-005-0183-1
Donis, J, M Kitenberga, G Snepsts, R Matisons, J Zarins, and A Jansons. 2017. The forest fire regime in Latvia during 1922–2014. Silva Fennica 51: 7746. doi. 10.14214/sf.7746
Donoghue, LR, WA Main. 1985. Some factors influencing wildfire occurrence and measurement of fire prevention effectiveness. Journal of Environmental Management 20: 87–96.
Dupire, S, T Curt, S Bigot, and T Fréjaville. 2019. Vulnerability of forest ecosystems to fire in the French Alps. European Journal of Forest Research 138: 1–18. doi. 10.1007/s10342-019-01206-1
EAGRI. 2019. http://eagri.cz/public/web/mze/ Accessed 1 Apr 2019
Elia, M, M D'Este, D Ascoli, V Giannico, G Spano, A Ganga, G Colangelo, R Lafortezza, and G Sanesi. 2020. Estimating the probability of wildfire occurrence in Mediterranean landscapes using Artificial Neural Networks. Environmental Impact Assessment Review 85: 106474. doi. 10.1016/j.eiar.2020.106474
Elia, M, V Giannico, G Spano, R Lafortezza, and G Sanesi. 2020. Likelihood and frequency of recurrent fire ignitions in highly urbanised Mediterranean landscapes. International journal of wildland fire 29: 120-131. doi. 10.1071/WF19070
Ellenberg, H. 1996. Vegetation Mitteleuropas mit den Alpen in okologischer, dynamischer und historischer Sicht. Stuttgart: Ulmer
Emmons, HW. 1973. Heat transfer in Fire. Journal of Heat Transfer 95:145-151. doi.org/10.1115/1.3450012
Engelmark, O. 1987. Fire history correlations to forest type and topography in northern Siberia. Annales Botanici Fennici 24: 317–324.
Engelmark, O. 1993. Early post-fire tree regeneration in a Picea-Vaccinium Forest in Northern Sweden. Journal of Vegetation Science 4:791–794. doi.org/10.2307/3235616
European Environment Agency. 2018. https://www.eea.europa.eu/data-and-maps/indicators/forest-fire-danger-2/assessment Accessed 1 Apr 2019
Field, RD, GR van der Werf, T Fanin, EJ Fetzer, R Fuller, H Jethva, R Levy, NJ Livesey, M Nathaniel, M Luo, O Torres, and HM Worden. 2016. Indonesian fire activity and smoke pollution in 2015 show persistent nonlinear sensitivity to El Nino-induced drought. Proceedings of The National Academy of Sciences of The United States of America 113: 9204-9209. doi.10.1073/pnas.1524888113
Flannigan, MD, and JB Harrington. 1988. A study of the relation of meteorological variables to monthly provincial area burned by wildfire in Canada 1953–80. Journal of Applied Meteorology 27:441-452. doi. 10.1175/1520-0450(1988)027<0441:ASOTRO>2.0.CO;2
Flannigan, MD, BM Wotton, GA Marshall, WJ de Groot, J Johnston, N Jurko, and AS Cantin. 2016. Fuel moisture sensitivity to temperature and precipitation: climate change implications. Climatic Change 134:59–71. doi. 10.1007/s10584-015-1521-0
Flannigan, MD, KA Logan, BD Amiro, WR Skinner, and BJ Stocks. 2005. Future area burned in Canada. Climatic Change 72(1):1-16. doi. 10.1007/s10584-005-5935-y
Futao, G, Z Su, G Wang, L Sun, F Lin, and A Liu. 2016. Wildfire ignition in the forests of southeast China: Identifying drivers and spatial distribution to predict wildfire likelihood. Applied Geography 66: 12-21. doi.org/10.1016/j.apgeog.2015.11.014
Ganteaume, A, A Camia, M Jappiot, J San-Miguel-Ayanz, M Long-Fournel, and C Lampin. 2013. A review of the main driving factors of forest fire ignition over Europe. Environmental Management 51:651–662. doi. 10.1007/s00267-012-9961-z
Ganteaume, A, R Barbero, M Jappiot, and E Maillé. 2021. Understanding future changes to fires in southern Europe and their impacts on the wildland-urban interface. Journal of Safety Science and Resilience 2: 20-29. doi.org/10.1016/j.jnlssr.2021.01.001
Goldammer, JG. 2003. Towards international cooperation in managing forest fire disasters in the Mediterranean region. In Workshop on Security and the Environment in the Mediterranean in the 20th Century held at the 4th Pan-European Conference. Security and Environment in the Mediterranean, ed Brauch, HG, PH Liotta, A Marquina, PF Rogers, ME Selim. Canterbury: Univ Kent
Grissino-Mayer, HD, WH Romme, ML Floyd, and DD Hanna. 2004. Climatic and human influences on fire regimes of the southern San Juan Mountains, Colorado, USA. Ecology 85: 1708-1724. doi. 10.1890/02-0425
Hart, SJ, TT Veblen, N Mietkiewicz, and D Kulakowski. 2015. Negative Feedbacks on Bark Beetle Outbreaks: Widespread and Severe Spruce Beetle Infestation Restricts Subsequent Infestation. Plos one 10: e0127975. doi. 10.1371/journal.pone.0127975
Hasičský záchranný sbor České republiky. 2019. https://www.hzscr.cz/ Accessed 1 Apr 2019
Healey, SP. 2020. Long-term forest health implications of roadlessness. Environmental Research Letters 15: 104023. doi. 10.1088/1748-9326/aba031
Hicke, JA, MC Johnson, LHD Jane, and HK Preisler. 2012. Effects of bark beetle-caused tree mortality on wildfire. Forest Ecology and Management 271: 81-90. doi. 10.1016/j.foreco.2012.02.005
Holuša, J, R Berčák, K Lukášová, Z Hanuška, P Agh, J Vaněk, E Kula, and I Chromek. 2018. Forest fires in the Czech Republic – definition and classification: review. Zprávy Lesnického Výzkumu, 63: 20-27
Keeley, JE. 2004. Impact of antecedent climate on fire regimes in coastal California. International Journal of Wildland Fire 13: 173–182. doi.org/10.1071/WF03037
Koutsias, N, G Xanthopoulos, D Founda, F Xystrakis, F Nioti, M Pleniou, G Mallinis, and A Arianoutsou. 2013. On the relationship between forest fires and weather conditions in Greece from long-term national observations (1894–2010). International Journal of Wildland Fire 22: 493–507. doi. 10.1071/WF12003
Kovalčík, M. 2014. Value of forest berries and mushrooms picking in Slovakia ́s forests. Beskydy 7: 39–46
Lecomte, N, M Simard, Y Bergeron, A Larouche, H Asnong, and PJH Richard. 2005. Effects of fire severity and initial tree composition on understorey vegetation dynamics in a boreal landscape inferred from chronosequence and paleoecological data. Journal of Vegetation Science 16: 665–674. doi. /10.1111/j.1654-1103.2005.tb02409.x
Martinez J, C Vega-Garcia, and E Chuvieco. 2009. Human-caused wildfire risk rating for prevention planning in Spain. Journal of Environmental Management 90: 1241-1252. doi. 10.1016/j.jenvman.2008.07.005
Martinez-Fernandez J, E Chuvieco, N Koutsias. 2013. Modelling long-term fire occurrence factors in Spain by accounting for local variations with geographically weighted regression. Natural Hazards and Earth System Sciences 13:311–327. doi. 10.5194/nhess-13-311-2013
McKenzie, D, Z Gedalof, DL Peterson, and P Mote. 2004: Climatic change, wildfire, and conservation. Conservation Biology 18: 890–902. doi. /10.1111/j.1523-1739.2004.00492.x
Mohammadi Z, P Lohmander, J Kašpar, R Berčák, J Holuša, R Marušák R. 2021. The effect of climate factors on the size of forest wildfires (case study: Prague‑East district, Czech Republic). Journal of Forestry Research 33: 1391-1400. doi. 10.1007/s11676-021-01413-w
Molinari, C, V Lehsten, RHW Bradshaw, MJ Power, P Harmand, A Arneth, JO Kaplan, B Vannière, and MT Sykes. 2013. Exploring potential drivers of European biomass burning over the Holocene: a datamodel analysis. Global Ecology and Biogeography 22: 1248–1260. doi. /10.1111/geb.12090
Moreira, F, FC Rego, and PG Ferreira. 2001. Temporal (1958-1995) pattern of change in a cultural landscape of northwestern Portugal: Implications for fire occurrence. Landscape Ecology 16: 557–567. doi. 10.1023/A:1013130528470
Moritz, MA. 2003. Spatio-temporal analysis of controls on shrubland fire regimes: age dependency and fire hazard. Ecology 84: 351–361. doi: 10.1890/0012-9658(2003)084[0351:SAOCOS]2.0.CO;2
Müller, MM, L Vilà-Vilardell, and H Vacik. 2020. Towards an integrated forest fire danger assessment system for the European Alps. Ecological Informatics 60: 101151. doi. 10.1016/j.ecoinf.2020.101151
Niklasson M, and A Granström. 2000. Numbers and sizes of fires: long-term spatially explicit fire history in a Swedish boreal landscape. Ecology 81:1484–1499. doi.v10.1890/0012-9658(2000)081[1484:NASOFL]2.0.CO;2
Niklasson M, E Zin, T Zielonka, M Feijen. 2010. A 350-year tree-ring fire record from Bialowieza Primeval Forest, Poland: implications for Central European lowland fire history. Journal of Ecology 98.6.:1319–1329. Doi. 10.1111/j.1365-2745.2010.01710.x
Osvald, A. 1997. Požiarnotechnické vlastností dreva a materiálov na báze dreva. Vedecké štúdie 8/97/A, Zvolen: TU vo Zvolene
Parisien, MA, SA Parks, MA Krawchuk, MD Flannigan, LM Bowman, and MA Moritz. 2011. Scale-dependent controls on the area burned in the boreal forest of Canada, 1980-2005. Ecological Applications 21: 789–805. doi. 10.2307/23021627
Parks, SA, LM Holsinger, MH Panunto, WM Jolly, SZ Dobrowski, and GK. 2018. High-severity fire: evaluating its key drivers and mapping its probability across western US forests. Environmental research letters 13: 044037. doi. 10.1088/1748-9326/aab791
Pew, K, and CP Larsen. 2001. GIS analysis of spatial and temporal patterns of human-caused wildfires in the temperate rain forest of Vancouver Island, Canada. Forest Ecology and Management 140: 1–18. doi. 10.1016/S0378-1127(00)00271-1
Pokorná A, V Štrombachová, J Mužík, D Dolanová, P Búřilová, M Pospíšil, J Kučerová, J Gregor, M Komenda, L Dušek. 2019. Národní portál Systém hlášení nežádoucích událostí. Praha: Ústav zdravotnických informací ČR, 2016. Accessed 20 Dec 2019.
Rodrigues M, A Jiménez-Ruano, and J la Riva. 2020. Fire regime dynamics in mainland Spain. Part 1: Drivers of change. Science of the Total Environment 721: 135841. doi. 10.1016/S0378-1127(00)00271-1
Roy PS. 2003. Forest fire and degradation assessment using satellite remote sensing and Geographic Information System. In: Satellite Remote Sensing and GIS Applications in Agricultural Meteorology. Proceedings of a Training Workshop held 7–11 July 2003 in Dehra Dun, India, ed MVK Sivakumar. Geneva: World Meteorological Organisation
Shena H, G Wall. 2021. Social media, space and leisure in small cities. Asia Pacific Journal of Tourism Research 26: 73-80. doi. /10.1080/10941665.2020.1859057
Sisak, L, M Riedl, R Dudik. 2016. Non-market non-timber forest products in the Czech Republic—Their socio-economic effects and trends in forest land use. Land Use Policy 50: 390–398. doi. /10.1016/j.landusepol.2015.10.006
State Administration of Land Surveying and Cadastre. 2019. https://www.cuzk.cz/en Accessed 20 Dec 12.2021.
Sturtevant, BR, PA Zollner, EJ Gustafson, and DT Cleland. 2004. Human influence on the abundance and connectivity of high-risk fuels in mixed forests of northern Wisconsin, USA. Landscape Ecology 19, 235–254. doi. 10.1023/B:LAND.0000030420.87498.f7
Šutara, J, M Mikšík, and V Janda. 2009. Hřibovité houby [Mushroom-forming fungi]. Praha: Academia
Tanskanen, H. 2007. Fuel conditions and fire behavior characteristics of managed Picea abies and Pinus sylvestris forests in Finland. Dissertationes Forestales 40: 1-51. doi. 10.14214/df.40
Thomas, EA, and RS McAlpine. 2010. Fire in the Forest. New York: Cambridge University Press.
Tinner, W, M Conedera, B Ammann, and AF Lotter. 2005. Fire ecology north and south of the Alps since the last ice age. The Holocene 8:1214–1226. doi. /10.1191/0959683605hl892
Tolasz, R, T Míková, A Valeriánová, V Voženílek. 2007. Atlas podnebí Česka [Climate atlas of Czechia]. Olomouc: Palacky Univerzity and Czech Hydrometeorogical Institute
Turetsky, MR, B Benscoter, S Page, G Rein, GR van der Werf, and A Watts. 2015. Global vulnerability of peatlands to fire and carbon loss. Nature geoscience 8: 11-14. doi10.1038/NGEO2325
Ubysz B, and JC Valette. 2010. Flammability: Influence of Fuel on Fire. In: Towards Integrated Fire Management – Outcomes of the European Project Fire Paradox. Ed JS Silva, F Rego, P Fernandes, E Rigolot. European Forest Institute Report 23: 1-228.
Ubysz B, R Szczygiel, M Kwitkowski, J Piwnicki. 2012. Intrukcja ochrony przeciwpozarowej lasu. Warszawa: Panstwowe gospodarstwo lesne, Lasy panstwowe.
Unal, YS, E Tan, and SS Mentes. 2013. Summer heat waves over western Turkey between 1965 and 2006. Theoretical and Applied Climatology 112:339–350. doi. 10.1007/s00704-012-0704-0
Ústav pro hospodářskou úpravu lesů. 2019. http://www.uhul.cz/ Accessed 1 Apr 2019
Vaněk, J. 2017: Analýza parametrů lesních požárů na území Libereckého kraje [Analysis of parameters of forest fires in the territory of the Liberec region]. Bachelor thesis. Praha: Česká zemědělská univerzita v Praze.
Vanická H, J Holuša, K Resnerová, J Ferenčík, M Potterf, A Véle, and W Grodzki. 2020. Interventions have limited effects on the population dynamics of Ips typographus and its natural enemies in the Western Carpathians (Central Europe). Forest Ecology and Management 470: 118209. doi./10.1016/j.foreco.2020.118209
Vannière, B, D Colombaroli, E Chapron, A Leroux, W Tinner, and M Magny. 2008. Climate versus humandriven fire regimes in Mediterranean landscapes: the Holocene record of Lago dell’Accesa (Tuscany, Italy). Quaternary Science Reviews 27: 1181–1196. doi. /10.1016/j.quascirev.2008.02.011
Vasić, M. 1992. Wildfires. Belgrade: University of Belgrade, Faculty of Forestry.
Vasile, D, L Dinca, and CM Enescu. 2017. Impact of collecting mushrooms from the spontaneous flora on forest ecosystems in Romania. AgroLife Scientific Journal 6: 268-275.
Weinbrenner, M, R Weber, T Neumeyer, and V Altstadt. 2022. Influence of foam density on burning behavior of polymer bead foams. Journal of Applied Polymer Science 139: e51558. doi. 10.1002/app.51558
Westerling, AL. 2016. Increasing western US forest wildfire activity: sensitivity to changes in the timing of spring. Philosophical Transactions of the Royal Society B-Biological Sciences 371: 20150178. doi. doi. 10.1098/rstb.2015.0178
Wood, SN. 2017. Generalized Additive Models. New York: Chapman and Hall/CRC doi. 10.1201/9781315370279
Žáček J, and P Klč. 2008. Forest transport roads according to natural forest regions in the Czech Republic. Journal of Forest Science 54: 73–83.
Zachar, M. 2009. Vplyv ohrevu na termickú degradáciu vybraných druhov dreva [Influence of thermal loading at spruce wood lignin alteration]. Zvolen: Technická univerzita vo Zvolene.
Zumbrunnen, T, P Menéndez, H Bugmann, M Conedera, U Gimmi, and M Bürgi. 2012. Human impacts on fire occurrence: A case study of hundred years of forest fires in a dry alpine valley in Switzerland. Regional Environmental Change 12: 935–949. doi. 10.1007/s10113-012-0307-4

Appendix.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Forest fire occurrence shows clear relationships with the numbers of nearby inhabitants and tourists in the Czech Republic

Status:

Version 1

Abstract

Background

Results

Conclusions

Figures

Background

Methods

Study area

Forest Fire Database

The Dependent Variable

Predictor Variables

Analyses

Results

Temporal forest fire pattern

Forest Fires Drivers

Discussion

Climatic signal in the forest fire incidence

Effect Of Forest Conditions

Landscape-scale Drivers

Socio-economic Drivers

Conclusion

Declarations

References

Supplementary Files

Status:

Version 1